Professional Documents
Culture Documents
Handbook - of - Individual Differences in Social Behavior
Handbook - of - Individual Differences in Social Behavior
Handbook - of - Individual Differences in Social Behavior
IN SOCIAL BEHAVIOR
i
Handbook of
In d i v i d u a l
D iff e r e n c e s
in S o c i a l
B e h a v io r
■ • ▲ ♦ ■ • A *
edited by
Mark R . Leary
Rick H. Hoyle
No part of this book may be reproduced, translated, stored in a retrieval system, or transmitted,
in any form or by any means, electronic, mechanical, photocopying, microfilming, recording,
or otherwise, without written permission from the Publisher.
Handbook of individual differences in social behavior / edited by M ark R. Leary, Rick H. Hoyle,
p. cm.
Includes bibliographical references and index.
ISBN 9 7 8 - 1 -5 9 3 8 5 - 6 4 7 -2 (hardcover)
1. Social psychology. 2. Human behavior. I. Leary, M ark R. II. Hoyle, Rick H.
H M 1033.H 355 20 0 9
3 0 2 . 5 '4 — dc22
2008053706
About the Editors
I
I
I
p
Contributors
V anessa T . A nderson, PhD, D epartm ent o f Psychology, C olum bia University, N ew Y ork, N ew York
R o b ert F. B ornstein, PhD, D erner Institute, Adelphi University, G arden City, N ew York
Jen n ifer K . B o sso n , PhD, D epartm ent o f Psychology, University o f South F lorida, T am pa, Florida
Stephen H . Boyle, PhD, Behavioral M edicine R esearch Center, D epartm ent o f Psychiatry,
D uke University, D urham , N orth C arolin a
Pablo B rin ol, PhD, D epartm ent o f Social Psychology, U niversidad A utonom a de M ad rid ,
M ad rid , Spain
Joh n T . C acio p p o , PhD , Center for C ognitive and Social N euroscien ce and D epartm ent of
Psychology, University o f C h icago, C h icago, Illinois
C h arles S. C arver, PhD, D epartm ent o f Psychology, U niversity o f M iam i, C oral G ab les, Florida
D avid E. C onroy, PhD , D epartm ents of K in esiology and H um an D evelopm ent and Fam ily Studies,
Pennsylvania State University, University Park, Pennsylvania
Susan E. C ro ss, PhD , D epartm ent o f Psychology, Iow a State University, A m es, Iowa
C lau d ia D albe rt, PhD, D epartm ent of E ducation al Psychology, M artin Luther University
o f H alle-W ittenberg, H alle, G erm any
G eraldine Downey, PhD, D epartm ent o f Psychology, C olum bia University, N ew York, N ew York
Jo h n D uck itt, PhD , D epartm ent o f Psychology, University o f A uckland, A uckland, N ew Z ealan d
A lice H . Eagly, PhD , D epartm ent o f Psychology, N orth w estern University, Evanston, Illinois
Andrew J . Elliot, PhD, D epartm ent o f C linical and Social Sciences in Psychology, University
o f R ochester, Rochester, N ew York
A llan Fenigstein, PhD, D epartm ent o f Psychology, Kenyon C ollege, G am bier, O hio
viii C o n trib u to rs
Patrick H . Finan, M A , D epartm ent o f Psychology, A rizona State University, Tem pe, A rizona
Shira Fishm an , M A , D epartm ent o f Psychology, University o f M ary lan d , C ollege Park, M arylan d
Eugene M . Fodor, PhD , D epartm ent o f Psychology, C lark son University, P otsdam , N ew York
Paul T. F uglestad, B A , D epartm ent o f Psychology, University o f M in n eso ta, M in n eap olis, M in n esota
A drian Furn ham , D Sc, D Phil, D L itt, D epartm ent o f Psychology, University C ollege L ondon,
London, United K ingdom
Erin E . H ard in , PhD, D epartm ent o f Psychology, T exas Tech University, L u b b ock , T exas
Louise C . H aw kley, PhD , Center for C ognitive and Social N euroscien ce and D epartm ent
o f Psychology, University o f C h icago, C hicago, Illinois
C raig A. H ill, PhD , D epartm ent o f Psychology, Indiana U niversity-P urdue University Fort W ayne,
F ort W ayne, Indiana
R on ald R . H olden, PhD , D epartm ent o f Psychology, Q ueens University, K in gsto n , O n tario, C an ad a
R ick H . H oyle, PhD, D ep artm en t of Psychology and N euroscien ce, D uke University, D urham ,
N orth C arolin a
K ristin e M . Kelly, PhD, D epartm ent o f Psychology, W estern Illinois University, M acom b , Illinois
R an dy J . L arsen , PhD, D epartm ent o f Psychology, W ashington University in St. L o u is, St. Lo uis,
M issou ri
M a rk R . Leary, PhD , D epartm ent o f Psychology and N euroscien ce, D uke University, D urham ,
N o rth C arolin a
C h ris Loersch , M A , D epartm ent o f Psychology, O hio State University, C olum b u s, O hio
M ich ael J . M cC aslin , M A , D epartm ent o f Psychology, O hio State University, C olum b u s, O hio
R ow lan d S. M iller, PhD , D epartm ent o f Psychology and Philosophy, Sam H ouston State University,
H untsville, T exas
K ristin N eff, PhD, D epartm ent o f E ducation al Psychology, U niversity o f T exas at A ustin,
A ustin, T exas
Ju lie K . N orem , PhD, D epartm ent o f Psychology, Wellesley C ollege, Wellesley, M assach u setts
Jen n ifer Passey, M A , D epartm ent o f Psychology, Q ueens University, K in gston, O n tario, C an ad a
C o n trib u to rs ix
Delroy L . Paulhus, PhD, D epartm ent o f Psychology, University of British C olum b ia, V ancouver,
British C olum bia, C an ad a
B enjam in Peterson, PhD, D epartm ent o f Psychology, University o f U tah, Sait Lake City, U tah
R ich ard E. Petty, PhD, D epartm ent o f Psychology, O hio State University, C olum b u s, O hio
K avita S. Reddy, M A , D epartm ent o f Psychology, C olum bia University, N ew Y ork, N ew York
W illiam Revelle, PhD , D epartm ent o f Psychology, N orthw estern University, Evanston, Illinois
Frederick R h odew alt, PhD, D epartm ent o f Psychology, University o f U tah, Salt l ake City, U tah
Brent W. R o b erts, PhD, D epartm ent o f Psychology, University of Illinois at U rb an a-C h am p aign ,
C h am p aign , Illinois
R ain er R o m ero-C an yas, PhD, D epartm ent of Psychology, C olum bia University, N ew York,
N ew York
Phillip R . Shaver, PhD, D epartm ent o f Psychology, University o f C alifo rn ia D avis, D avis, C aliforn ia
M ark Snyder, PhD, D epartm ent o f Psychology, University of M in n eso ta, M in n eap olis, M in n esota
Jeffrey Stuew ig, PhD, D epartm ent o f Psychology, G eorge M aso n University, F airfa x , V irginia
Peter Suedfeld, PhD , D epartm ent o f Psychology, University of British C olum b ia, Vancouver,
British C olum bia, C an ad a
A ngelina R . Sutin, PhD, L ab orato ry o f Personality and C ogn ition , N atio n al Institute on A ging,
N atio n al Institutes of H ealth, Baltim ore, M arylan d
W illiam B. Sw ann, Jr., PhD, D epartm ent of Psychology, University o f T exas at A ustin, A ustin, T exas
B ern a G ercek Sw ing, M S , D epartm ent o f Psychology, Iowa State University, A m es, Iowa
Ju n e Price Tangney, PhD, D epartm ent o f Psychology, G eorge M ason University, F airfa x , V irginia
Renee M . T obin , PhD, D epartm ent o f Psychology, Illinois State University, N o rm al, Illinois
W illiam Tov, PhD, School o f Social Sciences, Singapore M an agem en t University, Singapore
T h o m as A. W idiger, PhD, D epartm ent o f Psychology, University of Kentucky, L exin gton , Kentucky
Wendy W ood, PhD , D epartm ent of Psychology and N euroscience, D uke University, D urham ,
N orth C arolin a
K erstin Y oum an , M A , D epartm ent o f Psychology, G eorge M ason University, F airfa x , V irginia
A lex J . Z au tra, PhD, D epartm ent o f Psychology, A rizona State University, Tem pe, A rizona
M arv in Z uck erm an , PhD, D epartm ent of Psychology, University o f D elaw are, N ew ark , D elaw are
Preface
view of what is known about a particular personality variable. Our goal in edit
ing the H an d book o f Individual D ifferences in Social B ehavior was to provide
a relatively comprehensive examination of nearly 40 personality variables that
have been studied by behavioral researchers. Dozens, if not hundreds, of person
ality attributes have been studied, and, by necessity, we had to be selective. But
we hope that the H an d b ook includes the personality variables that researchers
currently find to be the most interesting, important, and useful. Some of these
variables— such as extraversion, neuroticism, and achievement motivation— have
been studied extensively for decades. Other variables— such as desire for control,
self-compassion, and embarrassability— have been examined in fewer studies. The
literature on each is summarized by an acknowledged expert— the researcher who
either first popularized the construct, developed a commonly used measure of it,
or conducted influential research.
Although the study of individual differences is sometimes regarded as the pur
view of personality psychology, researchers cutting a wide swath across the social
and behavioral sciences have in fact shown an interest in dispositional factors.
Individual differences have been studied not only by social and personality psy
chologists, but also by researchers in developmental, clinical, educational, coun
seling, health, organizational, political, cognitive, and sport psychology, as well
as by researchers in marketing, management, law, education, political science,
sociology, psychiatry, nursing, and social work. We hope that this book serves
as an invaluable resource for scholars and students across all areas of social and
behavioral science.
R eferen ce
PART I. INTRODUCTION
1 . Situ ation s, D isp o sition s, and the Stu d y o f S o cial B eh avior 3
M a r k R . L e a r y and R i c k H . H o y l e
4 . A greeablen ess 46
W i l l i a m G . G r a z i a n o and R e n e e M . T o b i n
5 . A ttach m en t Styles 62
P h i l l i p R . S h a v e r and M a r i o M i k u l i n c e r
6 . In terperson al D ep en d en cy 82
R obert F. B o r n s t e i n
8 . G e n d er Identity 109
W e n d y W o o d and A l i c e H . E a g l y
I I . D ep ression 161
P a t r i c k H . F i n a n , H o w a r d T e n n e n , and A l e x J . Z a u t r a
xiii
xiv C o n ten ts
2 7 . B e lo n g in g M o tivation 400
M a r k R . L e a r y and K r i s t i n e M . K e l l y
3 1 . Sensation S e e k in g 455
M a rv in Z u c k e r m a n
3 3 . Psychological D efensiveness: R e p ressio n , B lu n tin g, and D efen sive P essim ism 480
J ulie K. N orem
3 5 . In dependen t, R e la tio n a l, and C o lle ctiv e-In terd ep en d e n t S elf-C o n stru a ls 512
S u s a n E. C r o s s , E r i n E. H a r d i n , and B e r n a G e r c e k S w i n g
3 7 . N arcissism 547
F r e d e r i c k R h o d e w a l t and B e n j a m i n P e t e r s o n
In t r o d u c t i o n
i
i
Ii
I
CHAPTER 1
Situations, Dispositions,
and the Study of Social Behavior
M a rk R . Lea ry
R ic k H . H o y l e
3
4 I. IN T R O D U C T IO N
tions and intrapsychic processes. But part gists considered the research of their social-
of the rift also stemmed from differences psychological colleagues, they suspected
in the prevailing research paradigms that that the effects of any particular experimen
dominated social and personality psychol tal treatment were rarely, if ever, obtained
ogy during most of the 20th century. Social on every participant in the experiment. The
psychologists relied primarily on experimen findings in social psychology were based on
tal methods in which participants were ran comparisons of the means of participants’
domly assigned to experimental conditions responses in various experimental condi
that consisted of variations in the social tions, and such comparisons obscured the
situation. In fact, the classic experiments fact that participants in a particular condi
in social psychology— many of them con tion differed among themselves in how they
ducted by founders of the field such as Asch, responded to the independent variable. Be
Sherif, Schachter, Festinger, and M ilgram — cause social psychologists were interested
exemplified the experimental approach and primarily in between-group situational ef
provided the dominant research paradigm fects, these individual differences were rel
for future generations of social psychologists egated to the error term of their statistical
that exists to this day. tests— with no effort to determine why some
The social psychologists’ reliance on ex people reacted differently to the independent
perimental studies, typically conducted variable than others.
under controlled laboratory conditions, When Lee Cronbach gave his presidential
emerged not only from their inherent inter address to the American Psychological Asso
est in situational influences on behavior but ciation in 1957, he addressed this entrenched
also from a philosophy of science that viewed schism between researchers who rely on ex
experimental work as inherently more “sci perimental versus correlational approaches.
entific” than other research approaches. Be Cronbach observed that the field of psy
cause research that relied solely on descrip chology had fractured into two distinct
tive or correlational methods, in which there disciplines— one defined by the experimen
is no random assignment of participants to tal method and the other defined by corre
conditions controlled by the researcher, was lational approaches. Furthermore, he noted
unable to draw firm conclusions about the that psychology was severely “limited by the
causal relationships between purported an dedication of its investigators to one or the
tecedents and consequences, it was viewed other method of inquiry rather than to scien
as less rigorous, definitive, and thus “scien tific psychology as a whole” (1957, p. 671),
tific” than research that employed true ex even though the primary difference between
periments. The experimentalists’ perspective those who subscribe to each of these “two
on the proper way to conduct psychological disciplines of scientific psychology” merely
research led social psychologists to look concerned whether the variability that they
askance at much of the research in person sought to explain preexists in the world or
ality psychology, which relied primarily on is created through experimental manipula
descriptive and correlational methods. tions.
The correlational tradition had a long and
respected history from the earliest days of
psychometrics and differential psychology The P erson-Situation Debate
in the early part of the 20th century. Galton,
Pearson, Spearman, and others made con Differences in scientific approach were ac
tributions to the development of statistical companied by lively debates regarding the
methods that could be used to understand relative importance of situational versus
relationships among existing variables (e.g., dispositional factors in understanding be
correlation and factor analysis) and then havior. Although a foreshadowing of this
used these methods to study individual dif controversy can be seen earlier (Ichheisser,
ferences in intellectual, psychological, and 1943), the opening volley was fired in Mis-
physical characteristics. But as social psy chel’s (1968) critique of the trait concept
chologists looked at results from correla (and of personality psychology more gener
tional research, they saw mostly confounded ally). After critically reviewing 50 years of
variables and indefinite causal conclusions. research that showed only small correlations
Conversely, when personality psycholo in people’s behaviors across situations and
1. S itu ation s, D isp osition s, and the Study o f S ocial B eh av io r 5
time, Mischel concluded that “highly gener strength of situational effects on behavior
alized behavioral consistencies have not been were comparably low. In an early documen
demonstrated and the concept of personality tation of this point, Funder and Ozer (1983)
traits as broad response dispositions is thus calculated the correlation effect sizes for
untenable” (p. 145). His recommendation some well-known situational effects in so
that psychologists abandon their efforts to cial psychology (including classic studies of
explain behavior with traits and focus in forced compliance, bystander intervention,
stead on situations was embraced by many and obedience) and found that all fell under
social psychologists. For example, after pos .40. Other researchers have documented
ing a scenario asking the reader to predict the same conclusion, suggesting that the
whether a hypothetical person, John, will strength of the relationships between mea
help someone he sees slumped in a doorway, sured dispositions and behavior are compa
Ross and Nisbett (1991) concluded: rable to those between situational manipula
tions and behavior.
A h alf century of research h as taugh t us that Second, Epstein (1979, 1983) noted that a
in this situation , and in m ost other novel situ single measure of behavior is not a reliable
ations, one can n ot predict w ith a n y a c c u r a c y indicator of a person’s general behavioral
how p articu lar people will behave. At least
tendencies. As a result, the magnitude of
one can n ot do so using inform ation about
an individual’s p ersonal disposition s or even correlations between measures of person
abou t that in d ivid ual’s p ast b e h a v io r.. . . W hile ality and specific behaviors are attenuated
know ledge about Joh n is o f surprisingly little by measurement error, which lowers the
value in predicting whether he will help the strength of statistical effects. When behav
person slum ped in the doorw ay, details con iors are aggregated across situations (just as
cerning the specifics o f the situation w ould be self-report responses are aggregated across
invaluable, (pp. 2 - 3 , em phasis added) the items on a personality questionnaire),
behavioral measures are more reliable, cor
The effects of Mischel’s (1968) book were relations are notably larger, and personality
powerful and immediate. For example, the does a better job of predicting behavior.
percentage of articles published in the Jo u r Third, research began to show that per
nal o f Personality and Social Psychology sonality relates more strongly to behavior in
(JPSP) that included any reference to indi some situations than in others. In “strong”
vidual differences, whether alone or in com situations that constrain people’s behavior
bination with experimental manipulations, and provide clear cues regarding how people
dropped from 50% to about 30% from 1966 should behave, most people tend to act simi
to 1977 (Swann & Seyle, 2005). During the larly. In contrast, when “weak,” unstruc
same time period, the percentage of articles tured, or novel situations offer few cues or
in JPSP that reported purely experimental norms to guide behavior, large individual
studies rose from about 50% to nearly 70% . differences emerge (Caspi & Moffitt, 1993;
At the same time, a great deal of time, Ickes, 1982). Importantly, the laboratory
energy, and ink were devoted to analyzing situations that researchers create to study
Mischel’s (1968) claims more deeply. This individual differences typically constrain be
discussion led to four important conclusions havior (and thus the manifestation of traits)
about the respective influences of situational because they are rigidly controlled, often
and dispositional factors and the relation with independent variables that are inten
ships among them (for overviews, see Bern & tionally designed to exert a strong influence
Funder, 1978; Cervone, Caldwell, & Orom, on people’s behavior. Even when situations
in press; Epstein & O ’Brien, 1985; Kenrick are strong, however, we often still see indi
& Dantchik, 1983; Magnusson & Endler, vidual differences. Even in experiments with
1977; Snyder & Ickes, 1985). powerful manipulations, such as M ilgram ’s
First, Mischel’s (1968) most damning ar (1963) studies of obedience to authority,
gument was that correlations between mea large individual differences in the degree
sures of personality and of behavior (and to which participants disobeyed the experi
between measures of behavior collected on menter were observed (Packer, 2008).
different occasions) typically hover around Fourth, theorists noted a fallacy in the rea
.30, seeming to reflect a very weak relation soning of those, including Mischel (1968),
ship. However, evidence emerged that the who used small personality-behavior corre
6 I. IN T R O D U C T IO N
lations to argue that situational factors play ences and behavior was independent of the
a more powerful role in behavior than per relationship between personality influences
sonality. They noted that the fact that per and behavior. Fleeson (2001, 2004) similar
sonality and behavior tend to correlate .30 ly showed that strong cross-situational con
does not indicate that any of the remaining sistencies in people’s modal or typical level
variance is produced by the situation. Per of a trait are, at the same time, accompanied
haps more important, they pointed out that by large variability in their reactions across
the strength of situational and dispositional different situations.
effects are not inversely related to one an Considerations such as these not only
other, as one might assume. Contrary to first helped to lead personality psychologists out
appearances, behavior can simultaneously of their crisis of confidence but also induced
be strongly affected by situational factors many social psychologists to consider per
and also demonstrate strong evidence of in sonality more seriously in their own work. By
dividual differences. the m id-1980s, the percentage of articles in
An example may help to make this point. JPSP that involved personality had regained
Imagine that we administer a measure their precrisis levels. In 2 0 0 2 , the last year
of dispositional fearfulness— the degree for which data are available, just over half of
to which people tend to feel anxious and the articles in JPSP included some measure
afraid— to a sample of 100 participants. of personality (Swann & Selye, 2005).
We then randomly assign them to either an
experimental condition in which they are
threatened with painful electric shocks or Uses o f Personality Variables
to a control condition in which no threat is in Behavioral Research
present and ask them to rate how anxious
they feel. An analysis of the between-group M ost social psychologists now acknowledge
differences in anxiety would undoubtedly that dispositional factors predict variation
show a very strong effect of experimental in people’s thoughts, feelings, and behaviors
condition indicating that participants who that cannot be explained by situational fac
were threatened with shocks reported more tors and that a consideration of personality
anxiety on average than those who were can thus contribute to our understanding of
not. At the same time, however, correlat social-psychological phenomena. Research
ing participants’ pretest fearfulness scores ers differ in the degree to which they incor
with their anxiety ratings (whether corre porate personality variables into their own
lated within each condition or for the entire work, but, overall, social psychology is more
sample) would undoubtedly reveal a large amenable to the study of personality than
correlation between dispositional fearful ever before (Swann & Selye, 2005). Specifi
ness and how much anxiety participants cally, individual-difference variables can be
reported while they waited to be shocked. used to address five basic types of questions
In such a case, a strong situational effect is about social thought, emotion, and behav
revealed via between-group differences in ior.
state anxiety, and a strong personality ef
fect is revealed via correlations between a
M ain Effects
measure of personality and state anxiety.
Funder (2006) demonstrated this effect The simplest and most straightforward
empirically. Using data from Funder and questions about the relationship between
Colvin (1991), he showed that, across 62 be personality and social behavior involve
haviors that were measured across two situ “main effect” relationships between a par
ations, 20 behaviors differed significantly ticular disposition and some socially rel
between the two situations at the same time evant thought, emotion, or behavior. In its
that 37 behaviors showed significant within- simplest form, these kinds of studies simply
person stability. Furthermore, the correla correlate trait measures of personality with
tion between the magnitude of between- measures of particular behaviors, cogni
situation differences and cross-situational tions, emotions, or physiological reactions.
stability in behavior was - .0 1 , showing that For example, in a study designed to under
the relationship between situational influ stand aspects of political behavior, Bizer
1. S ituation s, D isp osition s, and the Study o f S ocial B eh av io r 7
and colleagues (2004) found that individu Testing T heories about Situations
al differences in the need to evaluate— the
The fact that a particular experimental
tendency to chronically evaluate aspects of
manipulation influences some behavior of
one’s life and environment— predicted the
interest often does not provide a great deal
degree to which people relied on party iden
of insight into the causes of the obtained
tification to form attitudes toward political
effect. Even when the experiment was de
candidates, the likelihood of voting in na
signed to test a particular theory, obtaining
tional and state elections, and the degree to
results consistent with hypotheses does not
which participants had emotional reactions
unequivocally support the theory’s explana
to political candidates. M ain-effect findings
tion, because one cannot prove the anteced
such as these show how features of people’s
ent of a logical argument (the theory-based
personalities relate to social-psychological
predictions) by affirming the consequent (ob
phenomena.
Another strain of main-effect research taining results that support the hypothesis).
involves correlations between two or more Results may appear to support a hypothesis
personality characteristics that are relevant for reasons other than those that the theory
to social behavior. For example, in a study specifies, and science is filled with examples
that focused on the question of whether in of empirical findings that appeared to sup
dividual differences in religiosity are distinct port a theory that was eventually shown to
from individual differences in spirituality, be false (Wallach & Wallach, 1998).
Saucier and Skrzypinska (2006) found that One strategy for exploring the possible
individual differences in subjective spirituali mechanisms underlying a particular experi
ty were positively correlated with private self- mental effect involves determining whether
consciousness and absorption, but traditional a particular personality variable moderates
religiousness was not. In contrast, traditional the effects of an experimental manipulation
religiousness correlated with right-wing au in the manner predicted by theory. In such
thoritarianism, but subjective spirituality did instances, the researcher is not primarily
not. In studies such as these, relationships interested in the personality variable per se
among various individual-difference mea but uses it as a methodological tool to test
sures that are relevant to social-psychological a hypothesis regarding a situational effect.
phenomena are examined. Imagine, for example, that we are testing the
Much of the research that has been con hypothesis that a particular situational ef
ducted on gender differences also falls in this fect on behavior is caused by the fact that the
category. Although not a “personality” at situation increases people’s concerns about
tribute per se, gender is certainly a potent being rejected by other people. If, prior to
individual-difference variable that relates to manipulating the independent variable(s) of
a wide array of socially relevant thoughts, interest, we obtain participants’ scores on a
emotions, and behaviors (see Wood & dispositional measure of rejection sensitiv
Eagly, Chapter 8, this volume). The wealth ity (Downey & Feldman, 1996), we can ex
of data regarding how women and men dif amine whether people who score low versus
fer is reflected in the growing number of high in rejection sensitivity respond differ
meta-analyses that have examined gender ently to the experimental manipulation, as
differences in aggression, leadership, com they would be expected to do if the effect
munication, jealousy, conversational inter somehow involves concerns with rejection.
ruptions, and other interpersonal behaviors
(e.g., Eagly & Johnson, 1990; Harris, 2 0 0 3 ; Testing T heories about Dispositions
Hyde, 1984).
O f course, these main-effect analyses of the A parallel strategy may be used to test hy
relationship between personality and socially potheses about the nature of a particular
relevant outcomes can become much more personality disposition. Historically, per
complex as researchers investigate multiple sonality researchers have been interested
predictors of various outcomes, examine pos primarily in main-effect hypotheses about
sible interactions among individual-difference dispositions, which they have tested by cor
variables in predicting behavior, and test me- relating scores on a personality scale with
diational and path-analytic models. other scales or by comparing how people
8 I. IN T R O D U C T IO N
who score low versus high on the scale be ing similarities and differences in how low
have in some situation. versus high levels of the state and low versus
However, our understanding of the cog high levels of the trait manifest in thoughts,
nitive, emotional, or behavioral features of emotions, behaviors, or physiological reac
a personality variable can be enhanced by tions. For example, we can learn a great deal
studying how people who score differently about anxiety both by assessing people’s re
on the personality variable respond across actions to experimentally manipulated low
various experimentally created situations. and high threat and by comparing the reac
For example, to examine how optimists ver tions of people who score low versus high on
sus pessimists process negative emotional a measure of trait anxiety. Similarly, we can
stimuli differently, Isaacowitz (2005) had study the relationship between the motive to
participants complete a self-report measure obtain social approval and some behavior,
of optimism. Participants then viewed three either by experimentally varying factors that
types of visual stimuli while their eye move influence the desire for approval or by mea
ments were tracked. Optimists showed selec suring individual differences in the need for
tive inattention to the most negative stimuli, approval.
and this relationship remained significant When the results of experimental studies
after controlling for the effects of neuroti- of states converge with those of correlational
cism, anxiety, and other variables. In studies studies of traits, we have greater confidence
such as this, experimentally manipulating that we understand the processes involved.
features of the participants’ environment (in And, when they do not converge (and they
this case the nature of visual stimuli) pro often do not), interesting questions arise re
vides insight into the nature of the personal garding why the state and trait operational
ity variable of interest. definitions of seemingly analogous con
The strategy of combining manipulated structs are not equivalent.
independent variables and measured person
ality variables in a single study may result in
State-by-Trait Interactions
precisely the same research design whether
one is primarily interested in understanding M ost social psychologists realize that, be
the situational or the dispositional effects. In cause people differ in their reactions to so
both cases, one is interested in the interac cial stimuli, almost every general statement
tion of the experimental manipulation and about the effects of a particular situational
the measured trait, and whether we say that factor is at best incomplete and at worst
the personality variable moderated the ef misleading or wrong. Likewise, personal
fects of the independent variable or the inde ity psychologists seem to understand that,
pendent variable moderated the effects of the although general predictions can be made
personality variable depends on our focus. on the basis of a person’s position on a par
ticular trait dimension, how people actually
behave at any moment is typically influenced
State and Trait Convergence
to some degree by the situation in which
Certain situational variables create differ they are found. Thus explaining virtually
ences in people’s psychological states that any thought, emotion, or behavior at a given
are conceptually analogous to the individual moment in time requires attention to both
differences that we see among people who situational and dispositional factors.
possess different levels of a personality trait. Furthermore, situational and disposition
For example, just as mildly versus severely al factors not only exert separate, additive
threatening situations elicit different levels influences on people’s responses but also can
of state anxiety, trait-like differences exist potentially interact in a statistical sense in
in the degree to which people are generally that the effects of a particular situation may
anxious. Likewise, certain situations in vary across levels of a trait or the effects of
crease people’s motivation to obtain social a trait may vary across situations. In fact,
approval, and certain people are disposition- a particular trait may relate to behavior in
ally more motivated to obtain approval than only some situations, and a particular situa
are others. tion may influence the reactions of only peo
When conceptually analogous states and ple with a certain personality characteristic
traits exist, much can be learned by exam in (Bern & Funder, 1978). Thus many studies
1. S itu ation s, D isp osition s, and the Study o f S ocial B eh av io r 9
in social and personality psychology test for Magnusson, 1976; Endler & Parker, 1992).
person-situation (or trait-by-state) interac The two sets of influences not only combine
tions. to influence or predict behavioral outcomes
Behavioral researchers tend to love statis as just described, but they also influence one
tical interactions, which, for some reason, another in a dynamic, reciprocal fashion. In
tend to connote the presence of a more so dynamic interactionism (Endler, 1983), the
phisticated and elegant psychological pro distinction between antecedents and con
cess than the mere presence of simultaneous sequences (and independent and dependent
main effects. Yet, although interactions be variables) may not be appropriate because
tween situations and dispositions are often situations and traits mutually influence one
interesting and informative, they are also another in a variety of ways. For example,
notoriously difficult to obtain, and, when a person’s traits can change the nature of a
they occur, they tend to be quite small rela situation, such as when a highly agreeable
tive to main effects (Chaplin, 1997; Kep- person creates a friendly and cooperative so
pel, 1982). Several factors contribute to the cial environment or an aggressive child in
weakness of person-situation interactions. stigates widespread hostility on a previously
First, the reliability of an interaction term is peaceful playground. Furthermore, people
almost always lower than the reliability of its with different personality predispositions
constituents (Bohrnstedt & M arwell, 1977). sometimes choose different kinds of social
Because the strength of a statistical effect is settings (Snyder & Ickes, 1985). Unlike in
attenuated by measurement error, the lower experimental settings in which people are
reliability of interaction terms decreases the thrust into situations that they did not pick,
likelihood that interactions will be detected in everyday life people have a certain degree
even if they are present (McClelland & Judd, of flexibility and freedom to gravitate to
1993). Furthermore, statistical models that ward situations that are consistent with their
include interaction terms have lower degrees personalities. Once people are in those self
of freedom for the error term than models selected situations, one finds it meaningless
that contain main effects only, so that statis to ask whether their behavior is a function
tical significance is less likely. of the situations or of their personalities, be
We would add to these documented con cause personality has determined the situa
siderations the possibility that we live in a tion. Likewise, personality traits can change
predominantly main-effect world. Although when people are in certain situations. For in
people undoubtedly respond differently from stance, the classic Bennington study showed
each other in any particular situation, those that students became less conservative dur
differences are often scaled similarly across ing their college experience and remained
situations. Thus, rather than finding interac less conservative for years afterward (New
tions in which the effects of a situation are comb, Koenig, Flacks, & Warwick, 1967).
different for some people than for others, we Fortunately, the development of struc
often find two main effects that reveal a situ tural equation modeling and related statisti
ational influence that increases or decreases cal modeling strategies provides for the first
everyone’s reactions while the variability time a way to approach modeling these com
among people remains constant. In any case, plex, reciprocal influences. As described by
for these and other reasons, statistical inter Hoyle and Leary (Chapter 2 , this volume),
actions between situations and personality if data are gathered strategically (i.e., re
are relatively rare relative to main effects, peatedly, with appropriate spacing, across
and those that do occur generally account time and situations), it is possible to model
for relatively little variance. the strong, dynamic version of interaction
Personality and situational influences can ism that its proponents advocate (Endler &
combine, influence, and interact with one Parker, 1992).
another in much more complicated ways
than through simple statistical interactions
Nonlinearity
between experimental manipulations and
measured personality variables. Proponents A relatively uncharted direction for research
of “interactionism” point to the fact that situ on the interplay of personality and social
ational and personality influences are mutu behavior is the modeling of nonlinear rela
ally interdependent (Endler, 1983; Endler & tions. Following up on our suggestion that,
10 I. IN T R O D U C T IO N
for the most part, people inhabit a main- ant, while another branch declared that only
effect world, we suspect that the relation the forces that acted on matter were worth
ships between variables in that world are, by studying (and that those forces could be stud
and large, linear. However, just as interac ied without reference to the characteristics of
tion effects add nuance (and sometimes sig matter itself)? Can we imagine one group of
nificant variance accounted for) to models of chemists being interested only in chemical
personality and social behavior, the addition structure and another group being interested
of nonlinear terms to statistical models may only in interactions among chemicals with
add richness and subtlety to our understand out considering the structure of the constitu
ing of the relationship between dispositions ents? Could meteorologists function if some
and behavior. studied only the properties of relatively static
Nonlinear relations can range from rela weather systems and others studied only the
tively straightforward curvilinear effects forces that act on them? Fortunately, most
evaluated using power polynomials in multi behavioral scientists now agree that the rift
ple regression and trend analysis in analysis between social psychologists and personality
of variance to complex dynamical systems psychologists has been misguided and detri
that attempt to model the “chaos” and “ca mental to a full understanding of socially rel
tastrophe” evident in human social behav evant thought, emotion, and behavior.
ior (e.g., Tesser & Achee, 1994; Vallacher, This rapprochement does not mean that
Nowak, &C Kaufman, 1994). An example we should all start studying precisely the
of work in which potential curvilinear rela same things, of course. We need specialists
tions are explored is Jorm and Christensen’s in personality structure and process, as well
(2004) study of the relations between reli as those who specialize in studying the ef
giosity and Eysenck’s three-factor model of fects of the “actual, imagined, or implied
personality. In addition to a modest linear presence of others” (Allport, 1968, p. 3). But,
relation with one factor, they found quadrat in trying to understand the phenomena that
ic relations with all three factors, indicating constitute the science of human psychology,
similarity in the personalities of individuals devoting attention to both situational and
at the highest and lowest levels of religiosity. dispositional factors is the optimal strategy.
Tesser and Achee (1994) identified a number
of instances of catastrophe in the prediction
of social behaviors. In such cases, a seem R eferen ces
ingly linear relation between two variables
Allport, G. W. (1937). P erso n a lity : A p s y c h o lo g ic a l
quickly changes direction at a particular
in te rp r eta tio n . New York: Holt, Rinehart & W in
point before returning to a linear form like ston.
that before the “catastrophe.” Such dynami Allport, G. W. (1968). The historical background
cal systems analyses also offer a compelling of modern social psychology. In G. Lindzey & E.
means of connecting seemingly disparate Aronson (Eds.), H a n d b o o k o f s o c ia l p sy c h o lo g y
(Vol. 1, 2nd ed., pp. 1 - 8 0 ). Reading, M A : Addison-
levels of analysis, such as neurobiology and Wesley.
personality (Mandell & Selz, 1995). Such Bern, D., & Funder, D. (1978). Predicting more of the
findings contribute to conceptual models that people more of the time: Assessing the personality
offer more precise and nuanced accounts of of situations. P sy c h o lo g ic a l R e v iew , 8 5, 4 8 5 - 5 0 1 .
individual differences in social behavior. Bizer, G. Y., Krosnick, J. A., Holbrook, A. L., Wheeler,
S. C., Rucker, D. D., & Petty, R. E. (20 04 ). The
impact of personality on cognitive, behavioral, and
affective political processes: The effects of need to
Conclusions evaluate. J o u r n a l o f P erson ality , 7 2, 9 9 5 - 1 0 2 7 .
Bohrnstedt, G. W., & Marwell, G. (1977). The reli
ability of products of two random variables. In K.
We find it difficult to imagine scientists in
F. Schussler (Ed.), S o c io lo g ic a l m e t h o d o lo g y : 1 9 7 8
any other discipline falling into a contro (pp. 2 5 4 - 2 7 3 ) . San Francisco: Jossey-Bass.
versy that would be equivalent to the tradi Caspi, A., & Moffitt, T. E. (1993). When do individual
tional schism between social and personality differences matter?: A paradoxical theory of person
psychologists. Would one branch of physics ality coherence. P sy c h o lo g ic a l In q u iry , 4, 2 4 7 - 2 7 1 .
Cervone, D., Caldwell, T. L., & Orom, H. (in press).
declare that the most important topics in the Beyond person and situation effects: Intraindividu
field involved the nature of matter but that al personality architecture and its implications for
forces such as gravitation were unimport the study of personality and social behavior. In F.
1. S itu ation s, D isp osition s, and the Study o f S ocial B eh av io r 11
Rhodewalt (Ed.), F ro n tiers o f s o c ia l p sy c h o lo g y : Ickes, W. (1982). A basic paradigm for the study of
P erso n a lity a n d s o c ia l b eh a v io r . New York: Psy personality, roles, and social behavior. In W. Ickes
chology Press. & E. Knowles (Elds.), P erson ality , ro les, a n d s o
Chaplin, W. F. (19 97). Personality, interactive rela cia l b e h a v io r (pp. 3 0 5 - 3 3 1 ) . New York: Springer-
tions, and applied psychology. In R. Hogan, J. J o h n Verlag.
son, & Briggs, S. (Eds.), H a n d b o o k o f p er so n a lity Isaacowitz, D. M. (2005). The gaze of the optimist.
p s y c h o lo g y (pp. 8 7 3 - 8 9 0 ) . San Diego, CA: Aca P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , 31,
demic Press. 407-415.
Cronbach, L. J . (1957). The two disciplines of scientific Jorrn, A. F., & Christensen, H. (200 4). Religiosity
psychology. A m er ic a n P sy ch o lo g is t, 12, 6 7 1 - 6 8 4 . and personality: Elvidence for nonlinear associa
Downey, G., &C Feldman, S. I. (1996). Implications tions. P erso n a lity a n d In d iv id u a l D iffe r e n c e s , 3 6 ,
of rejection sensitivity for intimate relationships. 1 4 3 3 -1 4 4 1 .
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 70, Kenrick, D. T., & Dantchik, A. (1983). Interaction
1327-1343. ism, idiographics, and the social psychological in
Eagly, A. H., & Joh nson, B. T. (1990). Gender and vasion of personality. J o u r n a l o f P erson ality , 51,
leadership style: A meta-analysis. P sy c h o lo g ic a l 286-307.
B u lletin , 1 0 8 , 2 3 3 - 2 5 6 . Keppel, G. (1982). D esign a n d a n aly sis: A research er's
Endler, N. S. (1983). Interactionism: A personal h a n d b o o k (2nd ed.). Englewood Cliffs, NJ: Pren-
ity model, but not yet a theory. In M. M. Page tice-Hall.
(Ed.), N e b r a s k a S y m p osiu m o n M otiv atio n 1 982: Lewin, K. (1936). P rin ciples o f t o p o lo g ic a l p s y c h o l
P ers o n a lity — C u rren t th e o r y a n d resea rch (pp. 1 5 5 - ogy. New York: McGraw-Hill.
2 00 ). Lincoln: University of Nebraska Press. Magnusson, D., & Endler, N. S. (1977). Interactional
Endler, N. S., & Magnusson, D. (1976). Toward an in psychology: Present status and future propsects. In
teractional psychology of personality. P sy ch o lo g ica l D. Magnusson & N. S. Endler (lids.), P erson ality at
B u lletin , 8 3 , 9 5 6 - 9 7 4 . th e c r o s s r o a d s : C u rren t issu es in trait p s y c h o lo g y
Endler, N. S., & Parker, J . D. A. (1992). Interactionism (pp. 3 - 3 5 ) . Hillsdale, NJ: Erlbaum.
revisited: Reflections on the continuing crisis in the Mandell, A. J ., & Selz, K. A. (1995). Nonlinear dy
personality area. E u r o p ea n J o u r n a l o f P erson ality , namical patterns as personality theory for neurobi
6, 1 7 7 - 1 9 8 . ology and psychiatry. P sychiatry, 5 8 , 3 7 1 - 3 9 0 .
Epstein, S. (1979). The stability of behavior: I. On McClelland, G. H., & Judd, C. M. (1993). Statistical
predicting most of the people much of the time. difficulties of detecting interactions and moderator
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 37, effects. P s y c h o lo g ic a l B ulletin , 114, 3 7 6 - 3 8 9 .
1097-1126. Milgram, S. (1963). Behavioral study of obedience.
Epstein, S. (1983). Aggregation and beyond: Some lo u r n a l o f A b n o r m a l a n d S o c ia l P sy ch o lo g y , 67,
basic issues on the prediction of behavior. J o u r n a l 371-378.
o f P erson ality , 51, 3 6 0 - 3 9 2 . Mischel, W. (1968). P erson ality a n d a s se ss m e n t. New
Epstein, S., & O ’Brien, E. J. (1985). The person - York: Wiley.
situation debate in historical and current perspec Newcomb, T. M., Koenig, K. E., Flacks, R., & W ar
tive. P sy c h o lo g ic a l B u lletin , 9 8, 5 1 3 - 5 3 7 . wick, D. P. (1967). P ersisten ce a n d ch an g e. New
Eleeson, W. (2001). Towards a structure- and process- York: Wiley.
integrated view of personality: Traits as density Packer, D. J. (2008). Identifying systematic disobedi
distributions of states. J o u r n a l o f P erson ality a n d ence in M ilgra m ’s obedience experiments. P ers p ec
S o c ia l P sy ch olog y , 8 0 , 1 0 1 1 -10 2 7. tives on P sy c h o lo g ic a l S cien ce, 3, 3 0 1 - 3 0 8 .
Fleeson, W. (20 0 4). Moving personality beyond the Ross, I.., & Nisbett, R. E^. (1991). T h e p er so n a n d the
person-situation debate: The challenge and oppor situ ation . New York: McGraw-Hill.
tunity of within-person variability. C u rren t D ire c Saucier, G., & Skrzypiriska, K. (200 6). Spiritual but
tio n s in P sy c h o lo g ic a l S cien c e , 13, 8 3 - 8 7 . not religious?: Evidence for two independent dispo
Funder, D. C. (200 6 ). Towards a resolution of the per sitions. J o u r n a l o f P erson ality , 74, 1 2 5 7 - 1 2 9 2 .
sonality triad: Persons, situations, and behaviors. Snyder, M ., & Ickes, W. (1985). Personality and so
J o u r n a l o f R e se a rc h in P erson ality , 4 0 , 2 1 - 3 4 . cial behavior. In G. I.indzey & E. Aronson (Eds.),
Funder, D. C., & Colvin, C. R. (1991). Explorations in H a n d b o o k o f so c ia l p sy c h o lo g y (3rd ed., pp. 8 8 3 —
behavioral consistency: Properties of persons, situ 947). New York: Random House.
ations and behaviors. J o u r n a l o f P erson ality a n d Swann, W. B., Jr., & Seyle, C. (2 005). Personality
S o c ia l P sy ch olog y , 5 2 , 7 7 3 - 7 9 4 . psychology’s comeback and its emerging symbiosis
Funder, D. C., & Ozer, D. J. (1983). Behavior as a with social psychology. P erson ality a n d S o c ia l P sy
function of the situation. J o u r n a l o f P erson ality a n d c h o lo g y B u lletin , 3 1, 155 -1 65 .
S o c ia l P sy ch olog y , 4 4 , 1 0 7 - 1 1 2 . Tesser, A., & Achee, J . (1994). Aggression, love, co n
Harris, C. R. (2003). A review of sex differences in formity, and other social psychological catastro
sexual jealousy, including self-report data, psy- phes. In R. R. Vallacher & A. Nowak (Eds.), D y
chophysiological responses, interpersonal violence, n a m ic a l sy stem s in s o c ia l p s y c h o lo g y (pp. 9 6 - 1 0 9 ) .
and morbid jealousy. P erson ality a n d S o c ia l P sy San Diego, C'A: Academic Press.
c h o lo g y R e v iew , 7, 1 0 2 - 1 2 8 . Vallacher, R., No wak, A., Sc Kaufman, J. (1994). In
Hyde, J . S. (1984). How large are gender differences in trinsic dynamics of social judgment. J o u r n a l o f P er
aggression?: A developmental meta-analysis. D ev e l so n a lity a n d S o c ia l P sy ch olog y , 67, 2 0 - 3 4 .
o p m e n t a l P sy ch olog y , 2 0 , 7 2 2 - 7 3 6 . Wallach, M. A., & Wallach, L. (1998). When experi
Ichheisser, G. (1943). Misinterpretations of personal ments serve little purpose: Misguided research in
ity in everyday life and the psychologist’s frame of mainstream psychology. T h eo r y a n d P sy ch o lo g y , 8,
reference. C h a ra c ter a n d P erson ality , 12, 1 4 5 - 1 6 0 . 183-194.
CH APTER 2
R ick H . H o y l e
M a rk R . Lea r y
12
2. M ethods for the Study o f Individual D ifferen ces 13
questions, this momentary or situated reac across a range of studies, relationships be
tion, not the characteristic or dispositional tween attitudes and behaviors are strongest
form, is of interest. In such cases, a state when the context and specificity of the mea
measure is the appropriate choice. Alter sures are compatible (Kim & Hunter, 1993).
natively, if the research question concerns For instance, predictions of the use of birth
participants’ characteristic standing on a control pills improve markedly when the
psychological characteristic, then a trait specificity of the predictor and outcome are
measure is appropriate. comparable. Predictions of the use of birth
For some responses, separate trait and control pills from people’s self-reported atti
state measures have been developed. For ex tudes ranges from r = .08 for attitude toward
ample, the State-Trait Anxiety Inventory al birth control in general to r = .32 for attitude
lows the assessment of both the “transitory toward birth control pills to r = .52 for at
emotional state” of anxiety and “relatively titudes toward use of birth control pills to r
stable individual differences” in the degree = .57 for attitude toward using birth control
to which people tend to experience anxiety pills in the next 2 years (Davidson & Jaccard,
(Spielberger, 1983). Another widely used in 1979). Similarly, one solution to the problem
strument with which state and trait measure of low observed correlations between per
ment is possible is the Positive and Negative sonality and behavior (Mischel, 1968) is to
Affect Schedule, on which respondents can aggregate behaviors to produce a variable at
be asked to rate on a single set of adjectives a level of generality that is similar to stan
their feelings either “during the past week” dard measures of personality traits (Epstein,
(i.e., trait) or “right now, at this moment” 1980). For example, extraversion is more ap
(i.e., state; Watson, Clark, & Tellegen, propriately used as a predictor of trends in
1988). A similar tack has been taken with behavior in social settings across time and
the Self-Esteem Scale (Rosenberg, 1965). situations than of any single instance of ex
Although this scale is typically regarded as traverted behavior. In studies of individual
a measure of trait self-esteem, researchers differences in social behavior, the principle
have used it to assess state self-esteem by of compatibility dictates that, to the extent
asking respondents to indicate the extent to possible, the individual-difference variable
which the statements about their self-worth should be measured at a level of specific
apply “at the moment” that they complete ity that corresponds to the specificity of the
the scale (e.g., Kernis, Grannemann, & Bar measure of the variables to be predicted.
clay, 1989). Often, researchers find it benefi Unlike the development of state-level coun
cial to assess a construct at both a trait and terparts to trait measures, the development
a state level so that the relationship between of specific measures from general measures
general tendencies and specific behaviors (and vice versa) is not straightforward. A
can be examined. fundamental concern involves determining
the appropriate dimensions that should be
used to parse the general construct to pro
General versus Specific M easurem ent duce more specific variants. A rich example
is global versus domain-specific self-esteem.
Individual differences can be measured at Several researchers have attempted to parse
varying levels of specificity, from very gener global self-esteem into more specific “types”
al (e.g., trait anxiety) to highly specific (e.g., by identifying a relatively small number of
trait anxiety regarding interactions with dimensions on which people characteristi
people of the other sex). Typically, the spec cally evaluate themselves, such as physi
ificity of the measure used should fall at a cal appearance, social skills, and academic
point along this continuum that corresponds competence (e.g., Fleming & Courtney,
to the level of specificity at which other 1984; Harter, 1988). However, these dimen
variables in the study are measured. The sions can be further decomposed into even
practice of matching the specificity of mea more specific areas of self-evaluation, such
surement of variables follows from the prin as particular features of one’s appearance
ciple of compatibility originally proposed (e.g., Franzoi & Shields, 1984) or specific
in studying relations between attitudes and academic subjects (Marsh & O ’Neill, 1984).
behavior (Fishbein & Ajzen, 1975). Indeed, Following from the principle of compatibil
14 I. IN T R O D U C T IO N
ity, these more specific variants are most concerns about the confidentiality of sensi
fruitfully used when other variables in the tive information that is transmitted over a
study are equally specific. network.
Moving in the other direction— from spe Both paper-and-pencil and computer-
cific to general— is more straightforward. administered measures assume that respon
Measures of particular behaviors or of dents can read and understand what they are
domain-specific self-reports can be aggre asked to do. When these assumptions are
gated (i.e., typically summed or averaged) not reasonable or when self-administration
to produce measures of general tendencies is not viable for other reasons, administra
that are comparable in level of specific tion of measures by a researcher is preferred.
ity with general measures of personality. Although the relative cost of this modality is
For instance, the seven specific domains in high, it generally yields a high response rate
which contingency of self-worth is measured and high-quality data. Drawbacks to this
by the Contingencies of Self-Worth Scale modality include the potential for interview
(Crocker, Luhtanen, Cooper, & Bouvrette, er bias, concerns about anonymity, the time
2003) can be aggregated into two more gen required to interview respondents individu
eral domains— internal versus external con ally, and cost.
tingencies— and these two domains can be Because all measurement modalities have
further aggregated into a measure of overall shortcomings, it is worrisome that the over
contingency of self-worth. This general mea whelming majority of published research
sure would be compatible with more general on individual differences in social behavior
measures of other constructs, such as global is based on self-administered paper-and-
self-esteem and overall adjustment. pencil questionnaires. To help address con
cerns with the ubiquity of paper-and-pencil
measurement, measures should be validated
M easurem ent M odality across modalities of administration. When
measures perform comparably no matter
When planning studies to examine links be how they are completed by respondents, re
tween individual differences and social be searchers have added faith in the validity of
havior, researchers should consider the op the measures and are equipped to disentan
timal way to administer the measures. The gle modality-specific biases from genuine re
number of options for assessing traits con lations between variables by using multiple
tinues to increase as technology expands and modes of collecting data across studies in a
access to this technology improves. The tra research program (Campbell, 1969). They
ditional paper-and-pencil approach is giving also are in a position to choose the measure
way to computer administration on either ment modality that best fits the study popu
stand-alone or locally networked comput lation (e.g., college students, the poor, the
ers or via server over the World Wide Web. mentally ill), the setting (e.g., research lab,
Significant advantages of computer-based participants’ homes, school), and research
administration are low cost (e.g., no photo question (e.g., simple associations, change
copying, postage, or data entry), the ability over time, variability across situations) with
to shuffle the order of items or scales easily, out concern for validity.
and, if needed, the ability to score measures
on the fly so that subsequent aspects of the
study can be tailored to each participant’s Multiple Measures
score on a particular measure. Computer
administration also allows seamless integra No single measure fully captures the con
tion of questionnaires with audio and video struct that it operationally defines— a fre
stimuli. Flowever, these benefits are offset by quent but misguided assumption known as
the need for participants to have access to a definitional operationism (Campbell, 1969).
computer and the Internet, the potential for For example, no measure of extraversion
lost data due to computer malfunction, the truly, accurately, and completely assesses
inability of researchers to control the condi extraversion. Moreover, all measures are
tions under which people complete the mea influenced to some degree by extraneous
sures over the Web, and, in some instances, factors such as social desirability or biases
2. M ethods fo r the Study o f Individual D ifferences 15
in how particular respondents use the re that the effect is not due to a measurement
sponse scales. For this reason, it is unwise or method artifact.
for research findings to be based on only one Although using different measures and
measure of key constructs. Ideally, multiple measurement strategies across studies ad
measures should be used across (or within) dresses concerns across a program of research
studies in a research literature or program, or a research literature, it is advantageous
measures that differ not only in content but to include multiple measures and strategies
also in modality of administration, reporters in individual studies when possible. Refer
(e.g., self, peer, parent, teacher), and means ring back to the anxiety example, if these
of responding. Findings that obtain across three measures— a self-report, physiological
measures (particularly if the measures differ marker, and judges’ rating— were gathered
in response format or mode of administra in a single study, their commonality could
tion) are presumably more robust and repli be modeled as a latent variable (e.g., using
cable than those based on a single measure. structural equation modeling), which then is
Moreover, when effect sizes are combined correlated with other variables. In this ap
across such studies, the average should con plication, one assumes that, although each
verge on the true effect size, that is, the effect measure is subject to various extraneous
size uncontaminated by particular method influences that undermine its reliability, the
effects. three measures share in common the influ
The use of multiple measures can take var ence of anxiety. By separating this common
ious forms that address different concerns ality from unique and random error, one is
about the reliance on a single measure. In left with a relatively pure representation of
one form, which addresses concerns about the construct. Assuming significantly differ
the degree to which a single measure covers ent content and measurement strategies, any
the full content of a construct, two or more effects obtained using this approach cannot
measures of the same construct are used. For be attributed to the idiosyncrasies of indi
instance, three different self-report measures vidual measures or measurement strategies
of self-esteem ensure greater coverage of the (cf. DeShon, 1998).
construct and provide a means of removing
sources of bias that differ across measures.
Despite these important payoffs, this strat M ethodological Strategies
egy does not deal with any biases that are
related to self-reports of self-esteem in gen The goal of sound measurement is an opera
eral. Alternatively, a single measure might tional definition of the individual-difference
be used but completed by two or more re variable that allows an accurate estimate of
porters (i.e., collateral reports). For instance, the magnitude and form of the relationship
severity of depressive symptoms might be between the variable and relevant social
assessed by asking a parent and a teacher thoughts, emotions, or behaviors. Yet even
to complete the instrument for each partici the most careful and complete measure
pant, as well as the participant him- or her ment strategy does not ensure that this goal
self. This strategy does not address concerns is met. Measurement is undertaken in the
about the content coverage of the measure, context of a methodological strategy and, as
but it addresses concerns about the valid with measurement, no methodological strat
ity of self-reports. A third strategy is to use egy is perfect. Consequently, the strongest
two or more measures that differ in content evidence is obtained when different meth
coverage, modality, and/or reporter. For ex odological approaches are used to estimate
ample, anxiety might be measured through or model the relationships between indi
paper-and-pencil or computer-administered vidual differences and social-psychological
self-reports, assessment of relevant physi outcomes. In this section, we outline a set of
ological markers, and trained judges’ coding principles for choosing and evaluating find
of participants’ videotaped behavior in the ings from research methods that are com
presence of an anxiety-inducing stimulus. If monly used to study individual differences
anxiety as measured in these various ways in social behavior. We then describe four
correlates similarly to measures of social categories of methodological strategies for
behavior, then the researcher can conclude such research.
16 I. IN T R O D U C T IO N
quence variables in statistical models and age in the design. For instance, Schultz and
about inflated estimates of covariance are Moore (1988) compared loneliness scores
significant, and they limit the potential con and correlates for high school students, col
tribution of cross-sectional studies to a gen lege undergraduates, and retirees with the
eral understanding of the role of individual goal of shedding light on age-related chang
differences in social behavior. es in loneliness. These variations on the one-
Nonetheless, there is a place for cross- shot cross-sectional study offer some of the
sectional studies. The first question to be benefits of strategies that explicitly incorpo
settled when undertaking any new line of rate time, but the findings they yield must be
research is simply whether and how the vari interpreted with caution. For instance, com
ables of interest are associated, and more parisons of different age groups are subject
complex studies typically await a cross- to cohort effects, in which putative age ef
sectional assessment of this question. The fects are actually due to history and context
large sample sizes typical of cross-sectional effects. Similarly, in cross-sectional time-
research, coupled with the fact that most series studies, events may transpire during
variables are measured on quasi-continuous the course of a study that alter the state of
scales, allow efficient evaluations of the re the population (e.g., September 11, 2001),
lations among variables, including those at obfuscating effects of the event of interest.
tributable to moderation by other variables. Thus, although both cross-sectional time-
Cross-sectional studies often include a large series and age-based cross-sectional strate
number of variables, allowing the use of sta gies offer benefits beyond those of one-shot
tistical approaches (e.g., multiple regression cross-sectional studies, isolating the influ
analysis) that isolate the personality variable ence of specific variables remains a concern.
of interest from other variables with which
it is correlated. Thus, early in a research
E xperim enta l Strategies
program, when the strength and form of
relations involving an individual difference Although also subject to limitations, ex
are not clear, cross-sectional studies are use perimental strategies overcome many of the
ful— even ideal. shortcomings of cross-sectional strategies. A
Under certain circumstances and for cer major benefit of well-designed experiments
tain research questions, more sophisticated is the isolation of the putative cause from
cross-sectional strategies are available that alternative causal factors, accomplished by
allow firmer inferences. For example, when randomly assigning participants to levels of
one’s hypothesis concerns the behavior of a the independent variables and manipulating
population rather than the behavior of in one or more independent variables. Experi
dividuals within the population, one can mental methods also address concerns about
model changes in behavior associated with sequence in directional relations between
situational “interruptions” (either due to variables. Because standing on the indepen
naturally occurring events or planned inter dent variables is attributable to a random
ventions) by repeatedly randomly sampling process that occurs at a known moment in
(without replacement) from the population time (i.e., the introduction of the indepen
over time. For example, Palmgreen, Dono- dent variable), it cannot be attributed to
hew, Lorch, Hoyle, and Stephenson (2001) other systematic sources. As such, if the in
found that when the population of young dependent variable is statistically associated
viewers in a television market is divided with scores on other variables, the only logi
into subgroups of individuals low and high cal inference is that the independent variable
in sensation seeking, the high-sensation- is antecedent to those variables in a causal
seeking group is both more likely to use illicit sequence. This does not rule out the possi
substances and to be influenced by antidrug bility that, if the roles of the variables were
media campaigns than the low-sensation- reversed in another study so that the inde
seeking group. pendent variable was measured and one of
A variation on this strategy is to draw the other variables manipulated, we might
(ideally, random) samples of different ages also observe a relation. Such a pattern would
from a population at a single point in time, indicate a bidirectional relation between the
thereby incorporating time as reflected in variables.
18 I. IN T R O D U C T IO N
Research using experimental methods personality scores is preserved (see Aiken &
often is undertaken in “laboratory” condi West, 1991).
tions, a typically artificial setting in which The strength of experimental designs that
the experimenter controls the situational include personality characteristics can be
variables that might influence the participant increased by measuring and covarying one
during the experiment. Experiments need or more other personality variables that are
not occur in such settings, however, and potentially confounded with the personality
the field experiment is a compelling strategy variable of interest. Imagine, for example,
that brings to bear on a research question that one is interested in how people who
the strengths of the experimental method in vary in need for power (the personality vari
a setting in which the behaviors of interest able of interest) respond to a manipulated
might typically be enacted. threat to their authority (the independent
A true experiment involving only individ variable). Because need for power is likely
ual differences is not possible, because re correlated with need for control, one might
search participants self-select to levels of the wish to eliminate the confounding effects of
individual difference, bringing with them control motivation from the findings. To do
unknown other characteristics that cannot so, one could pretest participants on both
be ruled out as alternative explanations. need for power and control motivation and
Thus the usefulness of experimental meth then covary (i.e., partial) control motivation
ods as a means of isolating causal variables from analyses that test effects involving need
is lost. However, researchers often combine for power. Doing so would ensure that any
manipulated independent variables and mea obtained effects were not due to control mo
sured personality variables in the same study tivation and thus were more likely to reflect
in what has been called an expericorr design individual differences in need for power.
(Leary, 200 8 ). (They are called expericorr This inference is tentative, however, because
because they possess features of both a true other unidentified traits that correlate with
experiment and a cross-sectional correlation need for power might be responsible for any
design.) In such designs, participants are pre obtained effects.
tested on the personality variable of interest An increasingly common use of experi
and then randomly assigned to experimental mental strategies for research on individual
conditions. Such designs allow researchers differences involves tests of statistical media
to explore the possibility that the personal tion. In the measurement-of-mediation strat
ity variable moderates reactions to the inde egy (Spencer, Zanna, & Fong, 2005), the pu
pendent variable such that participants who tative cause is manipulated as an independent
score differently on the personality measure variable, and, prior to or concurrent with the
respond differently to the independent vari dependent variables, putative mediators (i.e.,
able. intervening variables) of the relationship be
Traditionally, such moderation effects tween the independent and dependent vari
were tested by splitting participants into ables are measured. Mediational hypotheses
low versus high groups using a median split, assume a strict sequence of causal influences
then entering this dichotomous personal from independent variable to mediator to
ity variable, along with the manipulated dependent variable. When the measurement-
independent variable(s), into an analysis of-mediation approach is implemented using
of variance and testing the personality-by experimental methods, the temporal order
independent-variable interaction. This ana from the independent variable to mediator is
lytical approach is now strongly discour fixed; however, the sequence from mediator
aged because of evidence that converting a to the outcome variable is not. The reason
rich continuous personality variable into a is that the mediator is measured rather than
dichotomy throws away a great deal of in manipulated. Because the mediator serves as
formative variability and greatly reduces both an outcome variable (to the independent
the power of statistical tests (M acCallum, variable) and an antecedent variable (to the
Zhang, Preacher, & Rucker, 2 0 0 2 ; more on dependent variable), this problem cannot be
this later). Tests of personality moderation solved in a single experiment. Instead, a pair
should be conducted using moderated multi of experiments is required. In the first, the
ple regression in which the continuity of the effect of the independent variable on the me
2. M ethods for the Study o f Individual D ifferen ces 19
diator is evaluated. In the second, the effect variance is not a function of other variables
of the mediator on the dependent variable is in the model. Thus, for example, in a study
evaluated. Because the mediator is measured in which individual differences in rejection
in the first instance and manipulated in the sensitivity are measured at Time 1 and as
second, this strategy requires a measure and pects of social interaction are assessed 2
a manipulation that have been shown to be weeks later at Time 2 , the statistical relation
equivalent operational definitions of the me between rejection sensitivity and social be
diator (e.g., private self-awareness; Fejfar & haviors might reflect nothing more than co-
Hoyle, 2000). variance between stable components of the
Because, under the right conditions, ex two variables rather than the fact that rejec
periments can isolate causal variables and tion sensitivity was antecedent to behaviors
establish sequence in the relations between observed in social interactions. Precisely the
variables, one might be tempted to conclude same findings might have been obtained had
that experimental strategies are always pre rejection sensitivity and social behavior been
ferred over other strategies for studying indi measured concurrently.
vidual differences in social behavior. How This concern is addressed rather simply
ever, as with all methodological strategies, by measuring social interaction concurrent
experimental methods have significant limi with rejection sensitivity at both Time 1 and
tations. For example, in contrast with cross- Time 2. In so doing, the stable component
sectional studies, experimental studies are of social interaction can be statistically es
relatively costly in terms of space, personnel, timated and separated from the component
and the amount of time required to complete that is subject to change. By including the
data collection. Furthermore, many topics measure of rejection sensitivity at the second
cannot be studied experimentally because of assessment, it is possible both to estimate the
ethical or logistical constraints. For exam degree of stability in sensitivity to rejection
ple, one could not conduct an experimental and to entertain the possibility that, to some
study of the relationship between parental degree, social interaction causes people to
punitiveness and children’s authoritarianism be more attuned to rejection. Panel studies
because doing so would require randomly (often called cross-lagged panel designs), in
assigning parents to punish their children which all measures are administered at all
with varying degrees of severity. In such points in time, allow for persuasive tests of
cases, the best we can do is to measure hy sequence (Farrell, 1994). Using data from
pothesized causes and outcomes and model such studies, cross-lagged panel analyses can
their relationships. be used to test directly for sequence (Hoyle
& Robinson, 2003). For example, Farrell
(1994) measured anger and alcohol use on
Strategies T h a t Incorporate T im e
three occasions and used cross-lagged panel
If sequencing the temporal order of variables analysis to show that the relation between
is a goal and experimental manipulation is anger and alcohol use can be attributed sole
not feasible, then nonexperimental strate ly to the effect of anger on alcohol use; when
gies that incorporate time are an attractive stability of these constructs is controlled,
alternative. In the prototypical use of this there is no lagged effect of alcohol use on
strategy, a hypothesized cause measured at anger.
one point in time is related to an outcome At least two additional concerns must be
measured at a later point in time. (If media addressed if longitudinal strategies are to
tion is also of interest, then an intermediate be used effectively in studies of individual
assessment of the hypothesized mediator is differences in social behavior. One concern
required.) It is important to recognize that stems from the fact that research participants
the simple form of this strategy— in which are not randomized to levels of the variables,
each variable is measured only at the time and therefore those variables are not isolated
it is hypothesized to operate in the psycho from other variables to which observed re
logical process— is inadequate for establish lations might be attributed. Thus, as with
ing temporal sequence. The reason is that a cross-sectional studies, statistical methods
portion of the variance in the measures is of isolation must be used. An additional
typically stable, and, by definition, stable complication is that the variables of inter
20 I. IN T R O D U C T IO N
est are measured more than once, raising latent growth models they can be estimated
the question of whether the variables to be for each research participant. Variance in
controlled must be measured repeatedly as these parameters reflects the fact that partic
well. If the influence of these confounding ipants vary in their intercepts (often, though
variables is expected to vary by time— either not always, defined as the first time point)
because scores on the variable change or be and vary in the slopes of their trajectories.
cause the variable’s influence on other vari These slopes can be treated like traditional
ables of interest varies from one time to the variables in statistical models and used as
next— then they must be measured repeat predictors, outcomes, or simple correlates.
edly and included in the statistical analysis Also, through the use of growth mixture
at the appropriate place in the model. If the modeling, participants can be grouped in
variables to be controlled are fixed charac terms of similarity in growth parameters
teristics of research participants (e.g., geno and these groups used to define and study
type), then they need be measured only once, subpopulations (Muthen & Muthen, 2000).
preferably at the initial assessment, but pos
sibly at any of the assessments.
Strategies f o r Studying Processes
A second concern is the spacing of assess
as T hey Naturally O ccur
ments. The goal of repeated assessment is
to observe and model change. On the one To this point we have described strategies by
hand, if the amount of time between assess which summary reports on dispositions and
ments is too short, it is possible that change behavior are provided in hindsight (cross-
attributable to or moderated by personality sectional and panel studies) or by which be
characteristics might not be observed. On haviors are observed in settings controlled by
the other hand, if the lag is too long, multi the researcher (experimental studies). In the
ple, undetected changes might take place be former case, it is possible that research par
tween assessments. Looking across the large ticipants do not, or cannot, accurately recall
number of longitudinal studies in the litera prior thoughts, feelings, or behaviors. In the
ture, it would seem that time between assess latter case, it is possible that the controlled
ments is more a function of convenience or environment, although powerful, does not
convention than a reasoned decision based readily generalize to the social environments
on hypotheses about the timing of a well- about which inferences are to be made. Both
articulated causal process. The strongest of these shortcomings are addressed using
longitudinal studies are those in which the strategies that allow the collection of data
spacing between assessments is thoughtfully on many occasions as the events and pro
and strategically determined. cesses of interest occur in the course of ev
For individual differences that are studied eryday life.
during a time at which they are still emerg Research using these methods is particu
ing or developing, studies that use time to larly well suited to the study of individual
model trajectories of change are useful. If differences when conceptualized as the typi
the individual-difference variable is mea cal response of the individual to behavioral
sured on three or more occasions, latent contingencies in the immediate, experienced
growth modeling can be used to distinguish environment (Mischel, Shoda, & Mendoza-
research participants in terms of their pat Denton, 2002). Although traditional mea
terns of change over time (Bollen & Cur sures of individual differences provide a
ran, 2005). For instance, data on disposi summary of these responses, they are not
tional optimism could be obtained from a suitable for capturing the typical expres
sample of children at the beginning and end sion of the individual difference and the
of their last year of middle school and first processes that account for its influence on
year of high school. Trajectories of change behavior. For this endeavor, research strat
across these four assessments could be esti egies are required that allow the detection
mated and the characteristic form (e.g., lin of within-person variance in situated behav
ear, curvilinear) determined. In the simplest ioral contingencies, expression of individual
case, the characteristic form is linear and differences, and behavior.
defined by two parameters— an intercept One such strategy is experience sampling
and a slope. It is typical to estimate these (Conner, Barrett, Tugade, & Tennen, 2007).
parameters for the sample as a whole, but in In the prototypic application of experience
2. M ethods for the Study o f Individual D ifferences 21
sampling, the experiences (i.e., thoughts, interested in the mean level of positive af
feelings, behaviors) of a relatively small fect, as well as in cross-situational variabil
number of individuals are sampled across ity around that value, for each research par
time and naturally occurring situations. ticipant (e.g., Fleeson, 2004). Alternatively,
The sampling can be random, scheduled, these descriptive statistics can be studied in
or contingent. Random sampling is accom relation to other situated variables (e.g., lo
plished by equipping research participants cation at the time of reporting) or, in multi
with an electronic device that signals them level applications, in relation to dispositions
a set number of times each day at random and other individual differences.
to provide data. In the earliest uses of this These strategies for studying naturally
strategy, participants carried pagers that sig occurring experience as it happens offer
naled them in response to randomly timed an intriguing alternative to strategies such
calls by research personnel (e.g., Csikszent- as cross-sectional or longitudinal studies,
mihalyi, Larson, & Prescott, 1977). Bor but they are not without complications and
rowing the colloquial terms for pagers, these limitations. Experience-sampling studies
studies came to be known as beeper studies. can be costly. They require signaling equip
By randomly sampling the experience of se ment, frequent interaction (planned and un
lected individuals, these studies allow the re planned) with research participants during
searcher to draw inferences about the more data collection, and expertise with statisti
general experience of the individual. cal methods for analyzing nested data. Al
In other applications, akin to the panel though methods have become more refined
studies described earlier, research partici and equipment more reliable, it still is not
pants are assessed at predetermined times uncommon to lose data from participants
(e.g., morning, evening) each day for sev due to equipment malfunction or unexpected
eral days. Signaling is required, but it may events in the lives of individual participants
be done by simply programming alarms on that alter their typical experience or make
wristwatches or mobile telephones. Because it difficult for them to faithfully provide
specific aspects of experience often are of in data. Because research participants provide
terest to researchers (and these may or may data on numerous occasions, they cannot
not be captured by random or scheduled be asked too much each time lest the experi
sampling), it may be more efficient to peg ence of participating in the study intrude on
data collection to the occurrence of specific and alter their typical experience. The con
events. For instance, if the research question cern about isolation of key variables that we
concerns social interaction, the researcher is have highlighted throughout the chapter ap
interested in the experience of research par plies in these studies as well. People cannot
ticipants only when they are engaged in so be randomized to situations; therefore, we
cial interactions. Event-contingent sampling cannot distinguish between characteristics
cannot be signaled by devices or the research of situations influencing their behavior and
team. Instead, it requires that research par their choice of those situations as opportu
ticipant be trained to recognize relevant nities to behave in desired ways. Because all
situations and accept the responsibility of variables are measured on all occasions, the
providing data when those situations arise. concern about sequencing of variables can
Data from three decades of research using be addressed using methods described ear
this strategy indicate that research partici lier for panel studies.
pants are generally reliable and responsible
in this role.
It is not uncommon for experience sam A Nod to D ata Analysis
pling studies to engage research participants
for 2 weeks, sampling six to eight times each Although our focus is methods, not analysis,
day. The researcher then possesses 80 to 100 the choice of method often dictates or con
(or more) observations of each research par strains the choice of analysis. For instance,
ticipant. How are these data to be used in re data from cross-sectional studies typically
search on individual differences? In some re are continuous, making them ill suited for
search applications, descriptive information mean-comparison analytical strategies that
at the level of the individual is of primary assume factors with two or three levels. Data
interest. For example, a researcher might be generated from panel studies with three or
22 I. IN T R O D U C T IO N
more assessments are not conducive to anal egy and data analysis, we note the impor
ysis using typical strategies such as multiple tance of drawing inferences from statistical
regression analyses. And the nonindepen results that fully account for the strengths of
dence in data from experience-sampling the methodological strategy used to gener
studies requires the use of analytical strat ate the data. Sophisticated statistical meth
egies appropriate for such data. In light of ods cannot overcome the limitations of re
this inherent link between how variables are search methods, as sometimes is assumed.
measured and how they are analyzed, we For instance, the most elegant and nuanced
recommend factoring data analysis concerns structural equation model estimated on
into decisions about how data will be col cross-sectional data cannot overcome the
lected. fact that research participants self-selected
In addition, we recommend accounting to levels on all variables and were assessed at
fully for the manner in which data were col one point in time. Conversely, the strengths
lected when analyzing them. Said differently, of research methods sometimes necessitate
researchers should avoid forcing data to fit an only rudimentary statistical analyses, as in
analytical strategy that was chosen without carefully designed experiments and research
consideration for the methodological strat questions that focus on a specific pattern of
egy by which the data were collected and means. This interplay between method and
the characteristics of those data. As noted, analysis requires that researchers keep one
a frequent and counterproductive form of in mind when considering the other.
this error is dichotomization of variables
measured on a continuum to allow means
comparisons (e.g., using ANOVA), a strate Conclusions
gy that persists despite the well-documented
loss in statistical power and potential in The strength and informativeness of evi
crease in Type I errors (Fitzsimons, 2008). If dence bearing on the relationships between
the nature of the research question and the individual differences and socially relevant
state of the literature point to comparisons responses is a direct result of the method
of means, then data should be collected in a ological strategies by which that evidence
form that anticipates this analysis, or, bet was produced. In this chapter, we have at
ter, strategies should be used for generating tempted to convey that all methods have
estimated means from analyses appropriate strengths and limitations that make them
for continuously measured variables (Aiken more or less useful, depending on the re
& West, 1991). search question and constraints that are
Other ill-advised data-analytic choices imposed by the topic, research context, and
are better categorized as missed opportuni sample. The strongest bodies of evidence
ties than as outright errors. For example, if, are those in which the individual difference
as we have advised, multiple measures (pref of interest has been measured in multiple
erably measured using different modes) of ways using multiple modes of measurement
key constructs are included, then the data and studied using a range of methodologi
analysis should capitalize on this strength cal strategies. This systematic and thorough
by modeling a latent variable that captures approach to studying individual differences
commonality in the measures while remov in social behavior ensures that variance and
ing uniqueness and random error. At a more covariance attributable to the way in which
basic level, if multiple items are available for they are measured and studied is not con
specific constructs, then a similar separation fused as variance and covariance attribut
of commonality and error can be done in able to the individual difference itself. M ore
data analysis, ensuring that effect size esti over, thoughtful designs and analyses make
mates are not attenuated by some forms of possible research syntheses that generate un
error. In short, the benefits of conscientious biased estimates of the magnitude and form
measurement are not realized until the data of relationships between individual differ
have been analyzed using methods that take ences and other variables. These estimates
full advantage of the measurement strategy. then allow more precise statements about
As a final word of caution related to the individual differences in theoretical models
association between methodological strat of their development and influence.
2. M ethods for the Study o f Individual D ifferences 23
Ii
Extraversion
J o s h u a W ilt
W il l ia m R e v e l l e
Big Five
Extraversion Agreeableness Conscientiousness Neuroticism Openness
talkative sympathetic organized tense wide interests
assertive kind thorough anxious imaginative
active appreciative planful nervous intelligent
energetic affectionate efficient moody original
-quiet -cold -careless -stable -commonplace
-reserved -unfriendly -disorderly -calm -simple
-shy -quarrelsome -frivolous -contented -shallow
-silent -hard-headed -irresponsible -unemotional -unintelligent
Characters of Theophrastus
talker anxious to please -hostile coward -stupid
chatty flatterer -shameless grumbler -superstitious
boastful -unpleasant -distrustful mean -boor
arrogant -outcast -avaricious unseasonable -gross
garrulous -offensive -reckless feckless ironical
N ote. Big Five adjectives from John (1990). The characters of Theophrastus are from Jebb’s (1909) translation. Words with
the symbol are reverse scored.
3. E xtrav ersio n 29
traversion with hysterical disorders and in dimension similar to Jung’s in that Intro
troversion with what today would be called version is described by reflective behavior.
mood disorders. Although the credit is usu However, the Extraversion pole of this scale
ally given to Jung for originating the mod is similar to EPI Extraversion, as extraverts
ern term extraversion, the less known but are described as lacking restraint and ex
very important work of Gerard Heymanns hibiting impulsive behavior. Another higher
(Eysenck, 1992) had already identified ex order factor identified by the G ZTS is called
traversion more accurately as a dimension Social Activity, which contains aspects simi
(rather than a type) along a continuum of lar to the sociability part of Eysenck’s ex
“strong” and “w eak” functioning. It is also traversion. Subsequent analyses of the struc
Heymanns whom we should credit with the ture of the EPI and the EPQ showed that the
integration of psychometric methods with biggest difference is that extraversion in the
experimental approaches to personality and EPI contains a roughly equivalent amount
with situating psychological research in the of sociability and impulsivity items, where
hypothetico-deductive method. Standing on as the EPQ contains many more sociability
the shoulders of Heymanns and those who than impulsivity items (Rocklin & Revelle,
came before him, Hans Eysenck demonstrat 1981).
ed the importance of extraversion as a fun Raymond Cattell laid the foundation for
damental dimension of personality in a se modern lexical analysis when he factor-
ries of experimental and taxom etric studies analyzed paragraph descriptors based on
in the late 1940s and early 1950s (Eysenck, Allport and Odbert’s (1936) list of traits (ex
1952; Eysenck & Himmelweit, 1947). tracted from an unabridged dictionary) to
derive 16 primary personality factors (Cat
tell, 1946), five of which cluster together to
The M easurem ent o f Extraversion form a higher order factor of Extraversion
(Cattell, 1957). The content of Cattell’s Ex
The descriptive tradition in personality, as traversion contains aspects of Eysenck’s,
mentioned before, has its roots in T heo Gray’s, and Guilford’s conceptualizations
phrastus and Galen. In the 20th century, of extraversion, as Cattell’s extravert is de
psychologists began serious efforts to mea scribed as highly impulsive, social, and as
sure the major dimensions of personality, cendant.
and all such efforts have identified extraver
sion as a major dimension.
Current Taxonomies
Eysenck was one of the first to try to de Warren Norman (1963) derived what has
scribe the core features of extraversion with come to be called the Big Five (Goldberg,
scales developed to assess personality, the 1990) factors of personality from a factor
Maudsley Personality Questionnaire (M PQ; analysis of English adjectives taken from
Eysenck, 1959), the Eysenck Personality In the dictionary. Norman’s work was based
ventory (EPI; Eysenck & Eysenck, 1968), on the prior work of Fiske (1949) and Tupes
the Eysenck Personality Questionnaire and Christal (1961) on peer ratings and his
(EPQ; Eysenck & Eysenck, 1975), and the own work on peer ratings, based on the
Eysenck Personality Profiler (EPP; Eysenck paragraph descriptors of Cattell. (These
& W ilson, 1991). Some of the items for the five factors, called Surgency— similar to
M PQ and EPI were adapted from Guilford extraversion— Agreeableness, Conscien
(Guilford & Zimmerman, 1949), which led tiousness, Neuroticism, and Openness, have
to an interesting debate as to the proper since been observed in the languages of
structure of extraversion. The instrument many different cultures; Goldberg, 1990.)
Guilford developed to measure personality, Many of the adjectives have high loadings
the Guilford-Zim m erm an Temperament on two (not one or three) factors (Hofstee,
Survey (GZTS; Guilford & Zimmerman, De Raad, & Goldberg, 1992), so that pairs
1949), identifies a higher order factor called of the Big Five dimensions have a circum-
Introversion-Extraversion, which reflects a plex structure. This structure is measured by
30 II. I N T E R P E R S O N A L D IS P O S IT IO N S
the Abridged Big Five Circumplex (AB5C), as markers of social adaptation and form a
which contains items that have a primary higher order factor resembling extraversion.
loading on one factor and secondary load The causal mechanism thought to give rise
ing on a second one. In the AB5C, Surgency to sociability and am bition are the evolu
is described mainly by the disposition to en tionary pressures “to get along” and “get
gage in approach behavior. ahead” (Hogan, 1982).
AB5C A R a d ia te joy
B FI A I see myself as so meo ne w ho is full o f energy.
GZTS A You are a happy -go-luck y individual.
H E X A C O -P I A Am usually active and full o f energy
M P Q (M ultidimensional) A Have a lot o f fun
N EO -FFI A I really enjoy ta lk in g to people.
BFAS B A m the first to act
BFM B Talkativ e
EPI B D o you like going out a lot?
EPQ B D o you like telling jokes and fun ny stories to your friends?
EPP B W ou ld you prefer to fight for you r beliefs than let an im p o rta n t
issue go unchallenged?
F F-N PQ B Picture of person riding a bu ck ing horse
IPIP B A m the life of the party
M P Q (Maudsley) B D o you like to m ix socially with people?
N EO -PI-R B I am d om in a n t, force ful, and assertive.
rkes &C Dodson, 1908), extraverts should Jeffrey Gray proposed an alternative causal
outperform introverts in highly arousing theory of extraversion, reinforcement sensi
situations (because extraverts should be less tivity theory (RST; Gray, 1970, 1 9 8 1 ,1 9 8 2 ).
prone to overarousability), and introverts Based on animal research, the original for
should outperform extraverts in low-arousal mulation of R ST postulated the existence
situations (because introverts should be less of three separate neural systems underly
prone to underarousability). For an elegant ing behavior: (1) the behavioral approach
test of this hypothesis within subjects, see system (BAS), (2) the behavioral inhibition
Anderson (1990). Second, based on Wundt’s system (BIS), and (3) the fight—flight system
notion that people try to maintain moderate (FFS). The primary emphasis was on the ef
arousal (Wundt & Judd, 1897), extraverts fects of the BIS and BAS. Sensitivity of the
should, on average, respond more and faster BAS was thought to underlie trait impulsiv-
(in order to increase their arousal) than in ity, and sensitivity of the BIS was thought
troverts during performance tasks. Indeed, to underlie trait anxiety. These traits were
the explanation of extraverted behavior as conceptualized as primary traits that to
arousal seeking provided a compelling ex gether could explain Eysenck’s higher order
planation for extraverts’ use of stimulant factor of Extraversion. Eysenck’s Extraver
drugs (cigarettes), sexual activities, and so sion was thought by Gray to be Impulsivity
cial interaction. minus Anxiety. Similar to Eysenck’s theory,
R ST makes predictions about performance,
but these predictions are more complicated
Jeffrey G ray and R einforcem ent
and harder to generalize to human research
Sensitivity T heory
because R ST was founded on animal data.
Over the past 50 years, Eysenck’s hypoth However, R ST does make straightforward
eses have generated thousands of studies predictions regarding learning and affect:
yielding varying degrees of support (M at Because extraverts should be more sensitive
thews & Gilliland, 1999). More interesting to reward than introverts, extraverts should
and more conducive to scientific progress condition faster to rewarding stimuli and
than tests of a single theory is the emergence experience more positive affect than intro
of competing theories. This happened when verts.
3. E x tra v e rsio n 33
evolutionary bases, and as is true for most facilitation is a threshold model in that dop
personality traits, extraversion is moderate amine must reach a certain level for approach
ly heritable, h2 - .4 5 -.5 0 , with little if any behavior to be elicited. Thus approach behav
shared environmental influence (Bouchard ior is thought to depend on one’s tonic level
& Loehlin, 2001). Support for extraversion of dopamine, as well as one’s phasic level
as having a substantial genetic basis is also (Depue, 1995). At present, evidence for this
garnered from the finding that extraversion model is inconsistent. The first support for
can be identified in many animal species; the theory was the finding that extraversion,
additionally, each FFM facet of extraversion as measured by the M PQ (Tellegen, 1982),
displays moderately high heritability, and correlated with prolactin indicators of dop
the relationships between extraversion fac amine functioning in 11 women (Depue, Lu-
ets are largely accounted for by genetic fac ciana, Arbisi, Collins, & Leon, 1994); this
tors (Jang, Livesley, Angleitner, Riemann, & finding was subsequently replicated with a
Vernon, 2002). There is some evidence that larger sample (Depue, 1995). Other studies
heritability for extraversion declines with do not support Depue’s theory. For example,
age (Bouchard 8c Loehlin, 2001), which Fischer, W ik, and Fredrikson (1997) mea
logically means that the environment be sured extraversion with a German adapta
comes a more important source of extraver tion (Ruch & Hehl, 1989) of the EPQ -R (S.
sion variation as people grow older. Find B. Eysenck, Eysenck, & Barrett, 1985) and
ing that extraversion is heritable is the first found that extraversion was negatively cor
step in uncovering specific genetic pathways related with subcortical brain activity in the
that influence extraversion’s development. caudate nucleus and the putamen, areas that
For example, recent research has identified have high concentrations of dopamine ter
genes that account for between-person vari minals. As it stands, the dopaminergic hy
ation in extraversion, one likely candidate pothesis provides an exciting avenue along
being AD H 4 (Luo, Kranzler, Zuo, Wang, & which to pursue the biological basis of agen
Gelernter, 2007). tic extraversion. Newly developed ways to
measure dopaminergic functioning nonin-
vasively, such as with electroencephalogra
Extraversion and Brain F u n ction /S tru cture
phy (FLEG), may serve to increase the rate at
Genes do not act directly on behavior; ge which research determines the relationships
netic effects are mediated by brain function between agentic extraversion and dopamine
and structure (Revelle, 1995). Eysenck and (Wacker, Chavanon, &c Stemmier, 2006).
Gray were the first to detail complex theo
ries about how this might be the case for
Neurophysiological and Neuroanatomical
extraversion, and recent empirical investiga
Underpinnings o f Extraversion
tions continue to advance our understanding
of the neurobiological basis of extraversion. It is clear from the section on measure
ment in this chapter that extraversion has a
positive affect component, but the biologi
The Dopaminergic Hypothesis
cal mechanisms underlying this association
o f Agentic Extroversion
are not well known. In an excellent review,
Recently, Depue (1995) developed a novel Canli (2004) describes neuroimaging stud
theory for a subcomponent of extraversion ies conducted with the aim of elucidating
labeled agentic extraversion because it en the extraversion-positive affect association.
compasses the achievement and ascendance Across a wide range of tasks, functional
aspects of extraversion (Depue & Collins, magnetic resonance imaging (fM RI) analy
1999).2 Depue’s theory closely resembles sis revealed that extraversion as measured
Gray’s original R ST in that a behavioral with the N EO -PI-R was associated with
facilitation system (BFS)— the function of greater activation in numerous areas of the
which is to increase the salience of positive brain (amygdala, caudate, mediofrontal
stimuli— is thought to be a causal basis for gyrus, right fusiform gyrus) when positive
agentic extraversion (Depue, 1995; Depue & stimuli, but not negative stimuli, were pre
Collins, 1999). Depue’s model of behavioral sented. One important implication of these
3. E xtra v e rsio n 35
studies, noted by Canli, is that personality ed by the finding that NE.O-FFI extraversion
factors such as extraversion are likely to be is inversely related to thickness of the right
widely distributed in the brain. anterior prefrontal cortex and the right fusi
Recent studies have added to our knowl form gyrus; low thickness in these areas has
edge about the activation patterns that cor been suggested as underlying impulsive and
relate with extraversion and have sought to disinhibited behavior (Wright et al., 2006).
explain such patterns. EPQ extraversion has
been associated with activation in the lateral
Tem peram ent
prefrontal cortex, lateral parietal cortex,
and right anterior cingulate cortex; each of It is clear that extraversion is associated
these brain areas is associated with task- with structure and function across many
focused self-control and discrepancy detec areas of the brain. The fact that extraversion
tion (Eisenberger, I.ieberman, & Satpute, has a strong biological component suggests
2005). Haas, Omura, Amin, Constable, and that precursors of trait extraversion should
Canli (2006) determined that the N EO -PI- appear early in development. The study of
R facets of excitement seeking and warmth temperament shows this to be the case. Tem
accounted for the association noted between perament refers to individual differences in
extraversion and anterior cingulate cortex reactivity and self-control that arise from a
activity (Canli, 2 0 0 4 ; Eisenberger et al., constitutional basis (Durbin, Klein, Hayden,
2005). Two other novel findings from this Buckley, & M oerk, 2 0 0 5 ; Rothbart, 1981).
work were that extraversion predicted func A temperament dimension of extraversion—
tional connectivity to the anterior cingulate positive affect (PA)— has been identified in
and that this association was mediated by infants as young as 3 months, in middle child
the facets of warmth, gregariousness, and hood, and even into adulthood (Rothbart,
positive emotions. The studies discussed Ahadi, & Evans, 2000). As its name implies,
up to this point have focused on predict this dimension shares characteristics with the
ing brain activity during task engagement. extraversion personality trait. For example,
Deckersbach and colleagues (2006) recently one study that factor-analyzed lower order
extended these findings by showing that, at components of temperament found that a
rest, extraversion measured by the N EO -FFI higher order extraversion/PA factor included
is associated with greater activity in the orb- sociability and positive affect components,
itofrontal cortex, which might play a part in as well as regulatory components such as in
shifting attention to positive incentives. hibitory control (Evans & Rothbart, 2007).
Differences in brain structures are also The inclusion of regulatory aspects makes
associated with extraversion, and such dif temperamental extraversion/PA especially
ferences may have diverse implications for interesting to study in the context of dynam
psychopathology, learning, and behavior. ic cognitive and behavioral processes (Evans
Magnetic resonance imaging (M RI) stud & Rothbart, 2007). In one of the few stud
ies have shown that N EO -PI-R extraversion ies to use a dynamic design, Derryberry and
is positively correlated with gray matter in Reed (1994) found that adult extraversion/
the left amygdala (Omura, Constable, & PA temperament (measured with a short ver
Canli, 20 0 5 ); as reductions in amygdalar sion of the EPQ) predicted difficulty in shift
gray matter predict depression, this finding ing attention away from positive stimuli but
may suggest that extraversion is a protec not from negative stimuli. It is interesting to
tive factor against depression (Omura et al., note that the previous findings hark back to
2005). N EO -FFI extraversion and thick notions from Eysenck’s and Gray’s conceptu
ness of orbitofrontal cortex are associated, alizations of extraversion. Inhibitory control
and extinction of fear retention mediates overlaps considerably with Eysenck’s em
the path from orbitofrontal thickness to ex phasis on the impulsivity component of ex
traversion (Rauch et al., 20 0 5 ), suggesting traversion (Eysenck, 1967), and R ST (Gray
that brain structure influences extraversion &C McNaughton, 2000) explicitly predicts
by influencing learning processes. One way that extraversion should relate to attentional
that brain structure relates to specific com biases toward positive stimuli and approach
ponents of extraverted behavior is illustrat behavior.
36 II. I N T E R P E R S O N A L D IS P O S IT IO N S
Extraversion and the A BC D s only about 30% of the total variance between
constructs (Watson, 2000). Second, behav
The previous sections can be thought of as ioral content is better represented than posi
the ontogeny of a trait, starting off as genes, tive affect in measures of extraversion (Pyt-
developing into biological structures and lik Zillig, Hemenover, & Dienstbier, 2002).
systems, and then being expressed early in Third, a study by Ashton and colleagues
life as temperament. We view the fully devel (2002) used the same method as in Lucas and
oped, higher order traits such as the Big Five colleagues (2000) and showed that the ten
as characteristic patterns of affect, behavior, dency to behave in ways that attract social
cognition, and desire. attention accounts for the common variance
among N EO -PI-R Extraversion facets.
Extraversion and positive affect might not
H ow D o Extraverts Feel?
be the same construct, but the robust rela
It is well established that extraverts feel tionship between the two calls for explana
higher levels of positive affect than intro tion. The explanations that have been of
verts (Costa &C M cCrae, 1980; Lucas & fered can be grouped into those postulating
Baird, 2 0 0 4 ; Watson & Clark, 1992). The either a primarily structural or an instru
relationship between trait extraversion and mental basis for the relationship. A struc
trait positive affect has emerged in many tural explanation means that extraverts pos
cultures with many different methods (Lucas sess some quality or characteristic that leads
&c Baird, 2 0 0 4 ), with the average correlation them to experience more happiness than in
found to be around r = .40 (Lucas & Fujita, troverts. The general structural explanation
200 0 ). Not only do measures of trait extra is described by the affect-threshold model
version predict trait positive affect, but trait (Rosenberg, 1998), which can be divided
extraversion also predicts aggregated mo into the affect-level model (Gross, Sutton,
mentary positive affect (Costa & M cCrae, & Ketelaar, 1998) and the affect-reactivity
1992a; Spain, Eaton, & Funder, 2 0 0 0 ), as model (Larsen & Ketelaar, 1991; Strelau,
well as single ratings of current positive af 1987). The affect-threshold model states
fect (Lucas & Baird, 2 0 0 4 ; Uziel, 2006). that extraverts have a lower threshold for
This means that extraverts are happier than experiencing positive affect than introverts;
introverts in general, over short time frames, that is, it should require less positive stimu
and even in the moment. lation to elicit positive affect from extraverts
It has even been proposed that extraver than from introverts. This model is general
sion is at its core the tendency to experience in that it does not distinguish between two
positive affect (Watson & Clark, 1997), and ways that equal positive stimulation could
there is some evidence to support this claim. lead to more positive affect for extraverts.
The covariation of extraversion components The first way is described by the affect-level
is accounted for by positive affect; once pos model (Gross et al., 1998), which states
itive affect is removed, the other components that because extraverts are closer to expe
of extraversion do not correlate with each riencing positive affect than introverts at
other. A similar finding reported recently baseline, they require relatively less positive
that extraversion facets that reflect reward stimulation to feel good. The second way
sensitivity load on a higher order Extraver is described by the affect-reactivity model,
sion factor that accounts for the correlations which states that extraverts and introverts
between the other facets of extraversion could feel the same amount of positive affect
(Lucas & Baird, 2004). Not only does trait at baseline but that extraverts react more
extraversion predict trait positive affect, but strongly to positive stimuli than introverts
both traits also predict similar outcomes do. It is clear that the affect-reactivity model
such as social activity, leadership, and num has its roots in R ST (Corr, 2 0 0 8 ; Gray, 1970,
ber of friends (Watson & Clark, 1997). 1981, 1982).
The evidence linking extraversion and pos Testing the two models requires identify
itive affect is very strong; however, at least ing circumstances under which they make
three findings suggest that it would be rash conflicting predictions. In the affect-level
to conceptualize extraversion and positive af model, it is assumed that extraverts have
fect as redundant constructs. First, they share a higher tonic level of positive affect; thus
3. E x tra v e rsio n 37
it predicts that extraverts should be hap spend the same amount of time in social
pier than introverts in negative-, neutral-, situations and that introverts experience just
and positive-valence situations. The affect- as much happiness as extraverts in social
reactivity model assumes that extraverts situations. It has been found that extraverts
and introverts have similar tonic levels of are happier than introverts across a variety
positive affect but that extraverts react more of both social and nonsocial situations (Di
strongly to positive stimuli; thus it predicts ener, Sandvik, Pavot, & Fujita, 1992). The
that extraverts should be happier than in between-person extraversion-positive affect
troverts in positive-valence situations only. relationship has recently been extended to
Gross and colleagues (1998) found support existing within persons as well. A within-
for both models in their seminal investiga person relationship means that an individ
tion, manipulating situation valence with ual’s momentary positive affect depends on
positive, neutral, and negative film clips. momentary levels of extraversion, or state
Recently, a meta-analysis of six studies re extraversion (Fleeson, M alanos, & Achille,
vealed that the accuracy of each model de 2002). Fleeson and colleagues (2002) found
pends on situational properties (Lucas & that all participants, regardless of trait-level
Baird, 2004). In support of the affect-level extraversion, were happier the more extra
model, extraverts were happier in neutral verted they acted. Recent studies continue to
situations. In support of the affect-reactivity support the strong link between state extra
model, extraverts’ activated positive affect version and state positive affect. Participants
(e.g., being awake, alert) but not pleasant felt more positive affect in experiments in
positive affect was more reactive to positive which participants were instructed to act
stimulation. An even more complex picture extraverted, suggesting that state extraver
emerges when the interaction of extraver sion causes state positive affect (McNiel &
sion with neuroticism on affective reactivity Fleeson, 2006). Additionally, state extraver
is taken into account, as emotionally stable sion was found to mediate the relationship
extraverts react to positive stimuli more between approach goals and state positive
strongly than neurotic extraverts (Rogers & affect (Heller, Komar, & Lee, 2007).
Revelle, 1998).
Another class of explanations for the
How D o Extraverts Behave?
extraversion-positive affect relationship
posits instrumental origins. Instrumental In the field of personality psychology, pri
explanations assume that the relationship mary importance has been placed on ex
between extraversion and positive affect is plaining behavior (Funder, 2001). Accord
based on differences in what extraverts and ing to Funder, despite the importance, little
introverts do in their daily lives. research has actually been conducted toward
Sociability theory (Watson, 1988; Watson, this aim; Funder (2001) even explicitly of
Clark, McIntyre, & Hamaker, 1992) posits fered extraversion as an example of a trait
both instrumental and structural explana that has not been investigated in relationship
tions for the extraversion-positive affect to actual behavior. However, this seems be a
relationship. Sociability theory’s intuitive in very narrow definition of behavior, restrict
strumental hypothesis is that extraverts are ed to laboratory situations, for it ignores
happier than introverts because they engage the earlier work of Eysenck, who examined
in more social activities; the complementa the factor structures of behavioral observa
ry structural explanation is that extraverts tions (Eysenck & Himmelweit, 1947), and
enjoy social activities more than introverts. the even earlier work of Heymans (Eysenck,
Some evidence has been found in support of 1992); but it w ould include the German
sociability theory, as Argyle and Lu (1990) Observational Study of Adult Twins proj
found that extraverts participate in more ect (GOSAT; Borkenau, Riemann, Angleit-
social activities than introverts and that ner, & Spinath, 2001) and Antill’s (1974)
the amount of social activity partially me observational study of talking behavior as
diated the extraversion-happiness relation a function of extraversion and group size.
ship. Some evidence, however, contradicts Recently, research has begun to address the
sociability theory. Pavot, Diener, and Fujita important goal of elucidating the content of
(1990) found that extraverts and introverts extraverted behavior.
38 II. I N T E R P E R S O N A L D IS P O S IT IO N S
As it is expected that personality traits as more extraverted (Mehl, Gosling, & Pen
manifest themselves in behavior (Funder, nebaker, 2006).
2001), the most straightforward hypothesis Although some research has been done
(relating to extraversion) resulting from this on how personality predicts actual behav
expectation is that trait extraversion should ior, there has been almost no research on
at least predict aggregate state extraversion. how personality affects dynamic patterns of
W hat little research exists suggests that in behavior in different situations. However,
dividuals with higher levels of trait extraver Eaton and Funder (2003) were able to con
sion are indeed predisposed to enact more duct a study that revealed how extraversion
extraversion states (Heller et al., 2007; influences dynamic social interactions. As in
Schutte, M alouff, Segrera, Wolf, & Rodgers, other studies, it was found that extraverts
2003). Research on how extraversion relates behaved more socially than introverts; it was
to more discrete categories of behavior is also found that extraverts influence the be
also lacking, a fact that motivated the devel havior, affect, and interpersonal judgments
opment of the Riverside Behavioral Q -Sort of those with whom they interacted, gener
(RBQ ) as a remedy (Funder, Furr, & Colvin, ally creating a more positive social environ
2000). The RBQ contains a list of behavior ment. The question of why extraverts are so
al items that can be rated for how much they socially adept is unresolved at this time, but
describe a participant’s behavior in social in one intriguing possibility is that extraverts
teractions. In a study using the R B Q , extra have certain abilities that are lacking in in
version measured with the N EO -PI (Costa & troverts. Support for this notion comes from
M cCrae, 1985) predicted behaviors that can a study that measured extraversion with the
be characterized as energetic, bold, socially EPI and found that extraverts are better at
adept, and secure (Funder et al., 2 0 0 0 ). Also nonverbal decoding than introverts when it
driven by the paucity of behavioral research, is a secondary task (Lieberman & Rosen
Paunonen and colleagues (Paunonen, 2003) thal, 2001), as may be the case in social situ
predicted various behavioral categories on ations.
the Behavioral Report Form (Paunonen &
Ashton, 2001) from extraversion as mea
H ow D o Extraverts Think?
sured by the N EO -PI-R , the N E O -FFI, and
the Five-Factor Nonverbal Personality Ques Individual differences in behavior can be as
tionnaire (FF-N PQ ; Paunonen & Ashton, sessed in various categories, as described pre
2002). Across scales, extraversion reliably viously; in contrast, individual differences in
predicted alcohol consumption, popularity, cognition are reflected in the different ways
parties attended, dating variety, and exercise that people categorize the world. Extraver
(Paunonen, 2003). sion has been found to predict differences in
One limitation of the research on specific categorization across various tasks. Broadly
behavior described thus far is that the behav speaking, extraversion relates to a relatively
iors were not collected in natural environ positive view of the world, as extraverts judge
ments. An exciting new methodology called neutral events more positively than introverts
Big EAR (electronically activated recorder; do (Uziel, 2006). Extraversion predicts cat
Mehl & Pennebaker, 2003) circumvents this egorization of words by their positive affec
problem. Big EAR is simply a small record tive quality rather than their semantic qual
ing device that is programmed to turn on ity (Weiler, 1992). For example, extraverts
and off throughout the day, recording for a are more likely to judge the words hug and
few minutes at a time, producing objective sm ile as more similar than the words smile
data in natural environments. In a study and face. Extraversion also predicts judging
using Big EAR to investigate behavioral positive valence words, for example, truth
correlates of extraverts, as well as judges’ and honesty, as more similar than negative
folk theories of extraverted behavior, it was valence words, for example, g rief and death,
found that extraversion as measured by the although extraverts are not faster to cat
Big Five Inventory (BFI; John & Srivastava, egorize positive words than negative words
1999) related to talking to and spending by valence (Rogers &c Revelle, 1998). This
time with people; additionally, judges rated finding suggests a categorization advantage
people who were more talkative and social for positive valence only when processes are
3. E xtrav ersio n 39
competing. Extraversion also does not relate area revealed that extraversion is generally
to classifying rewards faster than threats; associated with high motivation for social
however, among people scoring low on IPIP contact, power, and status (Olson & Weber,
extraversion, quickness to classify threaten 20 0 4), personal strivings (Emmons, 1986)
ing stimuli was related to experiencing nega for intimacy and interdependence (King,
tive affect in daily life (Robinson, Meier, & 1995), and wishing for higher levels of posi
Vargas, 2005). In this study, quickness to tive affect and interpersonal contact (King
classify threatening stimuli did not relate to & Broyles, 1997).
negative affect among individuals scoring It was recently suggested that the correct
high in extraversion, suggesting that extra level of abstraction for investigating the rela
version might be a protective factor against tionship between desire and a broad, higher
sensitivity to threat. order trait such as extraversion is probably
One concern that might be raised is that not at the relatively narrow level of con
concurrent mood might be responsible for cepts such as personal strivings and wishes
the cognitive differences described here. An but rather at the broad level of major life
example of how mood affects cognition is goals (Roberts & Robins, 2000). At this
given by a study finding that state positive level, N EO -FFI extraversion relates to hav
affect predicts classification of objects by ing more economic (e.g., status and accom
their broad, global features over their local plishment), political (e.g., influencing and
features (Gasper &c Clore, 2002). Studies leading), and hedonistic (e.g., fun and ex
examining the combined effects of extraver citement) goals (Roberts & Robins, 2000).
sion and positive affect are in their beginning These findings were subsequently replicated
stages, and, as such, results are quite compli in another study finding that N EO -FFI ex
cated as this point. Although EPQ Extraver traversion was related to social goals (R ob
sion had a positive main effect on choosing erts & Robins, 2004). This study also deter
positive-valence homophones over neutral mined that positive increases in extraversion
homophones, on completing open-ended sto in early adulthood were related to assigning
ries with more positive tone, and on recalling increased importance to economic, aesthet
more positive than neutral or negative words ic, social, economic, political, and hedonistic
in a free-recall task, this effect was positively goals. These initial findings suggest that mo
moderated by current positive affect when tivation, especially at the level of broad life
positive affect was experimentally induced, goals, is an area ripe for important discover
but not when mood was allowed to vary free ies that is largely untapped at this point.
ly (Rusting, 1999). A different study found
that an extraversion composite consisting of
the EPQ, BAS/BIS scales, and the General Extraversion and Psychopathology
ized Reward and Punishment Expectancy
Scales (GRAPES; Ball & Zuckerman, 1990) In general, the importance of studying the
was related to beliefs that positive events relationships between normal personality
were more likely in the future (Zelenski & and psychopathology rests on the possibility
Larsen, 2002). Extraversion in this study that personality factors could indicate early
did not interact with naturally occurring or and persistent risk for the development of
experimentally manipulated positive mood, psychopathology (Krueger, Caspi, M offitt,
but a unique main effect of positive affect Silva, & M cGee, 1996; M arkon, Krueger,
emerged when mood was experimentally & Watson, 2005). Recently renewed inter
manipulated. Future research will need to est in the relationships between normal and
employ clever methods in order to clarify the abnormal personality have led to investiga
complex relationships of extraversion and tions of how extraversion relates to various
positive affect to cognition. forms of psychopathology (Widiger, 2005).
As a general dimension of personality, ex
traversion most obviously has implications
W h at D o Extraverts W ant?
for personality disorders; a personality dis
Comparatively little work has examined order is defined by the D SM -IV-TR as “an
motives and goals that are associated with enduring pattern of inner experience and
extraversion. Initial investigation into this behavior” that is “stable and of long dura
40 II. I N T E R P E R S O N A L D IS P O S IT IO N S
tion, and its onset can be traced back at least optimistic that the coming research on ex
to adolescence or early adulthood” (Ameri traversion will prove even more innovative
can Psychiatric Association, 2 0 0 0 , p. 689). and important, and we offer three areas that
In general, low extraversion is negatively promise to be particularly fruitful. First, re
correlated with the presence of personality search should investigate how extraversion
disorders, but this finding is not universal, is implicated in ongoing functioning. We
as there are some studies implicating high echo Funder’s call for more behavioral stud
extraversion in certain personality disor ies employing both self- and other reports
ders (Widiger, 20 0 5 ); see Costa and Widi- (Funder, 2001), as well as the continued
ger (2002) for a diverse set of reviews. That development of unobtrusive methods such
both high and low extraversion relate to per as Big EAR (Mehl & Pennebaker, 2003).
sonality disorders is reminiscent of N ettle’s O f particular interest will be studies that
suggestion that both poles of normal person investigate social processes in terms of the
ality dimensions involve costs and benefits dynamic state manifestations of behavior,
(Nettle, 2006). feelings, thoughts, and desires. A second
Although Hans Eysenck had examined area of investigation that we believe shows
the importance of extraversion in psychiatric great promise is testing the new R ST (Gray
diagnoses (Eysenck & Himmelweit, 1947) & McNaughton, 2000). We believe that
and continued to emphasize the application RST could become the unifying theory for
of normal personality traits to psychopa extraversion research, as it has implications
thology (Eysenck, 1957), recent investiga for studies at every level of personality re
tions of the relationships between normal search, from genetics and brain structure
personality and psychopathology outside of to patterns of thoughts and behavior. We
the personality disorders began in earnest encourage future investigations to integrate
with the groundbreaking study of Trull and research between different levels in the at
Sher (1994). They measured normal person tempt to elucidate mediating pathways; for
ality with the N EO -FFI and showed that example, it may be possible to find genetic
low extraversion, unique among the FFM markers of brain structures that are impli
dimensions, predicted depression and anxi cated in the BIS, BAS, and FFS (Corr, 2 0 0 8 ;
ety. Krueger and colleagues (1996) exam Smillie, 2008). The third area we highlight
ined how M PQ (Tellegen, 1982) dimensions is the growing availability of public-domain
were related to psychological disorders; in personality assessments, specifically the IPIP
regard to extraversion, the Social Close item pool (Goldberg et al., 20 0 6 ). The abil
ness scale was negatively related to conduct ity to obtain a large quantity of data in a rel
disorder, affective disorders, and substance atively short period of time (Goldberg et al.,
use disorders, whereas the Social Potency 2006) makes public-domain assessment the
scale was positively related to conduct dis method of choice for investigating the fol
order and substance abuse disorders. More lowing questions: W hat extraversion scales
recent research has looked specifically at and items have the best predictive validity
extraversion’s role in anxiety and depressive for various domains such as health, occupa
disorders, with one study finding EPI E xtra tional success, and interpersonal function
version to be negatively related to anxiety ing? W hat are the lower order facets or as
and major depressive disorder but that the pects that extraversion encompasses? How
relationship to anxiety did not remain when does extraversion content fit into higher
statistically controlling for gender, age, and order factors of personality? The first data
education (Jylha & Isometsa, 20 0 6 ). using public domain assessment to address
these questions have recently been reported
(DeYoung et al., 2007; Grucza & Goldberg,
Extraversion and the Future 2 0 0 7 ; Revelle, W ilt, & Rosenthal, in press).
their propensity to act bold, talkative, and Argyle, M ., & I.u, L. (1990). The happiness of ex
traverts. P erson ality a n d In d iv id u a l D iffe r e n c e s ,
assertive. Twenty-five hundred years later,
//(10), 1011-1017.
psychologists armed with advanced psycho Ashton, M. C., & Lee, K. (2001). A theoretical basis
metric techniques are building a scientific for the major dimensions of personality. E u r o p ea n
paradigm around the construct in which J o u r n a l o f P erson ality , 15(5), 3 2 7 - 3 5 3 .
Ashton, M. C., Lee, K., & Paunonen, S. V. (2002).
the Greeks were interested. Rooted in one’s
What is the central feature of extraversion?: Social
genes, brain structure and function, and attention versus reward sensitivity. J o u r n a l o f P er
early temperament is the personality trait of so n a lity a n d S o c ia l P sy ch olog y , §3(1), 2 4 5 - 2 5 1 .
extraversion. Similar to any other personal Ball, S. A., & Zuckerman, M . (1990). Sensation seek
ity trait, extraversion is expressed in indi ing, Eysenck’s personality dimensions and reinforce
ment sensitivity in concept formation. P erson ality
vidual differences in a person’s character a n d In d iv id u a l D iffe r e n c e s , 11(4), 3 4 3 - 3 5 3 .
istic patterns of feelings, actions, thoughts, Borkenau, P., Riemann, R., Angleitner, A., & Spinath,
and goals. We are encouraged by the recent F. M. (2001). Genetic and environmental influences
progress and growing interest in extraver on observed personality: Evidence from the German
observational study of adult twins. J o u r n a l o f P er
sion, and we are confident that, as personal so n a lity a n d S o c ia l P sy ch olog y , 80(4), 6 5 5 - 6 6 8 .
ity theory and research methods continue to Bouchard, T. J ., & I.oehlin, J. C. (2001). Genes, evo
become more accurate and precise, an even lution, and personality. B e h a v io r G e n e tic s , i f (3),
greater array of extraversion’s implications 243-273.
Buss, D. M. (1995). Evolutionary psychology: A new
across a wide variety of social, occupational,
paradigm for psychological science. P sy ch o lo g ica l
and clinical contexts will be revealed. In q u iry , 6(1), 1 -3 0.
Canli, T. (20 04 ). Functional brain mapping of extra
version and neuroticism: Learning from individual
A ckn o w ledgm en t differences in emotion processing. J o u r n a l o f P er
so n a lity , 72(6), 1 10 5-1 1 3 2 .
Cattell, R. B. (1946). D escrip tio n a n d m ea su re m en t o f
W e w o u ld like t o t h a n k A llen R o s e n t h a l f o r help
p er so n a lity . Oxford, UK: World Book.
w i th ea r l i e r d ra ft s . Cattell, R. B. (1957). P erson ality a n d m o tiv a tio n str u c
tu re a n d m ea su re m en t. Oxford, UK: World Book.
Corr, P. J . , & McNaughton, N. (2008). The reinforce
N otes ment sensitivity theory. In P. J. Co rr (Ed.), T h e
r e in fo r c e m e n t sen sitiv ity th e o r y o f p er so n a lity
(pp. 1 5 5 -18 7). Cambridge, UK: Cambridge Univer
1. Alth ough occasion ally one will see extro ver-
sity Press.
sio n-in troversion , the preferred spelling in psy
Costa, P. T., & Mc Crae, R. R. (1980). Influence of ex
ch ological research is e x traversion -in tr oversion.
traversion and neuroticism on subjective well-being:
For pu rposes o f brevity, we refer to the bipolar Happy and unhappy people. J o u r n a l o f P erson ality
d imension o f in tro versio n-extraversion by refer a n d S o c ia l P sy ch o lo g y , 3S(4), 6 6 8 - 6 7 8 .
ring to just one end o f it, extraversion . Costa, P. T., & Mc Crae, R. R. (1985). N E O PI p r o fe s
2 . T h e ne urobiology o f D ep ue’s “affiliative e x s io n a l m a n u a l. Odessa, FL: Psychological Assess
travers ion,” enco m pass in g w arm th and social ment Resources.
clos enes s, has only recently received research Costa, P. T., & M cCrae, R. R. (1992a). Four ways five
atte ntio n but is generally th o ug ht to be based factors are basic. P erson ality a n d In d iv id u a l D iffe r
on op iate functioning (Depu e & M o rro n e- e n c e s , 13(6), 6 5 3 - 6 6 5 .
Strupinsky, 2 0 0 5 ) Costa, P. T., & McCrae, R. R. (1992b). N E O P I-R
p r o fe s s io n a l m a n u a l. Odessa, FL: Psychological As
sessment Resources.
Co sta, P. T., & Widiger, T. A. (2002). P erso n a lity d is
R eferen ces o r d er s a n d th e fiv e - fa c t o r m o d e l o f p e r s o n a lity (2nd
ed.). Washington, DC: American Psychological As
Allport, G. W., & Odbert, H. S. (1936). Trait names: sociation.
A psycholexical study. P sy c h o lo g ic a l M o n o g r a p h s , Deckersbach, T., Miller, K. K., Klibanski, A., Fis-
47(211). chman , A., Dougherty, D. D., Blais, M. A., et al.
American Psychiatric Association. ( 2 00 0 ). D ia g n o stic (20 0 6). Regional cerebral brain metabolism corre
a n d s ta tistic a l m a n u a l o f m e n ta l d is o r d e r s (4 th ed., lates of neuroticism and extraversion. D ep r es sio n
text rev.). Washington, DC: Author. a n d A n x iety , 2.5(3), 1 3 3 - 1 3 8 .
Anderson, K. J . (1990). Arousal and the inverted-w Depue, R. A. (1995). Neurobiological factors in per
hypothesis: A critique of Neiss’s reconceptualizing sonality and depression. E u r o p e a n J o u r n a l o f P er
arousal. P s y c h o lo g ic a l B u lletin , 207(1), 9 6 - 1 0 0 . so n a lity , 9(5), 4 1 3 - 4 3 9 .
Antitl, J. K. (1974). The validity and predictive power Depue, R. A., & Collins, P. F. (1999). Neurobiology of
of introversion-extraversion for quantitative as the structure of personality: Dopamine, facilitation
pects of conversational patterns. D iss er ta tio n A b of incentive motivation, and extraversion. B e h a v
stracts In te rn a tio n a l, i5(1-B), 5 32 io r a l a n d B rain S cien ces, 22(3), 4 9 1 - 5 6 9 .
42 II. I N T E R P E R S O N A L D IS P O S IT IO N S
Depue, R. A., Luciana, M ., Arbisi, P., Collins, P., & Eysenck, H. J., & Himmelweit, H. T. (1947). D im en
Leon, A. (1994). Dopamine and the structure of sio n s o f p e r s o n a lity : A r e c o r d o f resea rch c a r r ie d
personality: Relation of agonist-induced dopamine o u t in c o lla b o r a tio n w ith H. T. H im m elw e it [an d
activity to positive emotionality. J o u r n a l o f P ers o n oth ers]. London: K. Paul, Trench.
a lity a n d S o c ia l P sy ch olog y , 67(3), 4 8 5 - 4 9 8 . Eysenck, H. J ., & Wilson, G. D. (20 00 ). T h e E y sen ck
Depue, R. A., & Morrone-Strupinsky, J. V. (2005). P erson ality P ro filer (Version 6). Worthing, UK: Psi-
A neurobehavioral model of affiliative bonding: Press.
Implications for conceptualizing a human trait of Eysenck, S. B., & Eysenck, H. J . (1975). M an u al o f the
affiliation. B e h a v io r a l a n d B rain S cien c es, 28(3), E y sen ck P erson ality Q u estio n n a ir e. London: Hod-
313-395. der & Stoughton.
Derryberry, D., &c Reed, M . A. (1994). Temperament Eysenck, S. B., Eysenck, H. J . , & Barrett, P. (1985). A
and attention: Orienting toward and away from revised version of the Psychoticism scale. P e r s o n a l
positive and negative signals. J o u r n a l o f P erson ality ity a n d In d iv id u a l D iffer en c es , 6(1), 2 1 - 2 9 .
a n d S o c ia l P sy ch olog y , 6 6 (6 ), 1 1 2 8 - 1 1 3 9 . Fischer, H., Wik, G., & Fredrikson, M . (1997). E x
DeYoung, C. G., Quilty, L. C., & Peterson, J. B. (2 007). traversion, neuroticism and brain function: A PET
Between facets and domains: 10 aspects of the Big studv of personality. P erso n a lity a n d In d iv id u a l
Five. J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , D iffe r e n c e s , 2 3 (2 ), 3 4 5 - 3 5 2 .
93(5), 8 8 0 - 8 9 6 . Fiske, D. W. (1949). Consistency of the factorial struc
Diener, E., Sandvik, E., Pavot, W., & Fujita, F. (1992). tures of personality ratings from different sources.
Extraversion and subjective well-being in a U.S. na J o u r n a l o f A b n o r m a l a n d S o c ia l P sy ch olog y , 4 4 ,
tional probability sample. J o u r n a l o f R e se a rc h in 329-344.
P erson ality , 26(3), 2 0 5 - 2 1 5 . Fleeson, W., Malanos, A. B., & Achille, N. M. (2002).
Digman, J. M. (1990). Personality structure: Emer An intraindividual process approach to the rela
gence of the five-factor model. A n n u al R e v ie w o f tionship between extraversion and positive affect:
P sy ch olog y , 4 1 , 4 1 7 - 4 4 0 . Is acting extraverted as “good” as being extravert
Durbin, C., Klein, D. N., Hayden, E. P., Buckley, M. ed? J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y ,
E., & Mo erk, K. C. (2 005). Temperamental em o 8 3(6), 1 4 0 9 - 1 4 2 2 .
tionality in preschoolers and parental mood dis Funder, D. C. (2001). Personality. A n n u al R e v iew o f
orders. J o u r n a l o f A b n o r m a l P sy ch olog y , 114(1), P sy ch olog y , 5 2 , 1 9 7 - 2 2 1 .
28-37. Funder, D. C., Furr, R., & Colvin, C. (20 00 ). The Riv
Eaton, I.. G., & Funder, D. C. (2003). The creation and erside Behavioral Q-So rt: A tool for the description
consequences of the social world: An interactional of social behavior. J o u r n a l o f P erson ality , 68(3),
analysis of extraversion. E u r o p ea n J o u r n a l o f P er 451-489.
so n a lity , 17(5), 3 7 5 - 3 9 5 . Gasper, K., & Clore, G. L. (2002). Attending to the
Eisenberger, N. I., Lieberman, M. D., & Satpute, A. big picture: M ood and global versus local process
B. (2 005). Personality from a controlled processing ing of visual information. P s y c h o lo g ic a l S cien ce,
perspective: An f M R I study of neuroticism, extra 1.5(1), 3 4 - 4 0 .
version, and self-consciousness. C o g n itiv e, A ffe c Goldberg, L. R. (1990). An alternative “description of
tive a n d B e h a v io r a l N eu r o s c ie n c e , 5(2), 1 69 -1 8 1 . personality”: The Big-Five factor structure. J o u r
Emmons, R. A. (1986). Personal strivings: An ap n a l o f P erson ality a n d S o c ia l P sy ch o lo g y , 5 9(6),
proach to personality and subjective well-being. 1216-1229.
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , Goldberg, L. R. (1992). The development of markers
51(5), 1 0 5 8 - 1 0 6 8 . for the Big-Five factor structure. P sy c h o lo g ic a l As
Evans, D. E., & Rothbart , M . K. ( 2 00 7). Developing a sessment, 4(1), 2 6 - 4 2 .
model for adult temperament. J o u r n a l o f R e se a rch Goldberg, L. R. (1999). A broad-bandwidth, pub
in P erson ality , 41(4), 8 6 8 - 8 8 8 . lic domain, personality inventory measuring the
Eysenck, H. J. (1952). T h e scie n tific stu dy o f p e r s o n a l lower-level facets of several five-factor models. In I.
ity. London: Routledge & Kegan Paul. Mervielde, I. Deary, F. De Fruyt, & F. Ostendorf
Eysenck, H. J. (1957). T h e d y n a m ic s o f a n x ie ty a n d (Eds.), P erson ality p s y c h o lo g y in E u r o p e (Vol. 7,
h y s te ria : An e x p e r im e n ta l a p p lic a tio n o f m o d e rn pp. 7 - 2 8 ). Tilburg, The Netherlands: Tilburg Uni
lea rn in g th e o r y to p sy ch ia try . Oxford, UK: Prae- versity Press.
ger. Goldberg, L. R ., Joh nson, J. A., Eber, H. W., Hogan,
Eysenck, H. J . (1959). The “Maudsley Personality In R., Ashton, M. C., Cloninger, C. R., et al. (2 006).
ventory” as determinant of neurotic tendency and The International Personality Item Pool and the fu
extraversion. Z e its c h r ift fu r E x p e r im e n te lle u n d ture of public-domain personality measures. J o u r
A n g ew a n d te P sy c h o lo g ie , 6, 1 6 7 - 1 9 0 . n a l o f R e se a rc h in P erson ality , 40(1), 8 4 - 9 6 .
Eysenck, H. J . (1967). T h e b io lo g ic a l b a sis o f p e r s o n Gray, J. A. (1970). The psychophysiological basis of
ality . Springfield, IL: Thomas. introversion-extraversion. B e h a v io u r R e se a rc h a n d
Eysenck, H. J . (1981). General features of the model. T h era p y , 8(3), 2 4 9 - 2 6 6 .
In H. J. Eysenck (Ed.), A m o d e l fo r p e r s o n a lity Gray, J. A. (1981). A critique of Eysenck’s theory of
(pp. 1 -3 7 ). Berlin, Germany: Springer-Verlag. personality. In H. J . Eysenck (Ed.), A m o d e l fo r p e r
Eysenck, H. J . (1992). A h u n d re d y ea rs o f p e r s o n a lity s o n a lity (pp. 2 4 6 - 2 7 7 ) . Berlin, Germany: Springer-
resea rch , fr o m H ey m a n s to m o d e r n tim es. Houten, Verlag.
The Netherlands: Bohn. Gray, J. A. (1982). N eu r o p s y c h o lo g ic a l th e o r y o f a n x i
Eysenck, H. J ., Sc Eysenck, S. B. G. (1968). M an u al fo r e ty : A n in v estig ation o f th e s e p t a l-h ip p o c a m p a l sy s
th e E y sen c k P erso n a lity In v en to ry . San Diego, CA: tem . Cambridge, UK: Cambridge University Press.
Educational and Industrial Testing Service. Gray, J . A., & McNaughton, N. (20 00 ). T h e n e u r o p
3. E xtrav ersio n 43
sy c h o lo g y o f a n x iety : An en q u iry in to th e fu n c tio n s King, L. A., & Broyles, S. J . (1997). Wishes, gender,
o f th e s e p t o - h ip p o c a m p a l sy stem (2nd ed.). Oxford, personalitv, and well-being. J o u r n a l o f P erson ality ,
UK: Oxford University Press. 65(1), 4 9 - 7 6 .
Gross, J . J . , Sutton, S. K., & Ketelaar, T. (1998). Re Krueger, R. F\, Caspi, A., Moffitt, T. F.., Silva, P. A.,
lations between affect and personality: Support & M cGee , R. (1996). Personality traits are differ
for the affect-level and affective reactivity views. entially linked to mental disorders: A multitrait-
P erson ality a n d S o c ia l P sy c h o lo g y B u lletin , 24(3), multidiagnosis study of an adolescent birth cohort.
279-288. J o u r n a l o f A b n o r m a l P sy ch olog y , 105(3), 2 9 9 -
Grucza, R. A., & Goldberg, L. R. ( 2 0 0 7 ). The compar ill.
ative validity of II modern personality inventories: Larsen, R. J ., & Ketelaar, T. (1991). Personality and
Predictions of behavioral acts, informant reports, susceptibility to positive and negative emotional
and clinical indicators. J o u r n a l o f P erson ality As states. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o l
sessment, 89(2), 167-187. ogy, 67(1), 1 3 2 - 1 4 0 .
Guilford, J . P., & Zimmerman, W. S. (1949). T h e Lee, K., & Ashton, M . C. (20 04 ). Psychometric prop
G u ilfo r d - Z im m e r m a n tem p er a m en t su rvey. O x erties of the HF1XACO personality inventory. M u l
ford, UK: Sheridan Supply. tiv a ria te B eh a v io r a l R esea rc h , 3 9 (2 ), 3 2 9 - 3 5 8 .
Haas, B. W., Omura, K., Amin, Z ., Constable, R., & Lieberman, M. IX, & Rosenthal, R. (2001). Why intro
Canli, T. (20 0 6). Functional connectivity with the verts can’t always tell who likes them: Multitasking
anterior cingulate is associated with extraversion and nonverbal decoding. J o u r n a l o f P erso n a lity a n d
during the emotional Stroop task. S o c ia l N e u r o s c i S o c ia l P sy ch olog y , 8 0 (2 ), 2 9 4 - 3 1 0 .
en c e , 1(1), 1 6 - 2 4 . Lucas, R. E., & Baird, B. M. (2 004). Extraversion and
Heller, D., Komar, J., & Lee, W. B. ( 2 0 07 ). The dy emotional reactivity. J o u r n a l o f P erson ality a n d S o
namics of personality states, goals, and well-being. cia l P sy ch olog y , 86(3), 4 7 3 - 4 8 5 .
P erso n a lity a n d S o c ia l P sy ch o lo g y B u lletin , 33(6), Lucas, R. E., Diener, E., Grob, A., Suh, F. M., & Shao,
898-910. L. ( 2 0 00 ). Cross-cultural evidence for the funda
Hofstee, W. K., De Raad, B., & Goldberg, I.. R. mental features of extraversion. J o u r n a l o f P ers o n
(199 2). Integration of the big five and circumplex a lity a n d S o c ia l P sy ch olog y , 79(3), 4 5 2 - 4 6 8 .
approaches to trait structure. J o u r n a l o f P erson ality Lucas, R. FI., & Fujita, F. ( 2 0 00 ). Factors influencing
a n d S o c ia l P sy ch olog y , 6 J(1 ), 1 4 6 - 1 6 3 . the relation between extraversion and pleasant af
Hogan, R. (1982). A socioanalytic theory of personal fect. Jo u r n a l o f P erson ality a n d S o c ia l P sy ch olog y ,
ity. In R. A. Dienstbier & M . M. Page (Eds.), P er 79(6), 1 0 3 9 - 1 0 5 6 .
s o n a lity : C u rren t th e o r y a n d research . N e b r a s k a Luo, X ., Kranzler, H. R., Zu o, L., Wang, S., & Gelern-
S y m p osiu m o n M otiv atio n (Vol. 3 0 , pp. 5 5 - 8 9 ) . ter, J. (20 07 ). Personality traits of agreeableness and
Lincoln: University of Nebraska Press. extraversion are associated with A D H 4 variation.
Hull, C. L. (1943). P rin cip les o f b e h a v io r : An in tr o B io lo g ic a l P sychiatry, 61(5), 5 9 9 - 6 0 8 .
d u c tio n to b e h a v io r th eo ry . Oxford, UK: Appleton- MacDon ald, K. (1995). Evolution, the five-factor
Century. model, and levels of personality. J o u r n a l o f P erso n
Jang, K. L., Livesley, W., Angleitner, A., Riemann, R., ality, 63(3), 5 2 5 - 5 6 7 .
& Vernon, P. A. (2002). Genetic and environmen Markon, K. E., Krueger, R. F., &c Watson, D. (2005).
tal influences on the covariance of facets defining Delineating the structure of normal and abnormal
the domains of the five-factor model of personal personality: An integrative hierarchical approach.
ity. P erso n a lity a n d In d iv id u a l D iffer en c es , 33(1), J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y ,
83-101. 88(1), 139-157.
John, O. P. (1990). The “ Big Five” factor taxonomy: Matthews, G. (1992). Extroversion. In A. P. Smith &
Dimensions of personality in the natural language D. M . Jones (Eds.), H a n d b o o k o f h u m a n p e r fo r
and in questionnaires. In I.. A. Pervin & O. P. Joh n m a n c e : Vol. 3. S ta te a n d trait (pp. 9 5 - 1 2 6 ) . San
(Fids.), H a n d b o o k o f p e r s o n a lity : T h e o r y a n d r e Diego, CA: Academic Press.
se a r c h (pp. 6 6 - 1 0 0 ) . New York: Guilford Press. Matthews, G., & Gilliland, K. (1999). The personality
John, O. P., Donahue, E. M., & Kentle, R. L. (1991). theories of H. J. Flysenck and J. A. Gray: A compara
T h e B ig F iv e In v e n to ry — V ersion s 4 a a n d 54. Berke tive review. P erson ality a n d In d iv id u a l D iffer en c es ,
ley: Institute of Personality and Social Research, 26(4), 5 8 3 - 6 2 6 .
University of California, Berkeley. McCrae, R. R., & Costa, P. T. (1997). Personality trait
Joh n, O. P., & Srivastava, S. (1999). The Big Five trait structure as a human universal. A m erica n P sy ch o l
taxonomy: History, measurement, and theoretical o g ist, 52(5), 5 0 9 - 5 1 6 .
perspectives. In L. A. Pervin & O. P. John (Eds.), Mc Niel, J., & Fleeson, W. (20 0 6). The causal effects
H a n d b o o k o f p e r so n a lity : T h e o r y a n d research of extraversion on positive affect and neuroticism
(2nd ed., pp. 1 0 2 - 1 3 8 ). New York: Guilford Press. on negative affect: Manipulating state extraver
Jung, C. G. (1971). P s y c h o lo g ic a l ty p es: C o lle c te d sion and state neuroticism in an experimental ap
w o r k s (Vol. 6). Princeton, NJ: Princeton University proach. J o u r n a l o f R esea rch in P erson ality , 4 0 (5 ),
Press. (Original work published 1921) 529-550.
Jylha, P., & Isometsa, E. (20 06 ). The relationship of Mehl, M. R., Gosling, S. D., 8c Pennebaker, J. W.
neuroticism and extraversion to symptoms of a n x i (2 0 06 ). Personality in its natural habitat: Manife s
ety and depression in the general population. D e tations and implicit folk theories of personality in
p re ss io n a n d A n x iety , 2 3 (5 ), 2 8 1 - 2 8 9 . dailv life. J o u r n a l o f P erson ality a n d S o c ia l P sy
King, L. A. (1995). Wishes, motives, goals, and person c h o lo g y , 90(5), 8 6 2 - 8 7 7 .
al memories: Relations of measures of human moti Mehl, M. R., & Pennebaker, J. W. (2 003). The sounds
vation. J o u r n a l o f P erson ality , 6.3(4), 9 8 5 - 1 0 0 7 . of social life: A psychometric analysis of students’
44 II. I N T E R P E R S O N A L D IS P O S IT IO N S
daily social environments and natural conversa Revelle, W., Humphreys, M . S., Simon, L., & Gil
tions. J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , liland, K. (1980). The interactive effect of personal
8 4 (4 ), 8 5 7 - 8 7 0 . ity, time of day and caffeine: A test of the arousal
Morley, H. (1891). C h a ra c ter w ritin gs o f th e s e v e n model. J o u r n a l o f E x p e r im e n ta l P sy ch o lo g y : G e n
tee n th cen tu ry . London: Kessinger. era l, 109, 1—31.
Nettle, D. (2 005). An evolutionary approach to the Revelle, W., Wilt, J ., & Rosenthal, A. (in press). Per
extraversion continuum. E v o lu tio n a n d H u m an B e sonality and cognition: The personality-cognition
h a v io r, 2 6 (4 ), 3 6 3 - 3 7 3 . link. In A. Gruszka, G. Matthews, & B. Szymura
Nettle, D. ( 2 00 6). The evolution of personality varia (Eds.), H a n d b o o k o f in d iv id u a l d iffe r e n c e s in c o g
tion in humans and other animals. A m er ica n P sy n itio n : A tten tio n , m e m o r y a n d ex e c u tiv e c o n tro l.
c h o lo g is t, 61(6), 6 2 2 —631. New York: Springer.
No rm an, W. T. (1963). Toward an adequate taxonomy Roberts, B. W., Kuncel, N. R ., Shiner, R., Caspi, A.,
of personality attributes: Replicated factors struc & Goldberg, L. R. (20 07 ). T he power of personal
ture in peer nomination personality ratings. J o u r ity: The comparative validity of personality traits,
n a l o f A b n o r m a l a n d S o c ia l P sy ch o lo g y , 6 6(6), socioeconomic status, and cognitive ability for pre
574-583. dicting important life outcomes. P ersp ectiv es o n
Olson, K. R., & Weber, D. A. (200 4). Relations be P sy c h o lo g ic a l S c ien ce, 2(4), 3 1 3 - 3 4 5 .
tween big five traits and fundamental motives. Psy Roberts, B. W., &C Robins, R. W. (20 00 ). Broad dispo
c h o lo g ic a l R e p o r ts , 95(3), 7 9 5 - 8 0 2 . sitions, broad aspirations: The intersection of per
Omura, K., Constable, R., & Canli, T. (2005). sonality traits and major life goals. P erson ality a n d
Amygdala gray matter concentration is associated S o c ia l P sy ch o lo g y B u lletin , 26(10), 1 2 8 4 - 1 2 9 6 .
with extraversion and neuroticism. N e u r o R e p o r t, Roberts, B. W., & Robins, R. W. (2 004). Person-
16, 1 9 0 5 - 1 9 0 8 . environment fit and its implications for personality
Ortony, A., Norman, D. A., & Revelle, W. (2 005). Ef development: A longitudinal study. J o u r n a l o f P er
fective functioning: A three-level model of affect, so n a lity , 72(1), 8 9 - 1 1 0 .
motivation, cognition, and behavior. In J. Fellous & Robinson, M. D., Meier, B. P., & Vargas, P. T. (2005).
M . Arbib (Eds.), W ho n ee d s e m o t io n s ?: T h e brain Extraversion, threat categorizations, and negative
m ee ts th e m a c h in e (pp. 1 7 3 - 2 0 2 ). New York: O x affect: A reaction time approach to avoidance moti
ford University Press. vation. J o u r n a l o f P erson ality , 73(5), 1 3 9 7 - 1 4 3 6 .
Ozer, D. J ., & Benet-Marti'nez, V. ( 2 0 06 ). Personality Rocklin, T., & Revelle, W. (1981). T he measurement of
and the prediction of consequential outcomes. A n extraversion: A comparison of the Eysenck Person
n u al R e v ie w o f P sy ch olog y , 57, 4 0 1 - 4 2 1 . ality Inventory and the Eysenck Personality Ques
Paunonen, S. V. (2003). Big five factors of personal tionnaire. B ritish Jo u r n a l o f S o c ia l P sy ch olog y ,
ity and replicated predictions of behavior. Jo u r n a l 20 (4 ), 2 7 9 - 2 8 4 .
o f P erso n a lity a n d S o c ia l P sy ch olog y , 8 4 (2 ), 4 1 1 - Rogers, G. M ., & Revelle, W. (1998). Personality,
422. mood, and the evaluation of affective and neutral
Paunonen, S. V., &C Ashton, M. C. (2001). Big five fac word pairs. J o u r n a l o f P erso n a lity a n d S o c ia l P sy
tors and facets and the prediction of behavior. J o u r c h o lo g y , 74(6), 1 5 9 2 - 1 6 0 5 .
n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 81(3), Rosenberg, E. L. (1998). Levels of analysis and the
524-539. organization of affect. R e v ie w o f G e n e ra l P sy ch o l
Paunonen, S. V., & Ashton, M . C. (2 002). T h e n o n og y , 2(3), 2 4 7 - 2 7 0 .
v er b a l a s s e ss m e n t o f p e r s o n a lity : T h e N P Q a n d the Rothbart, M. K. (1981). Measurement of temperament
FF-N P Q ,. Ashland, OH : Hogrefe & Huber. in infancy. C h ild D e v e lo p m e n t, 5 2(2), 5 6 9 - 5 7 8 .
Pavlov, 1. P. (19 27). C o n d it io n e d r e flex e s: A n in v esti Rothb art, M. K., Ahadi, S. A., &c Evans, D. E. (20 00 ).
g a tio n o f th e p h y s io lo g ic a l a ctiv ity o f th e c e r e b r a l Temperament and personality: Origins and out
c o r te x . Oxford, UK: Oxford University Press. comes. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o l
Pavot, W., Diener, E., & Fujita, F. (1990). Extraversion og y , 78(1), 1 2 2 - 1 3 5 .
and happiness. P erson ality a n d In d iv id u a l D iffe r Ruch, W., & Hehl, F. J. (1989, June). P sy ch o m etric
en c e s, 11(12), 1 2 9 9 - 1 3 0 6 . p r o p e r tie s o f th e G e rm a n v ersio n o f the E P Q -R .
Pytlik Zillig, L. M ., Hemenover, S. H., & Dienstbier, Paper presented at the 4th annual meeting of the In
R. A. (20 02 ). W hat do we assess when we assess ternational Society for the Study of Individual Dif
a Big 5 trait?: A content analysis of the affective, ferences. Heidelberg, Germany.
behavioral and cognitive processes represented in Rusting, C. L. (1999). Interactive effects of personality
the Big 5 personality inventories. P ers o n a lity a n d and mood on emotion-congruent memory and judg
S o c ia l P sy c h o lo g y B u lletin , 2 8 (6 ), 8 4 7 - 8 5 8 . ment. J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y ,
Rauch, S. L., Milad, M . R., Orr, S. P., Quinn, B. T., 77(5), 1 0 7 3 - 1 0 8 6 .
Fischl, B., & Pitman, R. K. (2005). Orbitofrontal Schutte, N. S., Malouff, J. M ., Segrera, E., Wolf, A., &
thickness, retention of fear extinction, and extra Rodgers, L. (2 003). States reflecting the big five di
version. N e u r o R e p o r t, 16, 1 9 0 9 - 1 9 1 2 . mensions. P erso n a lity a n d In d iv id u a l D iffe r e n c e s ,
Revelle, W. (1995). Personality processes. A n n u al R e 3 4 (4 ), 5 9 1 - 6 0 3 .
v iew o f P sy c h o lo g y , 4 6 , 2 9 5 - 3 2 8 . Smillie, L. D. (20 08 ). What is reinforcement sensi
Revelle, W. (20 08 ). T he contribution of reinforcement tivity?: Neuroscience paradigms for approach -
sensitivity theory to personality theory. In P. J. avoidance processes in personality. E u r o p ea n J o u r
Corr (Ed.), T h e re in fo r c e m e n t sen sitiv ity th e o r y o f n a l o f P erson ality , 2 2 (5 ), 3 5 9 - 3 8 4 .
p e r s o n a lity (pp. 5 0 8 - 5 2 7 ) . Cambridge, UK: C a m Smillie, L. D., Pickering, A. D., & Jackson, C. J. (200 6).
bridge University Press. The new reinforcement sensitivity theory: Implica
3. E x tra v e rsio n 45
tions for personality measurement. P erson ality a n d analyses of positive and negative affect: Their re
S o c ia l P sy c h o lo g y R e v iew , 10(4), 3 2 0 - 3 3 5 . lation to health complaints, perceived stress, and
Spain, J . S., Eaton, L. G., & Funder, D. C. (200 0). Per daily activities. J o u r n a l o f P erson ality a n d S o c ia l
spectives on personality: The relative accuracy of P sy ch olog y , 5 4 (6 ), 1 0 2 0 - 1 0 3 0 .
self versus others for the prediction of emotion and Watson, D. ( 2 00 0). M o o d a n d tem p er a m en t. New
behavior. J o u r n a l o f P erson ality , 6 8(5), 8 3 7 - 8 6 7 . York: Guilford Press.
Stelmack, R. M ., & Stalikas, A. (1991). Galen and the Watson, D., & Clark, L. A. (1992). On traits and
humour theory of temperament. P erson ality a n d In temperament: General and specific factors of em o
d iv id u a l D iffe r e n c e s , 12(3), 2 5 5 - 2 6 3 . tional experience and their relation to the five-factor
Strelau, J. (1987). Fimotion as a key concept in temper model. J o u r n a l o f P erson ality , 6 0(2), 4 4 1 - 4 7 6 .
ament research. J o u r n a l o f R e se a rch in P erson ality , Watson, D., & Clark, L. A. (1997). Extraversion and
2 1(4), 5 1 0 - 5 2 8 . its positive emotional core. In R. Hogan, (.Jo hn son ,
Tellegen, A. (1982). B r ie f m a n u a l f o r th e D iffe r e n tia l & S. Briggs (Eds.), H a n d b o o k o f p e r s o n a lity p s y
P erson ality Q u estio n n a ir e. Minneapolis: Univer c h o lo g y (pp. 7 6 7 - 7 9 3 ). San Diego, CA: Academic
sity of Minnesota. Press.
Tellegen, A. (1985). Structures of mood and person Watson, D., Clark, L. A., McIntyre, C. W., & Hamak-
ality and their relevance to assessing anxiety, with er, S. (1992). Affect, personality, and social activ
an emphasis on self-report. In A. H. Turna & J. D. ity. J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y ,
Maser (Eds.), A n x iety a n d the a n x ie ty d iso rd er s 6 3(6), 1 0 1 1 - 1 0 2 5 .
(pp. 6 8 1 - 7 0 6 ) . Hillsdale, NJ: Erlbaum. Weiler, M. A. (1992). S en sitiv ity to a ffe c tiv e ly va-
Theophrastus. (1909). T h e c h a r a c te rs o f T h eo p h r a stu s le n c e d stim u li. Unpublished doctoral dissertation,
(R. C. Jebb, Trans. & J. E. Sandys, Ed.). London: Northwestern University.
Macmillan. Widiger, T. A. (2 005). Five factor model of personality
Tooby, J ., & Cosmides, L. (1990). On the universality disorder: Integrating science and practice. J o u r n a l
of human nature and the uniqueness of the individ o f R e se a rc h in P erson ality , 39(1), 6 7 - 8 3 .
ual: T he role of genetics and adaptation. J o u r n a l o f Wright, C. I., Williams, D., Feczko, E., Barrett, L. F.,
P erso n a lity , 58(1), 1 7 -6 7 . Dickerson, B. C., Schwartz, C. E., et al. (20 06 ).
Trull, T. J ., & Sher, K. J . (1994). Relationship between Neuroanatomical correlates of extraversion and
the five-factor model of personality and Axis 1 dis neuroticism. C e r e b r a l C o r te x , 26(12), 1 8 0 9 -1 8 1 9 .
orders in a nonclinical sample. J o u r n a l o f A b n o r Wundt, W., & Judd, C. H. (1897). O u tlin es o f p s y c h o l
m a l P sy ch olog y , 103(2), 3 5 0 - 3 6 0 . ogy. Oxford, UK: Engelmann.
Tupes, F.. C., & Christal, R. E. (1961). R ecu rren t p e r Yerkes, R., 8c Dodson, J. (1908). T he relation of
s o n a lity fa c to r s b a s e d on trait ratin gs (USAF ASI) strength of stimuli to rapidity of habit-formation.
Tech. Rep. No. 61-97). Lackland Air Force Base, lo u r n a l o f C o m p a ra tiv e N eu r o lo g y a n d P sy c h o l
T X : U.S. Air Force. og y , 18, 4 5 9 - 4 8 2 .
Uziel, L. ( 2 00 6). The extraverted and the neurotic Zelenski, J. M., &c Larsen, R. J. (2 002). Predicting
glasses are of different colors. P erson ality a n d In d i the future: How affect-related personality traits
v id u a l D iffe r e n c e s , 41(4), 7 4 5 - 7 5 4 . influence likelihood judgments of future events.
Wacker, J., Chavanon, M .-L ., & Stemmier, G. (200 6). P erso n a lity a n d S o c ia l P sy ch olog y B u lletin , 2 8 (7 ),
Investigating the dopaminergic basis of extraversion 1 0 0 0 - 1010.
in humans: A multilevel approach. J o u r n a l o f P er Zinbarg, R., & Revelle, W. (1989). Personality and
so n a lity a n d S o c ia l P sy ch olog y , 97(1), 171-187. conditioning: A test of four models. J o u r n a l o f P er
Watson, D. (1988). Intraindividual and interindividual so n a lity a n d S o c ia l P sy ch olog y , 5 7 (2 ), 3 0 1 - 3 1 4 .
CHAPTER 4
Agreeableness
W il l i a m G . G r a z i a n o
R e n e e M . T o b in
46
4. A greeableness 47
persons who like them. At a process level, on agreeableness began as a result of reliable
this principle is almost certainly true, but empirical regularities arising in descriptions
it is also true that some people are liked by of others and later in self-descriptions (Dig-
virtually everyone. Liking is both a personal man & Takem oto-Chock, 1981). Because of
and an interpersonal process. Presumably its bottom-up empirical origins, there were
people would avoid environments contain debates about its correlates and even a suit
ing persons they do not like. However, if we able label for this hypothetical construct.
looked closely at the mutuality of liking in Other labels used to describe the dimension
group of people in vivo, we would discover are tender-m indedness, friendly com pliance
a lower match than expected. Some people versus hostile n on com plian ce , likeability,
would have friends and interaction partners com m union , and even love versus hate.
with whom they seemed not to match. Labeling constructs has consequences.
These considerations apply directly to For example, the term com pliance has a
agreeableness. Persons high in agreeable process-based meaning in social psychology
ness are well liked and popular with their that often places it on a continuum of social
peers, in part because they project positiv- influence with internalization and identifica
ity onto others and make excuses for others’ tion (e.g., Petty & Wegener, 1999). That va
shortcomings (Graziano & Tobin, 2002). riety of compliance is considerably different
Persons high in agreeableness expect oth from the one used more casually in person
ers to be pleasant and likeable and appear ality to imply tendencies to follow rules and
to elicit such behavior from their partners. norms. Friendly compliance might imply a
This pattern is consistent with the reciproc general conforming personality, but there is
ity of attraction principle, but it suggests the no experimental or even correlational evi
need to look past the personality moderator dence that persons high in agreeableness are
approach. In particular, it points to the need more responsive to social influence per se.
for attention to social interdependence and Recently, Habashi and Wegener (2008)
to other social-cognitive processes underly manipulated the quality of arguments
ing interpersonal interaction. The Person x (strong vs. weak) in a study of persuasive
Situation approach is a step in that direc communication. They found that persons
tion (Graziano, Habashi, Sheese, & Tobin, low in agreeableness were less influenced
2007). That is, rather than treating agree by persuasive communications than their
ableness as a variable that merely raises or peers high in agreeableness, regardless of
lowers the level of situational effects, agree the quality of argument. For persons high in
ableness enters the stage as an equal partner. agreeableness, however, strong arguments
In some cases, the presence of persons at led to greater attitude change than weak ar
different levels of agreeableness can funda guments. Taken together, these data suggest
mentally alter the situations themselves. We that agreeableness is related to being respon
discuss this in the section on cooperation sive to others, including their communica
and competition. tions. Responsiveness may be a prerequisite
for social influence, but it is certainly not the
same construct as compliance.
H istorical Origins o f Agreeableness Another issue related to the labeling of
compliance is socially desirable responding.
From ancient times, writers have commented Virtually every positive self-report marker or
on the value of agreeableness in social rela statement for agreeableness is more socially
tionships (e.g., Aristotle’s Akrasia). In mod desirable than its supposed bipolar oppo
ern scientific research, agreeableness has a site. Warm and kind are more desirable than
curious history relative to many other recog cold and unkind. It is possible that agree
nized dimensions of personality. Unlike the ableness primarily indexes self-favoring bias
supertraits of extraversion and neuroticism, and social desirability rather than basic indi
agreeableness did not initially receive system vidual differences in social dispositions, but
atic empirical research because of deductive the data do not support this interpretation.
top-down theorizing about its link to biology First, agreeableness differences initially en
or to especially conspicuous social behaviors tered the scientific literature through regu
(Feigl, 1970). Instead, systematic research larities in observer ratings (e.g., Digman &
48 II. IN T E R P E R S O N A L D IS P O S IT IO N S
Takem oto-Chock, 1981). Observer ratings ableness? Perhaps persons high in agreeable
are not entirely invulnerable to social desir ness are enhanced versions of persons low
ability problems, but their bias problems will in agreeableness. Exactly what process do
be different from those affecting self-report. persons low in agreeableness lack that per
When observer ratings and self-report con sons high in agreeableness possess? (For a
verge despite differences in bias, as is the parallel discussion of the dimensionality of
case with agreeableness ratings, it suggests self-monitoring, see Graziano & Waschull,
some validity for both assessment methods. 1995, pp. 2 3 8 -2 4 2 ). This issue is discussed
Second, the empirical literature does not for the case of agreeableness in the sections
support a self-favoring-bias artifact expla on helping, on prejudice, and on the overall
nation. In three studies using observational, motivational model of agreeableness.
correlational, and experimental methods (N As we inch upward from the soil of obser
= 979), Graziano and Tobin (2002) found vation toward theory, another issue involves
that other dimensions of the Big Five (Con configurations of personality dimensions.
scientiousness, Neuroticism) had more sig In talking about agreeableness, it may not
nificant correlations with various indexes of be safe to ignore other aspects of personal
self-favoring bias (impression management, ity, whether they are correlated with agree
self-deception, all three self-monitoring fac ableness or not (e.g., Goldberg, 1999). Is it
tors) than Agreeableness. Some measures of reasonable to expect the same pattern of,
self-favoring bias (self-deception) were un say, aggressive behavior from persons high
related to agreeableness. A social relations in agreeableness who are also high in extra
analysis (Kenny, 1996) also found that, by a version as from persons high in agreeable
ratio of almost 4:1, the larger part of agree ness who are high in neuroticism (but not
ableness variance was in the perceiver effect, extraversion)? Research showed that retal
relative to the target effect. That is, the larg iatory aggression was related (inversely) to
er source of variance in rated agreeableness agreeableness (Gleason, Jensen-Campbell,
was due to perceivers’ attributing the quali & Richardson, 2004), but is there a con-
ties of agreeableness to targets. In another figural aspect to this story? Their rationale
study participants were randomly assigned was derived from a different interactive
to conditions in which they were told that it theoretical perspective, but Ode, Robinson,
was bad to be agreeable or good to be agree and Wilkowski (2008) presented data show
able or were given no instructions. Partici ing that at higher levels of agreeableness, the
pants actually increased their self-ratings of anger-neuroticism link was considerably re
agreeableness when they were told it was duced. In a set of studies, Ode and Robinson
a bad quality. If agreeableness is somehow (2008) found a similar moderating effect for
related to self-favoring bias, being seen in a agreeableness on the relation between neu
socially positive light is not a major part of roticism and depressive symptoms. Similarly,
it. Overall, these and other outcomes suggest in a resistance-to-temptation study, Jensen-
that agreeableness effects are probably not Campbell and Graziano (2005) showed
artifacts of self-favoring bias. that higher levels of conscientiousness could
Other measurement issues have implica partially compensate for lower levels of
tions for construct validity. One is the as agreeableness (and vice versa) in predicting
sumed dimensionality of the construct itself. cheating in adolescents. Interestingly, in all
High internal consistency and even coherent of these cases the substantive, focal concern
loadings in factor analyses do not guarantee was affect regulation. The configuration of
that one and only one dimension underlies personality patterns (vs. one personality di
phenotypic expressions of agreeableness. At mension at a time) is at the leading edge of
least conceptually, the pattern of behavior personality theory and measurement, gener
exhibited by persons low in agreeableness ally under the rubric of the Abridged Big Five
may require a set of variables completely Circumplex (AB5C), but it clearly suggests
different from those used to describe the be avenues for refinement of our understanding
haviors exhibited by persons high in agree of links among personality dimensions and
ableness. Operationally, is a person low their collective relation with behavior (De
in agreeableness merely a person who is a Raad, 2 0 0 0 ; De Raad, Hendriks, & Hofst-
deficient version of a person high in agree ee, 1994). For the purpose of this chapter,
4. A greeableness 49
however, we do not discuss AB5C issues un .15 (vs. sex and extraversion at .20). Among
less immediately relevant to the topic. The children, teachers rate girls as more agreeable
configuration issue is important in the dis than boys, but in self-reports from the same
cussion of our new motivational approach to children, we find little evidence that boys
agreeableness. and girls consistently differ. Agreeableness
is correlated more highly with psychological
femininity (but not psychological masculin
M easuring Agreeableness ity) than with biological sex per se. This is
consistent with the Spence and Helmreich
Agreeableness differences can be mea (1979) view that psychological femininity is
sured through observation by knowledge tied closely to interpersonal and expressive
able informants such as spouses (Costa & motives and interests in relationships (Len-
M cCrae, 1988), employment supervisors ney, 1991). We find no evidence that minor
(Hogan, Hogan, & Roberts, 1996), and ity children differ systematically from ma
teachers (e.g., Digman & Takem oto-Chock, jority children in agreeableness (Graziano et
1981). Agreeableness may even be manipu al., 1998; Hair & Graziano, 2003). Usually,
lated experimentally as an independent agreeableness is correlated negatively with
variable (Jensen-Campbell, Graziano, & neuroticism, in the range o f - .2 0 to - .3 0 .
West, 1995). However, the method most The process of searching for measurement
commonly used is self-report, and this can artifacts and correlates of agreeableness
be accomplished through several different could be endless apart from the focus pro
instruments. Goldberg (1992) offered a set vided by relevant theory (Feigl, 1970). Some
of adjective markers that can be used even guidance comes from work on personality
with children (Graziano, Jensen-Campbell, development. Agreeableness may be tied dis
Steele, & Hair, 1998). Goidberg and his col tinctively to systems of self-regulation, espe
leagues have translated their instruments cially as they apply to frustration regulation
into many different languages (see http:// in social relations (Jensen-Campbell & Gra
ipip.ori.org/newltem Translations.htm ). An ziano, 20 0 5 ; Jensen-Campbell & M alcolm,
other option is to use questionnaire-format 2007).
measures such as the Big Five Inventory Ahadi and Rothbart (1994) offer a de
(BFI; John & Srivastava, 1999), the Interna velopmental hypothesis linking an early-
tional Personality Item Pool (IPIP; Goldberg appearing temperamental process, effortful
et al., 20 0 6 ), or one of the versions of the control, to subsequent personality structure
N EO (Costa &C M cCrae, 1988). Overall, the in children, adolescents, and adults. They
measures show some differences, but more propose that effortful control is part of a
remarkable is their convergence. An indi common developmental system underly
vidual scoring high on agreeableness on the ing two of the major dimensions in the Big
Goldberg markers is likely to score high on Five structural model of personality, namely
the BFI as well (John & Srivastava, 1999). Agreeableness and Conscientiousness (Gra
Confidence that most Big Five instruments ziano, 1994; Graziano & Eisenberg, 1997).
can measure agreeableness is enhanced by Specifically, Rothbart and her colleagues
evidence that such measures converge posi (e.g., Rothbart & Bates, 2 0 0 6 ; Rothbart &
tively with corresponding individual differ Posner, 1985) proposed that effortful control
ences in prosocial motives and negatively modulates other temperament systems as the
with antisocial tendencies, suggesting that frontal cortex matures. Effortful control is
personality differences and motives are re related to early-appearing differences in the
lated systematically (e.g., Finch, Panter, & ability to sustain and shift attention and the
Caskie, 1999). Evidence for divergent validi ability to initiate and inhibit action volun
ty also supports the claim that agreeableness tarily (e.g., Kochanska, Murray, & Coy,
warrants research as a separate construct. 1997). Effortful control seems to be related
The simple correlation between agreeable to the ability to suppress a dominant behav
ness and sex varies from sample to sample ior to perform a subdominant response or
and across age groups. In self-report data even an opposing dominant response, as is
from college-age students, in our samples of commonly the case for agreeableness. Jensen-
30 0 or more, the correlation rarely exceeds Campbell and colleagues (2002) found that
50 II. I N T E R P E R S O N A L D IS P O S IT IO N S
both agreeableness and conscientiousness with others (Graziano & Eisenberg, 1997)
were associated with traditional assessments and if such motivation reflects underlying
of self-regulation (e.g., on Stroop and W is processes of control and regulation (Jensen-
consin Card Sorting tasks). Campbell et al., 2 0 0 2 ; Jensen-Campbell,
Knack, Waldrip, &C Campbell, 2 0 0 7 ; Tobin,
Graziano, Vanman, & Tassinary, 2000),
Agreeableness and then in each of these domains agreeableness
Interpersonal Behaviors would be an important moderator of behav
ioral output. In each case, outcomes pointed
Like most psychological constructs, agree toward the need for more complicated ways
ableness can be understood in terms of of conceptualizing each of the classes of in
thoughts, feelings, and behaviors that are terpersonal behaviors. Fourth, stronger ex
related to it (Shadish, Cook, & Campbell, perimental studies linking agreeableness to
2002). Overall, agreeableness seems to be interpersonal processes have entered the lit
positively related to adaptive social behav erature, enhancing the credibility of the sug
iors (e.g., conflict resolution, emotional gestive (but inherently weaker) correlational
responsiveness, helping behavior) and neg studies. These experimental studies receive
atively related to maladaptive social behav special attention.
iors (i.e., prejudice, stigmatization). In this
section we discuss the links between agree
Agreeableness and Conflict
ableness and four classes of interpersonal
behaviors: interpersonal conflict, interper Theoretically, agreeableness maps onto the
sonal cooperation, helping and prosocial major motivational system of communion,
behavior, and prejudice. First, we focus on or the desire for unity, intimacy, and solidar
these four classes of interpersonal behavior ity with others (Wiggins, 1991). Consistent
because each involves important elements with this theoretical link, Graziano, Jensen-
of social motivation and social cognition. If Campbell, and H air (1996) found that most
humans did not live communally and had no people tend to endorse negotiation and dis
prospect of continuing interaction with each engagement tactics in resolving conflict but
other, then each day would become a totally that the difference between persons low in
new blank slate, as in G roundhog Day, the agreeableness and persons high in agreeable
movie starring Bill Murray. There would be ness was maximal when destructive tactics
little utility to social cognition and efforts to (e.g., power assertion tactics such as physical
understand the motivation of others. With force) were at issue. In particular, persons
the prospect of future interaction and social low in agreeableness reported that destruc
exchange, any given act of bias, helping, or tive tactics were generally more acceptable
conflict has implications for future interac than did persons high in agreeableness. Fur
tion. These implications inspire researchers thermore, individuals high in agreeableness
to search for underlying motives. tended to perceive less conflict in their social
Second, each of these types of behaviors interactions, report more liking of interac
has been linked to processes of control and tion partners, and elicit less conflict from
regulation. It makes sense to look for the fin their partners. They are perceived by others
gerprints of agreeableness in behaviors con as displaying less tension in their interac
taining elements of control and regulation. tions relative to their peers.
Third, in keeping with Kurt Lewin’s general Building on these findings, Jensen-
theoretical perspectives, we assume that in Campbell and Graziano (2001) conducted
terpersonal behaviors share basic underly multimethod research, including a diary
ing processes, so classification into discrete study, examining agreeableness as a moder
classes serves only a temporary pragmatic ator of middle-school children’s conflict pat
function (see Graziano & Waschull, 1995). terns. Consistent with their previous work,
Nevertheless, each of these kinds of inter individuals high in agreeableness reported
personal behavior has a distinct behavioral engaging in more constructive conflict res
topography. Our starting position was that olution tactics in their day-to-day interac
if agreeableness is related to differences in tions relative to their peers. Thus, across age
motivation for maintaining positive relations ranges and methodologies, agreeableness is
4. A greeableness 51
linked to positive resolution of conflict, pre Studies 2 and 3 translated the findings from
sumably because of its underlying motiva this vignette study into laboratory-based ex
tion to get along with others. periments in which participants were given
opportunities to volunteer help to a person
in need. Using Batson’s Katie Banks para
Agreeableness and Cooperation
digm (Coke, Batson, & McDavis, 1978),
Agreeableness is related to behavior in com Graziano, Habashi, and colleagues found
petitive and cooperative situations. Graziano that individuals high in agreeableness of
and colleagues (1997) examined patterns of fered to help outgroup members (i.e., a stu
cooperative and competitive behaviors in tri dent from a different university) more often
ads of college students. Overall, agreeableness than did individuals low in agreeableness.
was related negatively to competitiveness in They also offered help more often than their
groups and related positively to expectations low-agreeable counterparts even when their
of harmonious group interactions. Graziano attention was directed experimentally to the
and colleagues also found that competitive technical aspects (rather than the emotional
ness mediates the relation between agree aspects) of the situation, demonstrating that
ableness and cooperation, indicating that in an other-focused, empathic response is more
dividuals low in agreeableness tend to view automatic in persons high in agreeableness
themselves as less interdependent with other relative to their peers. Results of Study 3
group members and to respond with more shed additional light on the relation between
competitive behaviors relative to their high- agreeableness and helping by demonstrating
agreeable peers. Similar relations were also that empathic concern, but not personal dis
found earlier in development with 115 triads tress, mediates this relation in the technical-
of school-age children (Tobin, Schneider, focus condition.
Graziano, & Pizzitola, 2002). In both age Graziano, Habashi, and colleagues (2007,
groups, persons high in agreeableness seem Study 4) extended this line of research by
to transform competitive situations into co experimentally manipulating not only the
operative ones. This transformation is easier focus of participants’ attention (emotional
to accomplish if other group members are vs. technical aspects) but also the cost of
also high in agreeableness. helping. They found that when the cost of
helping is high, asking participants low in
agreeableness to focus on the emotional as
Agreeableness and H elp in g
pects of the situation reduces their willing
Agreeableness also plays an important ness to help. When cost of helping is low,
role in the experience of positive emotions however, the opposite pattern was obtained
within the context of interpersonal relation for individuals low in agreeableness: Persons
ships. Among the five dimensions of per low on this personality dimension increased
sonality, agreeableness is the only one that their helping when the cost of helping was
is significantly correlated with both of the low and they were instructed to focus on
major aspects of prosocial emotions, name emotion. Thus, when costs are low, a re
ly empathic concern and personal distress. minder to pay attention to others’ emotions
Zero-order correlations between agreeable facilitates helping behavior in people low in
ness and measures of self-reported empathy agreeableness, whereas this same reminder
are consistently strong and positive (e.g., decreases helping behavior when greater
Graziano, Habashi, et al., 20 0 7 ). Beyond costs of helping are placed on people low in
self-reports, agreeableness has also been agreeableness. These findings indicate that
connected to prosocial behaviors, such as helping may be increased in persons who or
volunteering to help others in need. In the dinarily do not offer to do so when the costs
first of a set of studies, Graziano, Habashi, to them are low. In contrast, the request to
and colleagues (2007) found that individu focus on emotions as opposed to the techni
als high in agreeableness are more likely to cal aspects of the situation did not yield a
report willingness to help a wider range of similar reduction in helping for individuals
others than those low in agreeableness when high in agreeableness when the cost of help
presented with scenarios in which they may ing was either high or low. Taken together,
offer to help a sibling, a friend, or a stranger. the Graziano, Habashi, and colleagues stud
52 II. IN T E R P E R S O N A L D IS P O S IT IO N S
ies demonstrated important links among with an overweight woman. In Study 3, the
agreeableness, empathy, and helping behav authors replicated and extended these find
ior. Thus research indicates that the mo ings by demonstrating that these prejudicial
tives underlying agreeableness are related to responses translated into discriminatory be
greater experience of empathy and that this haviors. Participants were presented with a
emotional experience is, in turn, related to photograph of either a typical or overweight
increased willingness to help. partner who was ostensibly similar to the
participant in personality and were given
the opportunity to change partners without
Agreeableness and Prejudice
penalty. Graziano, Bruce, and colleagues
Agreeableness-related responsiveness mo found that only men low in agreeableness in
tives have also been linked to biases in reac dicated a desire to switch partners, and they
tions to others. Graziano, Bruce, Sheese, and only did so when paired with an overweight
Tobin (2007) investigated whether the mo female partner.
tives underlying agreeableness lead individu The remaining two studies by Graziano,
als to respond differentially to persons from Bruce, and colleagues (2007) focused on
a stigmatized group (e.g., persons who are identifying conditions under which persons
overweight). In a five-study program, these high in agreeableness may be more likely to
researchers first examined how agreeable exhibit the prejudicial responses shown by
ness was related to perceived social norms individuals low in agreeableness. In Study
and personal endorsements of prejudice to 4, providing a justification for expressing
ward over 100 potential targets of prejudice. prejudice (i.e., the partner expressed coun-
They found that individuals high in agree ternormative negative sentiments about their
ableness did not differ from their peers in university) yielded increased prejudicial re
terms of their understanding of the social sponding from individuals high in agree
norms related to the acceptability of hold ableness, but only when they were paired
ing prejudiced feelings toward these groups; with overweight female partners. Although
however, they did differ in their personal persons high in agreeableness increased
endorsement of such prejudice. That is, in their prejudicial responding in this condi
dividuals high in agreeableness reported less tion, overall, participants low in agreeable
negative reactions to most groups, including ness expressed stronger prejudicial reactions
traditional targets of prejudice (e.g., homo than their high-agreeable counterparts. A
sexuals, Jews, Hispanics) relative to their similar pattern of results was obtained in a
peers. Thus the findings support the hypoth fifth study, when participants were provided
esis that agreeableness is related to preju with an even greater justification for the ex
diced reactions, at least in terms of verbal pression of prejudice, namely that the osten
self-reports. sible interaction partner created additional
Moving beyond these self-reports, Gra work for the participant. Individuals high
ziano, Bruce, and colleagues (2007) used an in agreeableness expressed prejudice toward
established experimental paradigm (Snyder overweight partners relative to typical weight
& Haugen, 1994) to investigate prejudicial partners, but only when the partner was at
reactions to specific interaction partners. In fault for a mistake that led to additional
this study, participants were partnered with workload for the participants. In contrast,
an unknown female participant for a “get individuals low in agreeableness expressed
ting acquainted conversation.” Before the more negative reactions to their partners re
conversation, participants were provided gardless of the cause of additional work.
with a photograph of the supposed partner.
This photograph was digitally altered so
that the partner appeared either overweight Em otion al Processes
or of typical weight. Participants reported Underlying Agreeableness
their prejudicial reactions to their partners and Interpersonal Behavior
using a social-distance measure following
the conversations. Male participants low in Agreeableness may not be highly related to
agreeableness responded with the most prej other major structural dimensions of person
udicial reactions, but only when partnered ality, but it is probably related to other dis
4. A greeableness 53
Yeadon, & Hesson-M clnnis, 2 0 0 4 ) and on proach and avoidance. In the helping area,
helping (Batson, 1991; Dijker & Koomen, studies show that “messy victims” (such as
20 0 7 ), but the processes may be more gen those who are bleeding) seem to activate
eral than previously recognized. Third, fur avoidance that blocks helping (e.g., Pili-
ther clarification of the apparent anomalies avin, Callero, & Evanset, 1982). In Batson’s
in the two literatures can be obtained using (1991) empathy-altruism model, self-focused
a dual-process, sequential-opponent motiva personal distress seems to block helping, es
tional system that incorporates agreeable pecially when escape from the helping situa
ness. We describe such a model here. tion is relatively easy (Batson et al., 1981). In
Let us first consider some of these anoma the stigma literature, Pryor and colleagues
lies, curiosities, and striking parallels. Re (2004) showed that people often have an ini
search on the social psychology of prejudice tially negative reflexive reaction to outgroup
lives in almost total isolation from research members. Within 3 0 0 -5 0 0 milliseconds,
on helping and altruism. At first glance, however, corrective reflective processes can
this makes sense. After all, prejudice is a come online and suppress the avoidance.
negative, even antisocial behavior, whereas The nature of the Pryor paradigm makes it
helping is a positive, constructive, prosocial clear that both reflexive avoidance and re
activity. With closer inspection, however, flective approach are operative.
the separate-but-equal status is harder to From a theoretical perspective, apparent
justify. First, the small number of research anomalies within each of these two areas are
ers who are active in both areas (e.g., Dovi- more informative than are similarities. In
dio, 1984) have found a few processes that Batson’s empathy-altruism approach, noted
seem to affect both prejudice and helping. previously, the self-focused emotion of per
Perhaps the most conspicuous example is sonal distress undermines helping, whereas
the ingroup-outgroup status of the victim the victim-focused emotion of empathic
(e.g., Graziano, Habashi, et al., 2 0 0 7 ; Pili- concern promotes helping. This relation has
avin, Dovidio, Gaertner, & Clark, 1982). been demonstrated in experimental studies
Members of outgroups are often targets of that manipulate perspective taking. Tech
prejudice, and they are also less likely to re nically, empathy refers to a set of related
ceive help. Digging a bit deeper, we find half components that include personal distress,
hidden assumptions about processes that empathic concern, and perspective taking
apply to both areas. Prejudice and helping (Davis, 1996). The last of these three pro
are relationship phenomena, usually operat vides a distinctively cognitive process that is
ing at the initiation phase of interpersonal relatively easy to manipulate experimentally.
attraction, at least as investigated by social In the typical experiment using the Batson
psychologists (Graziano & Bruce, 2008). paradigm, operationally the affective pro
Even within more focused debates about cesses of empathic concern are elicited from
interpersonal processes (e.g., Byrne, 1997; research participants by manipulating their
Rosenbaum, 1986), the typical assumption focus of attention (e.g., Coke et al., 1978;
was that target persons were placed along a Toi & Batson, 1982).
single continuum of positive to negative af The apparent anomaly here is that vir
fect (Graziano & Bruce, 2008). tually all studies that have measured both
Agreeableness research may help us un personal distress and empathic concern find
derstand the links between prejudice and that they are correlated positively, not nega
prosocial behavior. In particular, findings tively (e.g., Batson, O ’Quin, Fultz, Vanderp-
involving agreeableness allow us to see con las, & Isen, 1983; Graziano, Habashi, et al.,
nections between these two sets of behav 2007). For present purposes, we note that
iors that are not obvious. We elaborate on Batson and colleagues (1983) attempted to
this point later. For now, we note that even address this problem by categorizing “pre
without agreeableness, we can see connec ponderant motives.” Because personal dis
tions between the two. In addition to a few tress and empathic concern are both present
common influential variables and processes, in participants (and are correlated posi
both literatures acknowledge the possibility tively), Batson and colleagues assigned par
that phenomena in the prejudice and help ticipants to conditions based on their single
ing literatures contain elements of both ap most dominant motive (see pp. 711-712).
4. A greeableness 55
Thus Batson and his colleagues recognize plicitly sets high standards for altruism and
the operation of two potentially opposing does not allow for the operation of multiple
motives linked to avoidance and approach. motives. Given these limitations, the basic
Here we expand on this conceptualization proposition receives only mixed empirical
by connecting these avoidance and ap support (Schroeder, Dovidio, Sibicky, M at
proach motives to individual differences in thews, & Allen, 1988). In some cases, the
agreeableness. ease of escape seems to matter little (e.g.,
Another set of curiosities involves over Eisenberg et al., 1989; Habashi, 2008). At
compensation. The prejudice literature the least, the inconsistency in the effective
shows that often (but not always), research ness of the easy-difficult escape manipula
participants may provide exceptionally larg tions is an anomaly. As we see later, differ
er offers of help, assistance, or other benefits ences in agreeableness help us confront this
to outgroup members than to ingroup mem apparent anomaly.
bers (Dijker & Koomen, 20 0 7 ). As noted Another set of curiosities is associated
previously, Pryor and colleagues (2004) with individual differences in motivation.
found evidence of deliberative compensa The quests for the “prejudiced personal
tion. If Pryor and colleagues are describing ity” (Allport, 1954; Graziano, Bruce, et al.,
general processes extending beyond preju 20 0 7) and the “altruistic personality” (Bat
dice and stigmatization, then overcompensa son, Bolen, Cross, & Neuringer-Benefiel,
tion may be the result of two interdependent 1986; Carlo, Eisenberg, Troyer, Switzer, &
motivation processes that occur in sequence. Speer, 1991; Graziano & Eisenberg, 1997;
In the specific cases studied so far, the re Penner, Fritzsche, Craiger, & Friefeld, 1995)
flective processes generally lead away from have long histories. That many individual
prejudice toward prosocial behavior. At differences moderate helping or prejudice is
least in theory, there is no prior reason to as no longer controversial (Dovidio, Piliavin,
sume that reflective processes per se lead to Schroeder, & Penner, 2 0 0 6 ; Penner, Dovid
prosocial action, but agreeableness provides io, Piliavin, & Schroeder, 2005). More con
us with a motivational structure to support troversial is the generality of the influence of
these observations. Because the literature any given individual difference. For exam
as a whole shows less liking for outgroup ple, Pryor and colleagues (2004) found that
members, overcompensation results suggest prejudiced reactions to an H IV victim were
that at least two motivational processes are moderated by individual differences on the
involved in the outcome. Two likely candi Heterosexual Attitudes toward Homosexu
dates are approach and avoidance, and these als scale (Larson, Reed, & Hoffman, 1980),
processes theoretically connect to agreeable but there was no evidence that this individu
ness. al difference moderated prejudice against ex
Another apparent anomaly comes from convicts. Part of the problem is theoretical.
the Batson empathy-altruism model. It may It is not clear precisely what mechanisms are
appear to be a tangent at this point, but its responsible for mediation. Ideally, a single
relevance is explained in the subsequent in motive or set of motives would be identified
tegration section. The research literature that cut across substantive topic areas and
shows that many (maybe even most) forms provide a unifying thread. As noted previ
of helping are motivated by self-interest ously, Graziano and colleagues (Graziano,
(Batson, 1991; Cialdini et al., 1987). W hat is Bruce, et al., 2007; Graziano, Habashi,
still controversial is the frequency of helping et al., 2007) found evidence that a shared
motivated solely for the benefit of the vic mechanism links personality, namely agree
tim. Batson (1991) argued that the motives ableness, with both helping and prejudice.
underlying most acts of helping are difficult A final curiosity involved empathy as an
to discern because helping can relieve the emotion. Some studies treat empathy as a
provider’s distress, as well as the victim’s. category of emotion (e.g., Batson, 1991),
One situation, however, allows a clearer de whereas other studies treat it as part of a
lineation of helper’s motives. If a potential dimension based on negative-to-positive
helper can readily escape but still chooses emotion (e.g., Piliavin et al., 1982). As noted
to help, then altruistic, other-oriented mo previously, empirical work suggests that em
tives are now more plausible. This logic im pathy as a construct has a complex structure
56 II. I N T E R P E R S O N A L D IS P O S IT IO N S
consisting of both affective and cognitive correct, then agreeableness may not only re
processing (Davis, 1996). Within this struc late to sympathetic caregiving to the weak
tural approach, affect is further differentiat and disadvantaged but may also operate to
ed into a self-focused element and an other- suppress the responses associated with the
focused element (Eisenberg et al., 1989). In more primitive fight-flight system. Put dif
some research, the cognitive component is ferently, some agreeableness phenomena
used to activate the affective components. In may be fairly direct expressions of care and
Batson’s theory, perspective taking enhances others may be a product of care-based sup
empathic concern but inhibits personal dis pression of fight-flight. Concretely, persons
tress. This makes intuitive sense, but it seems high in agreeableness may feel empathic
to assume that empathic concern and person concern directly for victims of misfortune
al distress are separate categories of emotion (Graziano, Habashi, et al., 20 0 7 ), but they
or are negatively correlated. Furthermore, may also suppress (perhaps effortfully) nega
even from a dimensional approach to emo tive reactions to traditional targets of preju
tion, what is the affective valence of emotion dice generated by their fight-flight systems
once it is aroused? Is it primarily negative or (Graziano, Bruce, et al., 2007).
positive (e.g., Davis, 1996)? Perhaps its ini Taking the system a bit further past de
tial activation is experienced as negative, but scription, let us assume some connections
if help can be provided, it becomes positive. between the fight-flight and care systems of
We discuss this subsequently. potential relevance to agreeableness. If we
One step toward integrating these diver assume that both fight-flight and care sys
gent issues and building a model of agree tems are present in almost all people (but
ableness may be found in work by Dijker at varying strengths) and that fight-flight
and Koomen (2007). They proposed an in occurs faster than care on exposure to an
novative, integrative approach to stigmati environmental oddity, the two may operate
zation that included two evolved, preverbal as opponents to each other’s preponderant
systems of motivation. Each of these two re responsive activation tendencies. If so, we
flects human evolutionary history. The older can generate explanations for apparent para
component is a fight-or-flight system that we doxes and anomalies. In the helping context,
carry as part of our paleoreptilian heritage. personal distress may inhibit prosocial acts
Encounters with “unusual cases” (“devi because it is part of fight-flight, not care.
ance” in Dijker & Koomen, 2 0 0 7 ) activate Empathic concern promotes helping because
this system without conscious deliberation, it is part of care. Despite having opposite ef
priming a system that impels individuals to fects on helping, both personal distress and
flee from danger or to fight if forced to do so. empathic concern are present in most peo
The second system is newer in evolutionary ple, explaining the positive correlation. Per
time and is part of the parental care system sonal distress is the first response to a victim
associated with kin selection (Hamilton, because it is connected to the faster fight-
1964; Trivers, 1972). Furthermore, the two flight system. If there is an opportunity for
motivational systems have the capacity to easy escape from the victim when personal
elicit characteristic emotions when exposed distress is high, then the victim will not re
to certain specific environmental triggers. ceive help. If escape cannot occur quickly,
Because humans evolved in small groups of or if the observer must remain in proxim
genetically related individuals, aggressive re ity to the victim, then enough time may
actions to unusual cases had to be inhibited. pass for the slower empathic-concern sys
Some of the unusual cases probably involved tem to become active. This would suppress
kin, for whom repair of deviance would be the fight-flight system and increase chances
more beneficial than aggression or exclusion. the victim would receive help. This account
The care system has the capacity to suppress would explain why outcomes of research on
the fight—flight system. ease-difficulty of escape are unstable. The
The theoretical system presented by Di key variable— the time interval between ex
jker and Koomen (2007) may be expanded posure to the victim and the window of op
further toward agreeableness. Let us as portunity for escape— is unmeasured.
sume that agreeableness is the psychological Going one step further, the system we de
manifestation of the care system. If this is scribe may be a case of the opponent-process
4. A greeableness 57
Pure A
{Personal Distress)
PureB
* (Empathic Concern}
F IG U R E 4.1. O ppon en t-process model of m otivation. From Solom on and C orbit (1974). C opyright
1974 by the A m erican Psychological A ssociation . A dapted by perm ission.
58 II. I N T E R P E R S O N A L D IS P O S IT IO N S
and the initiation of relationships: A review of the Two faces of temptation: Differing motives for self-
empirical literature. In S. Sprecher, A. Wenzel, & J. control. M e r r ill-P a lm e r Q u arterly , 51, 2 8 7 - 3 2 4 .
Harvey (Eds.), H a n d b o o k o f re la tio n sh ip in itiation Jensen-Campbell, L. A., Graziano, W. G., 8c West, S.
(pp. 2 6 9 - 2 9 5 ) . New York: Psychology Press. G. (1995). Dominance, prosocial orientation, and
Graziano, W. G., Bruce, J . W., Sheese, B. E., 8c Tobin, female preferences: Do nice guys really finish last?
R. M . ( 2 0 07 ). Attraction, personality, and preju J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 6 8 ,
dice: Liking none of the people most of the time. 427-440.
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 93, Jensen-Campbell, L. A., Knack, J. M ., Waldrip, A. M.,
565-582. 8c Campbell, S. D. ( 2 00 7). D o Big Five personal
Graziano, W. G., 8c Eisenberg, N. (1997). Agreeable ity traits associated with self-control influence the
ness: A dimension of personality. In R. Hogan, J. regulation of anger and aggression? J o u r n a l o f R e
Joh nson, 8c S. Briggs (Eds.), H a n d b o o k o f p e r s o n a l s e a r ch in P erson ality , 41(2), 4 0 3 - 4 2 4 .
ity p s y c h o lo g y (pp. 7 9 5 - 8 2 4 ) . San Diego, CA: Aca Jensen-Campbell, L. A., 8c Malco lm , K. T. (200 7).
demic Press. The importance of conscientiousness in adolescent
Graziano, W. G., Habashi, M . M., Sheese, B. E., 8c interpersonal relationships. P erson ality a n d S o c ia l
Tobin, R. M . (20 07 ). Agreeableness, empathy, and P sy ch o lo g y B u lletin , 33(3), 3 6 8 - 3 8 3 .
helping: A person x situation perspective. J o u r n a l Jensen-Campbell, L. A., Rosselli, M ., Workman, K.
o f P ers o n a lity a n d S o c ia l P sy ch olog y , 93, 5 8 3 - 5 9 9 . A., Santisi, M ., Rios, J . D., 8c Bojan, D. (2002).
Graziano, W. G., Hair, E. C., 8c Finch, J . F. (1997). Agreeableness, conscientiousness and effortful con
Competitiveness mediates the link between person trol processes. J o u r n a l o f R e se a rc h in P erson ality ,
ality and group performance. J o u r n a l o f P ers o n a l 36, 476-489.
ity a n d S o c ia l P sy ch olog y , 73, 1 3 9 4 - 1 4 0 8 . John, O. P., 8c Srivastava, S. (1999). T he Big Five trait
Graziano, W. G., Jensen-Campbell, L. A., 8c Hair, E. taxonomy: History, measurement, and theoretical
C. (1996). Perceiving interpersonal conflict and re perspectives. In L. A. Pervin 8c O. P. John (Eds.),
acting to it: The case for Agreeableness. J o u r n a l o f H a n d b o o k o f p e r s o n a lity : T h e o r y a n d re se a r ch
P erson ality a n d S o c ia l P sy ch olog y , 70, 8 2 0 - 8 3 5 . (2nd ed., pp. 1 0 2 -1 3 8 ). New York: Guilford Press.
Graziano, W. G., Jensen-Campbell, L. A., Steele, R. Kelley, H. H „ Holmes, J. G., Kerr, N. L., Reis, H. T„
G., 8c Hair, E. C. (1998). Unknown words in self- Rusbult, C. E., 8c Van Lange, P. A. M. (2003). An
reported personality: Lethargic and provincial in a t la s o f in te r p e r s o n a l situ ation s. New York: C a m
Texas. P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , bridge University Press.
24, 893-905. Kenny, D. A. (1996). The design and analysis of social-
Graziano, W. G., 8c Tobin, R. M . (2 002). Agreeable interaction research. A n n u al R e v ie w o f P sy ch olog y ,
ness: Dimension of personality or social desirability 47, 5 9 - 8 6 .
artifact? J o u r n a l o f P erson ality , 70, 6 9 5 - 7 2 7 . Kochanska, G., Murray, K., & Coy, K. C. (1997). In
Graziano, W. G., 8c Waschull, S. B. (1995). Social hibitory control as a contributor to conscience in
development and self-monitoring. In N. Eisenberg childhood: From toddler to early school age. C h ild
(Ed.), S o c ia l d e v e lo p m e n t (pp. 2 3 3 - 2 6 0 ) . Thousand D e v e lo p m e n t, 6 8 , 2 6 3 - 2 7 7 .
Oaks, CA: Sage. K ohnstamm, G. A., Halverson, C. F., Jr., Mervielde,
Haas, B. W., Omura, K., Constable, R. T., 8c Canli, T. L, 8c Havill, V. L. (1998). Analyzing parental free
(200 7). Is automatic emotion regulation associated description of child personality. In G. A. Koh n
with agreeableness? A perspective using a social stamm, C. F. Halverson, Jr., I. Mervielde, 8c V. L.
neuroscience approach. P sy c h o lo g ic a l S cien ce, 18, Havill (Eds.), P a ren ta l d es c rip tio n o f ch ild p e r s o n
130-132. a lity : D e v e lo p m e n ta l a n te c e d e n ts o f th e B ig F iv e?
Habashi, M . M. (20 08 ). S ep a ra tin g a ltru istic a n d e g o (pp. 1-19). Mahwah, NJ: Erlbaum.
istic m o tiv es f o r h elp in g : A n in d iv id u a l d iffe r e n c e Krueger, R. F., Hicks, B. N., 8c McGue, M . (2001).
a p p r o a c h . Unpublished doctoral dissertation, Pur Altruism and antisocial behavior: Independent ten
due University. dencies, unique personality correlates, distinct eti
Habashi, M . M ., 8c Wegener, D. (2008). P relim in ary ologies. P sy c h o lo g ic a l S cien ce, 12, 3 9 7 - 4 0 2 .
e v id e n c e th at a g r e e a b le n e s s is m o r e clo s ely r e la ted Lang, P. J., Bradley, M . M ., 8c Cuthbert, B. N. (1995).
to resp o n s iv en e ss th an co n fo r m ity . Unpublished I n te r n a tio n a l A ffec tiv e P ictu re S ystem (IA P S ):
manuscript, Purdue University. T ec h n ic a l m a n u a l a n d a ffe c t iv e ratings. Gainesville:
Hair, E. C., 8c Graziano, W. G. (2003). Self-esteem, University of Florida, Center for Research in Psy
personality, and achievement in high school: A pro chophysiology.
spective longitudinal study in Texas. J o u r n a l o f P er Larson, K. S., Reed, M ., 8c Ho ffm an, S. (1980). At
so n a lity , 71, 9 7 1 - 9 9 4 . titudes of heterosexuals toward homosexuals: A
Hamilton, W. D. (1964). The genetical evolution of so Likert-type scale and construct validity. J o u r n a l o f
cial behaviour: 1 , 11. J o u r n a l o f T h e o r e t ic a l B io lo g y , S ex R e se a rc h , 16, 2 4 5 - 2 5 7 .
7, 1 - 3 2 . Lenney, E. (1991). Sex roles: The measurement of
Hogan, R., Hogan, J ., 8c Roberts, B. W. (1996). Per masculinity, femininity and androgyny. In J . P.
sonality measurement and employment decisions: Robinson, P. R. Shaver, 8c L. S. Wrightsman (Eds.),
Questions and answers. A m er ica n P sy ch olog ist, M easu res o f p e r s o n a lity a n d s o c ia l p s y c h o lo g ic a l
51(5), 4 6 9 - 4 7 7 . a ttitu d es : Vol. 1. M easu res o f s o c ia l p s y c h o lo g ic a l
Jensen-Campbell, L. A., 8c Graziano, W. G. (2001). attitu d es (pp. 5 7 3 - 6 6 0 ) . San Diego, CA: Academic
Agreeableness as a moderator of interpersonal con Press.
flict. J o u r n a l o f P erson ality , 69, 3 2 3 - 3 6 1 . Ode, S., 8c Robinson, M. D. (20 0 8). C an a g r e e a b le
Jensen-Campbell, L. A., 8c Graziano, W. G. (2005). n ess turn g ra y s k ie s b lu e f: A r o le fo r a g r e e a b le
4. A greeableness 61
Attachment Styles
P h il l i p R . S h a v e r
M a r io M ik u l in c e r
ttachment theory (Ainsworth, Ble- 2. One of these behavioral systems, the at
A har, Waters, & W all, 1978; Bowlby,
1969/1982) was initially proposed as a way
tachment system, is responsible for estab
lishing primary social connections and
of understanding why close relationships in calling on them in times of stress or dif
the family and the loss of such relationships ficulty.
are among the most important determi 3. The history of a person’s close relation
nants of later social adjustment and mental ships shapes the parameters of his or her
health. The originator of the theory, John attachment system, leaving an important
Bowlby, was a British psychoanalyst with an residue in the form of “internal working
unusual interest in ethology and cognitive models” of self, partners, and relation
and developmental psychology. He was for ships. This developmental process results
tunate to form a working relationship with in each person having a measurable “at
an American developmental psychologist, tachment style” (Hazan & Shaver, 1987)
Mary Ainsworth, who added psychometric that influences the nature and outcomes
and research skills to Bowlby’s astute clini of subsequent relationships, including
cal observations and exceptional ability to those with romantic/sexual partners,
integrate diverse scientific literatures in the close friends, offspring, and even co
service of what, by today’s standards, is a workers and subordinates in social orga
“grand theory.” nizations (e.g., Davidovitz, Mikulincer,
The key components of the theory are few, Shaver, Ijzak, & Popper, 2007).
and they are relatively easy to describe:
In this chapter we describe the theory in
1. Humans and other primates evolved be more detail, explain how its key constructs
havioral and motivational systems that are measured in studies of adolescents and
allow them to survive and reproduce, de adults, and provide a brief summary of re
spite vulnerabilities associated with being search findings. A much more detailed ac
born prematurely, taking a long time to de count of the theory and the research it has
velop to maturity, and needing the protec generated can be found in our book, A ttach
tion, assistance, and cooperation of other m ent in A dulthood (Mikulincer & Shaver,
species members across the lifespan. 2007a).
62
5. A tta ch m e n t Styles 63
A ttachm ent T h eory: Basic Concepts the prolonged dependence of human infants
on “stronger and wiser” others (often, but
T h e A ttachm ent Behavioral System
not always, parents), who could defend them
In A ttachm ent and L oss — one of the from predators and other dangers. Because
most cited series of books in contempo human (and other primate) infants seem
rary psychology— Bowlby (1973, 1980, naturally to look for and gravitate toward
1969/1982) attempted to map and under particular others (those who are familiar
stand the profound impact that the quality and, at least sometimes, helpful) and to pre
of early relationships with primary caregiv fer them over alternative caregivers, Bowlby
ers has on personality development and in used the terms affection al bon d and attach
dividual differences in social behavior across ment, which is the reason for calling his for
the lifespan. As a psychoanalyst, Bowlby was mulation attachm ent theory. Although the
well aware that Freud and his followers had attachment system is most important and
already explored this issue, but he was also most visible in behavior during the early
aware that his fellow psychoanalysts had not years of life, Bowlby (1988) claimed that it
effectively integrated their work and their in is active across the lifespan and is frequently
terpretive approach to human problems with manifested in seeking support and love from
the rest of scientific psychology and psychia close relationship partners. This inspired
try. By considering a vast array of empirical various researchers (e.g., M ain, Kaplan, &
and theoretical writings ranging from clinical Cassidy, 1985; Shaver, Hazan, & Bradshaw,
observations of infants deprived of maternal 1988) to extend the theory into the domain
care to primate ethology and Piaget’s theory of adult relationships.
of cognitive development, Bowlby came to The purported goal of the attachment
the conclusion that a person’s fundamental system is to maintain a sense of safety or
sense of safety, self-worth, coping efficacy, security (called “felt security” by Sroufe
and well-being rests on the quality of his or & Waters, 1977). In Bowlby’s (1969/1982)
her social interactions with close relationship view, the attachment behavioral system is
partners, beginning with primary caregivers particularly activated by events that threat
in infancy. He also concluded that when a en the sense of security, such as encounters
person does not have reliable, trustworthy, with actual or symbolic threats or noticing
supportive relationships with close others, that an attachment figure is not sufficiently
personality development is distorted in ways near, interested, or responsive. In such cases,
that have serious negative consequences. a person is automatically motivated to seek
In explaining the motivational processes and reestablish actual or symbolic proximity
involved in personality development, which to an attachment figure (a process Bowlby,
Freud attempted to do using concepts such as 1969/1982, called the attachment system’s
sexual and aggressive “drives” or “instincts,” “primary strategy”). These bids for prox
Bowlby (1969/1982) borrowed from primate imity persist until protection and security
ethology the concept of behavioral systems — are attained. The attachment system is then
species-universal, biologically evolved neu deactivated, and the person can calmly and
ral programs that organize behavior in ways coherently return to other activities, which
that increase the likelihood of survival and Bowlby considered to be under the control
reproduction. He portrayed these systems as of other behavioral systems (e.g., explora
similar to cybernetic control systems, which tion, affiliation, caregiving). In infancy,
do not follow drive principles. According to attachment-system activation includes non
Bowlby, one of the key behavioral systems verbal expressions of distress, need, and de
is the attachm ent system , which has the sire for proximity (e.g., crying, calling) and
biological function of protecting a person locomotor behaviors aimed at reestablishing
(especially during infancy and early child and maintaining proximity (Ainsworth et
hood) from danger by ensuring that he or al., 1978). In adulthood, the primary attach
she maintains proximity to caring and sup ment strategy does not necessarily require
portive others (whom Bowlby, 1969/1982, actual proximity-seeking behavior, although
called attachm ent figures). In Bowlby’s view, often such behavior is initiated; it can also
the need to seek out and maintain proximity involve the internal activation of comforting
to attachment figures evolved in relation to mental representations of relationship part
64 II. I N T E R P E R S O N A L D IS P O S IT IO N S
ners who regularly provide care and protec to be available and supportive; I will experi
tion (Mikulincer & Shaver, 2004). These ence relief and comfort as a result of proxim
cognitive representations can create a sense ity to this person; I can then return to other
of safety and security and help a person deal activities.”
successfully with threats. When an attachment figure is not physi
cally or emotionally available in times of
need, not responsive to one’s bids for prox
Individual D ifferences
imity, or poor at alleviating distress or pro
in A ttachm ent-System Strategies
viding a secure base, attachment-system
Although nearly all children are born with functioning is disrupted, and the individual
normal attachment systems, which moti does not experience com fort, relief, or felt
vate them to pursue proximity and security security. Rather, the distress that initially
from an attachment figure in times of need, activated the system is compounded by seri
the quality of attachment-system function ous doubts about the feasibility of attaining
ing also depends on the availability of such a sense of security: “Is the world a safe place
a figure in times of need; his or her sensi or not? Can I trust others in times of need?
tivity and responsiveness to bids for close Do I have the resources necessary to manage
ness, com fort, and support; and his or her my own negative emotions?” These worries
ability and willingness to alleviate distress about self and others can maintain the at
and provide a secure base from which the tachment system in a continually activated
child can return calmly to other activities state, keep a person’s mind preoccupied with
(Bowlby, 1973, 1988). As Cassidy (1999) threats and the need for protection, and in
noted, “Whereas nearly all children become terfere drastically with other activities.
attached (even to mothers who abuse them; Frustrating interactions with inadequately
Bowlby, 1956), not all are securely attached” available or unresponsive attachment figures
(p. 7). According to attachment theory, the indicate that the attachment system’s oper
quality of interactions with attachment fig ating parameters need to be adjusted. This
ures in times of need is the major cause of implies that certain secondary attachm ent
individual differences in attachment-system strategies need to be adopted rather than
functioning. (There may also be genetic continuing to rely only on the primary strat
causes, as shown recently by Crawford et egy, confident proximity seeking. Attach
al., 2007, and Donnellan, Burt, Levendosky, ment theorists (e.g., Cassidy & Kobak, 1988;
& Klump, 2 0 0 8 , a possibility that was men M ain, 1990) have emphasized two such sec
tioned early on by Bowlby, 1969/1982.) ondary strategies: hyperactivation and d ea c
When an attachment figure is available, tivation of the attachment system. Hyperac-
sensitive, and responsive to an individual’s tivating strategies emerge from interactions
proximity bids, the individual is likely to feel with attachment figures who are sometimes
an inner sense of security— a sense that the responsive but only unreliably so, placing the
world is a safe place, that others are help attached person on a partial reinforcement
ful when called on, and that it is possible to schedule that seems to reward energetic,
explore the environment curiously and con strident, noisy proximity bids because they
fidently and engage rewardingly with other sometimes seem to succeed. In such cases,
people. This sense is an inner signal that people do not easily give up on proximity
the attachment system is functioning well seeking, and in fact they intensify it as a way
and that proximity seeking is an effective to demand the attachment figure’s love and
emotion-regulatory strategy. Moreover, peo support. The main goal of these strategies is
ple acquire important procedural knowledge to get an attachment figure, viewed as unre
about distress management, which becomes liable or insufficiently available and respon
organized around a relational script (Wa sive, to pay attention and provide protection
ters, Rodrigues, & Ridgeway, 1998; Waters or support. The chosen way to pursue this
& Waters, 2 0 0 6 ). The secure form of this goal is to maintain the attachment system in
script includes the following if-then propo a chronically activated state. This involves
sitions: “If I encounter an obstacle and/or exaggerating appraisals of danger and signs
become distressed, then I can approach a of attachment-figure unavailability and in
significant other for help; he or she is likely tensifying demands for attention, affection,
5. A tta ch m e n t Styles 65
and assistance. When repeatedly practiced, m odels o f others) and representations of the
this secondary strategy becomes what we self’s lovability and competence (working
call an anxious attachment style. m odels o f self). He argued that “if an indi
Deactivating strategies are another reac vidual is to draw up a plan to achieve a set-
tion to an attachment figure’s unavailabil goal not only does he have to have some sort
ity, and they seem to arise in conjunction of working model of his environment, but he
with attachment figures who disapprove must have also some working knowledge of
of and punish closeness and expressions his own behavioral skills and potentialities”
of need or vulnerability. In such relation (1969/1982, p. 112). Thus the attachment
ships, an individual learns to expect better system, once it has been activated repeatedly
outcomes if signs of need and vulnerability during interactions with a specific attach
are hidden or suppressed, proximity bids ment figure, includes representations of the
are weakened or blocked, the attachment availability, responsiveness, and sensitivity
system is deactivated despite a sense of se of such a figure, as well as representations of
curity not being achieved, and attempts are the self’s own capabilities for mobilizing the
made to handle threats by oneself (a strategy attachment figure’s support and one’s feel
Bowlby, 1969/1982, called “compulsive self- ings of being loved and valued by this figure.
reliance”). The primary goal of deactivating Because working models, at least initially,
strategies is to keep the attachment system are based on the internalization of specific
down-regulated to avoid the distress caused interactions with a particular attachment
by an attachment figure’s unavailability or figure, a person can hold multiple working
rejection. This deactivation requires denying models that differ in the outcome of the inter
attachment needs, steering clear of closeness action (success or failure to attain security)
and interdependence in relationships, and and the strategy used to deal with the distress
distancing oneself from threats that might caused by attachment-figure unavailability
cause unwanted activation of the attachment (hyperactivating or deactivating, anxious or
system. avoidant). Like other cognitive representa
tions, these working models form excitatory
and inhibitory associations with each other
A ttachm ent W orking Models
(e.g., experiencing or thinking about a secu
Beyond characterizing individual differences rity attainment activates memories of con
in attachment-system functioning during in gruent episodes of successful proximity bids
teractions with attachment figures, Bowlby and renders memories of attachment-figure
(1973) also proposed that such interactions unavailability less accessible), and these as
can be incorporated into mental structures sociations favor the formation of more ab
that eventually become relatively stable per stract and generalized representations of a
sonality patterns. At the core of these mental relationship with a specific partner. Thus
structures are what Bowlby called internal models with a specific attachment figure
w orking m odels. The term w orking has two (relationship-specific models) are created,
meanings in attachment theory. One is that and through excitatory and inhibitory links
the models are not static representations but with models representing interactions with
rather are the basis of social expectations, other attachment figures, even more generic
inferences about the likely outcomes of alter working models are formed to summarize
native social behaviors, and behavioral pro different kinds of relationships. The result of
grams that can be enacted in relationships. this process can be conceptualized as a hi
The other meaning of w orking is that the erarchical associative memory network that
models are based on past experiences and includes episodic memories, relationship-
can be revised based on new experiences. specific models, and generic models of se
This characteristic is what makes personality curity attainment, hyperactivation, and
change and successful relationship-oriented deactivation. As a result, with respect to a
psychotherapy possible (Bowlby, 1988). particular relationship and across different
Bowlby thought that interactions with at relationships, most people can sometimes
tachment figures were stored in at least two think about interpersonal relations in secure
kinds of working models: representations terms and at other times think about them
of attachment figures’ responses (working in less secure terms. In a 2003 paper, Over
66 II. IN T E R P E R S O N A L D IS P O S IT IO N S
all, Fletcher, and Friesen provided empirical formation of relatively stable individual dif
support for this hierarchical structure of at ferences in the operating parameters of the
tachment working models. attachment system. These stable and general
Each working model within the hierarchi ized individual differences can be empirical
cal network differs in cognitive accessibili ly examined by measuring a construct called
ty— that is, in the ease with which it can be attachm ent style — a person’s characteristic
activated and used to guide the functioning pattern of expectations, needs, emotions,
of the attachment system in a given social and behavior in social interactions and close
interaction. As with other cognitive repre relationships (Flazan & Shaver, 1987). De
sentations, the strength or accessibility of pending on how it is measured, attachment
each model is determined by the amount of style characterizes the way people behave
experience on which it is based, the number in a particular relationship (relationship-
of times it has been applied in the past, and specific style) or across relationships (global
the density of its connections with other attachment style).
working models (e.g., Baldwin, 1992; Shav The concept of attachment style, although
er et al., 1996). At a relationship-specific not given that name, was first proposed by
level, the model representing the typical in Ainsworth (1967) to describe infants’ pat
teraction with an attachment figure has the terns of responses to separations from and
highest accessibility in subsequent interac reunions with their mothers in a laboratory
tions with that person. At a generic level, “Strange Situation” assessment procedure.
the model that represents interactions with Based on this procedure, infants were classi
major attachment figures (e.g., parents and fied into one of three style categories: secure,
romantic partners) becomes the most chron anxious, or avoidant. M ain and Solomon
ically accessible working model and has the (1990) later added a fourth category, “dis
strongest effect on attachment-system func organized/disoriented,” which included odd,
tioning across relationships and over time. awkward behavior and unusual fluctuations
Consolidation of a chronically acces between anxiety and avoidance. Ainsworth
sible working model is the most important and colleagues (1978) noticed that the dif
psychological process accounting for the ferent infant attachment patterns can be ar
enduring, long-term effects on personality rayed in a two-dimensional space based on
functioning of attachment-relevant interac the anxiety and avoidance dimensions. This
tions during infancy, childhood, and adoles possibility has also been pursued in subse
cence (Bowlby, 1973; Fraley, 2 0 0 2 ; Waters, quent studies of romantic and global attach
M errick, Treboux, Crowell, & Albersheim, ment styles.
2000 ). Given a fairly consistent pattern of Infants classified as secure seem to pos
interactions with primary caregivers during sess chronically accessible working models
childhood, the most representative or proto of secure attachment, and their pattern of
typical working models of these interactions responses to separation and reunion reflects
become part of a person’s implicit proce a stable sense of attachment security. Spe
dural knowledge, tend to operate automati cifically, they react to separation from their
cally and unconsciously, and are resistant to mothers with overt expressions of distress
change. Thus what began as representations but then recover quickly and continue to ex
of specific interactions with a primary care plore the environment with interest. When
giver during childhood become core person reunited with their mothers, they greet them
ality characteristics; they tend to be applied with joy and affection, respond positively to
in new situations and relationships, and they being held, and initiate contact with them
shape the functioning of the attachment sys (Ainsworth et al., 1978). Avoidant infants
tem in adulthood. seem to hold chronically accessible work
ing models of unsuccessful proximity bids
organized around attachment-system de
T h e Concept o f A ttachm ent Style
activation. During separation and reunion
According to attachment theory (Bowlby, episodes, they show little distress when
1988; M ikulincer & Shaver, 2007a), a par separated from their mothers and seem to
ticular history of attachment experiences actively avoid them on reunion (Ainsworth
and the resulting consolidation of chroni et al., 1978). Anxious infants also seem to
cally accessible working models lead to the hold chronically accessible working models
5. A tta ch m e n t Styles 67
of frustrated proximity bids, but these mod to discuss, are associated with having a child
els seem to be organized around attachment- with the new, fourth attachment pattern,
system hyperactivation. These infants show “disorganized/disoriented” (Main & Solo
overt expressions of distress and despair mon, 1990), which in turn is most strongly
during separation episodes and conflictual, related to later psychopathology. These is
angry responses toward their mothers on sues are among the most actively studied by
reunion (Ainsworth et al., 1978). The two clinically oriented attachment researchers
different insecure patterns can be viewed as because of their applied significance.
defensive styles, one based on attempting to Despite the great value of the AAI as a
shut down or deactivate the attachment sys method of studying adults’ attachment pat
tem in order to avoid punishment or frustra terns, the interview is difficult to administer
tion and the other based on attempting to and score, and it focuses almost exclusively
escalate the expression of negative emotion on an adult’s early relationships with parents.
until a more security-enhancing response Taking a different path into the domain of
from an attachment figure is attained. adult attachment, Hazan and Shaver (1987;
In the 1980s, researchers from different Shaver et al., 1988) applied Bowlby’s ideas to
psychological fields (developmental, clini the study of romantic relationships. Because
cal, personality, and social psychology) con they developed their ideas within the frame
structed new measures of attachment style in work of personality-social psychology, they
order to extend attachment research into ad began with a simple self-report measure of
olescence and adulthood. Based on a devel adult attachment style. This measure con
opmental and clinical approach, Main and sisted of three brief descriptions of feelings
her colleagues (George, Kaplan, & M ain, and behaviors in romantic relationships that
1985; M ain et al., 1985; see Hesse, 2 0 0 8 , were intended to be adult analogues of the
for a review) devised the Adult Attachment three infant attachment styles identified by
Interview (AAI) to study adolescents’ and Ainsworth et al. (1978). Participants were
adults’ mental representations of attachment asked to read the three descriptions and then
to their parents during childhood. In the place themselves into one of the three at
AAI, interviewees answer open-ended ques tachment categories according to their pre
tions about their childhood relationships dominant feelings and behavior in romantic
with parents and are classified into three relationships. The three descriptions were:
categories paralleling Ainsworth’s infant ty
pology: “secure” (or free and autonomous Secure : I find it relatively easy to get close
with respect to attachment), “dismissing” (of to others and am comfortable depending
attachment), or “preoccupied” (with attach on them and having them depend on me.
ment). A person is classified as secure if he or I don’t worry about being abandoned or
she describes parents as available and respon about someone getting too close to me.
sive and if his or her memories of relation A voidant : I am somewhat uncomfortable
ships with parents are presented in a clear, being close to others; I find it difficult to
convincing, and coherent manner. Dismiss trust them completely, difficult to allow
ing individuals play down the importance of myself to depend on them. I am nervous
attachment relationships and tend to recall when anyone gets too close, and often
few concrete episodes of emotional interac others want me to be more intimate than I
tions with their parents. Preoccupied people feel comfortable being.
are entangled in worries and angry feelings Anxious: I find that others are reluctant to
about parents and, although they can easily get as close as I would like. I often worry
retrieve negative memories, they have trouble that my partner doesn’t really love me or
discussing them coherently without becom won’t want to stay with me. I want to get
ing overwhelmed and disorganized by anger very close to my partner, and this some
or anxiety. In recent years, new categories times scares people away.
have been added to the AAI coding system,
because some adults seem either to be unre Hazan and Shaver’s (1987, 1990) initial
solved with respect to traumas or losses or studies were followed by hundreds of others
to be unclassifiable into any of the major at that used the simple forced-choice self-report
tachment categories. These patterns, which measure to examine the interpersonal and
would take us beyond our space limitations intrapersonal correlates of adult attachment
68 II. IN T E R P E R S O N A L D IS P O S IT IO N S
style (see Mikulincer & Shaver, 2007a, for a expectations of partner availability and re
review). Over time, attachment researchers sponsiveness, comfort with closeness and
made methodological and conceptual im interdependence, and constructive ways of
provements to the original self-report mea coping with threats and stressors. People
sure, improvements that included using Likert who score high on both dimensions (labeled
(agree-disagree) scales to rate the extent to “fearful avoidants” by Bartholomew &
which each of the three prototypes described Horowitz, 1991) are especially low in trust
one’s experiences in romantic relationships and seem more likely than other people to
(e.g., Levy & Davis, 1988); decomposition have been hurt or abused in important rela
of the descriptions into separate items that tionships (Shaver & Clark, 1994).
formed multi-item scales (e.g., Collins & The two attachment-style dimensions can
Read, 1990; Feeney & Noller, 1990); split be measured with the 36-item Experiences in
ting the avoidant category into “dismissing” Close Relationships inventory (EC R; Bren
and “fearful” subtypes, thus moving from nan et al., 1998), which is reliable in both the
a three- to a four-category classification internal-consistency and test-retest senses
scheme (Bartholomew & Horowitz, 1991); and has high construct, predictive, and dis
and rewording the instructions and items criminant validity (Crowell, Fraley, & Shav
to examine global attachment style in close er, 1999). Eighteen items tap the avoidance
relationships generally (not just romantic re dimension (e.g., “I try to avoid getting too
lationships), as well as relationship-specific close to my partner”; “I prefer not to show a
styles (e.g., Baldwin, Keelan, Fehr, Enns, & partner how I feel deep down”), and 18 tap
Koh Rangarajoo, 1996; La Guardia, Ryan, the anxiety dimension (e.g., “I need a lot of
Couchman, & Deci, 20 0 0 ). (The history of reassurance that I am loved by my partner”;
this kind of measurement is spelled out in “I resent it when my partner spends time
detail in Chapter 4, M ikulincer & Shaver, away from me”). (Slightly revised but simi
2007a.) lar versions of the scales, labeled the EC R -R ,
Today, adult attachment researchers were created by Fraley, Waller, & Brennan,
working from a personality-social perspec 2000.) The two scales were conceptualized
tive largely agree that attachment styles are as independent and have been found to be
best conceptualized as regions in a two- empirically uncorrelated in most studies.
dimensional (anxiety-by-avoidance) space. Hundreds of studies using self-report mea
The two dimensions are consistently ob sures of adult attachment style, some based
tained in factor analyses of attachment on three categories, some on four categories,
measures (e.g., Brennan, Clark, & Shaver, and some on two dimensions, have found
1998). Moreover, Fraley and Waller (1998) theoretically coherent attachment-style vari
demonstrated that dimensional representa ations in relationship quality, interpersonal
tions of attachment style are more accurate behavior, self-esteem, social cognitions,
than categorical representations. The first emotion regulation, ways of coping with
dimension, attachment-related anxiety , is stress, and mental health. In the remaining
concerned with a strong desire for close sections of this chapter, we provide brief ex
ness and protection, intense worries about amples of these studies (for a comprehensive
partner availability and one’s own value to review, see Mikulincer & Shaver, 2007a).
the partner, and the use of hyperactivating
strategies for dealing with insecurity and
distress. The second dimension, attachment- Individual Differences Related
related av oid an ce , is concerned with dis to A ttachm ent Style
com fort with closeness and dependence on
R elationship Quality
relationship partners, preference for emo
tional distance and self-reliance, and the use In the original studies of adult attachment
of deactivating strategies to deal with inse style, Hazan and Shaver (1987) provided
curity and distress. initial evidence for an association between
People who score low on both dimensions a person’s attachment style (measured with
are said to be secure or to have a secure at the three-category measure reproduced ear
tachment style. They enjoy a chronic sense lier in this chapter) and the way he or she
of attachment security, trust in partners and construes experiences of romantic love. Spe
5. A tta ch m e n t Styles 69
cifically, they found that people who classi 1990), stronger relational commitment (e.g.,
fied themselves as securely attached reported Shaver & Brennan, 1992; Simpson, 1990),
that their love relationships were friendly, and stronger relational cohesion (M ikulinc
warm, trusting, and supportive; they empha er &c Florian, 1999).
sized intimacy as the core feature of these Attachment style seems to be involved in
relationships; and they said they believed in several interpersonal processes that facilitate
the existence of romantic love and the possi or hinder the maintenance of a satisfactory
bility of maintaining intense love over a long couple relationship. For example, several
time period. People with an avoidant style studies have found that higher scores along
described their romantic relationships as the attachment anxiety or avoidance dimen
low in warmth, lacking friendly interactions, sion are associated with less constructive,
and low in emotional involvement; and they less mutually sensitive patterns of dyadic
said that romantic love fades with time. In communication (e.g., Feeney, 1994; Fitzpat
contrast, people who reported an anxious rick, Fey, Segrin, & Schiff, 1993). Moreover,
attachment style described their romantic secure partners have been found to maintain
relationships in terms of obsession and pas more positive patterns of nonverbal com
sion, strong physical attraction, desire for munication (expressiveness, pleasantness,
union with the partner, and proneness to fall attentiveness) than less secure partners (e.g.,
in love quickly and perhaps indiscriminately. Guerrero, 1996; Tucker & Anders, 1998)
At the same time, they characterized their and to be more accurate in expressing their
lovers as untrustworthy and inadequately feelings and noticing their partners’ nonver
supportive; they confessed to intense bouts bal signals (e.g., Feeney, 1994). A person’s
of jealousy and anger toward romantic part attachment style has also been related to the
ners, as well as worries about rejection and methods couples adopt to manage interper
abandonment. Subsequent studies have rep sonal tensions and conflicts (e.g., Gaines et
licated and extended these initial findings, al., 1997; Scharfe & Bartholomew, 1995).
indicating that anxiously attached individu Specifically, secure people rely more heavily
als are less confident than their more secure on effective conflict resolution strategies—
counterparts about being able to establish compromising and integrating their own and
successful relationships (e.g., Carnelley & their partners’ positions. They also display
Janoff-Bulm an, 1992; Pietromonaco & C ar greater accommodation when responding
nelley, 1994) and more likely to emphasize to partners’ anger or criticism. In contrast,
potential losses when thinking about rela insecure people tend to rely on less effec
tionships (Boon & Griffin, 1996). tive conflict resolution strategies that leave
There is good evidence that secure indi conflicts unresolved and may even lead to
viduals tend to maintain more stable roman conflict escalation. Whereas anxious hyper-
tic relationships than insecure people (either activating strategies lead people to intensify
anxious or avoidant) and report higher levels conflict, avoidant deactivating strategies lead
of relationship satisfaction and adjustment people to distance themselves from conflict-
(see Mikulincer & Shaver, 2007a). This pat ual interactions and to avoid engaging with
tern has been consistently obtained in studies their partners.
of both dating and married couples and can Attachment style is also associated with
not be explained by other personality fac sexual motivation and sexual behavior,
tors, such as the “Big Five” personality traits as would be expected based on Bowlby’s
or self-esteem (Mikulincer, Florian, Cowan, (1969/1982) contention that the attachment
& Cowan, 2 0 0 2 ; Noftle & Shaver, 20 0 6 ). behavioral system and the sexual behavioral
For example, Davila, Karney, and Bradbury system are intertwined in romantic/sexual
(1999) collected data every 6 months for relationships (e.g., Brennan & Shaver, 1995;
3 years from newlywed couples and found Mikulincer & Shaver, 2 0 0 7 b ; Tracy, Shaver,
that changes in husbands’ and wives’ reports Albino, & Cooper, 2003). Attachment secu
of secure attachment predicted concurrent rity is associated with sexual satisfaction and
changes in both partners’ reports of marital is conducive to genuine intimacy in sexual
satisfaction. Studies have also linked attach situations, including sensitivity and respon
ment security with greater intimacy (e.g., siveness to a partner’s wishes and openness
Collins &C Read, 1990; Feeney & Noller, to mutual sexual exploration. In contrast,
70 II. I N T E R P E R S O N A L D IS P O S IT IO N S
avoidant individuals tend to remain emo site-sex partners. As compared with secure
tionally detached during sexual activities, people, anxious ones reported higher levels
and anxiously attached individuals tend to of negative emotions and feelings of rejec
hyperactivate sex-related worries and engage tion, especially when interacting with oppo-
in sex primarily to placate a partner, feel ac site-sex partners. Tidwell and colleagues also
cepted, and avoid abandonment (Brassard, found that attachment anxiety was associ
Shaver, & Lussier, 20 0 7 ; Davis, Shaver, & ated with more variability or lability in emo
Vernon, 2 0 0 4 ; Schachner & Shaver, 2004). tional responses and closeness-promoting
Insecurely attached people’s approach behavior. Thus, whereas avoidant people
to sexual activities can also hinder marital seemed to steer clear of intimate exchanges
satisfaction by fostering relational tensions and feel uninvolved, tense, and bored dur
related to fidelity, betrayal, and jealousy. ing daily interactions, more anxious people
For example, Schachner and Shaver (2002) experienced and displayed greater levels of
found that avoidant attachment is associated distress and more ups and downs across
with “mate poaching”— attempts to attract interactions. This finding fits well with
someone who is already in a relationship other evidence concerning anxious people’s
and being open to being “poached” by oth ambivalence and the strong influence of
ers— and with low scores on a relationship perceived availability or unavailability of
exclusivity scale. In contrast, the tendency attachment figures on their emotional reac
of anxious individuals to hyperactivate vigi tions (e.g., Bartz & Lydon, 2 0 0 6 ; Pierce &
lance and concern regarding the possibility Lydon, 2001).
of losing their sexual partners can lead to in Interestingly, Gallo and Matthews (2006)
tense bouts of jealousy, which in turn endan showed that insecurely attached people’s
ger relationship stability and quality. There negative experiences of daily interpersonal
is extensive evidence that anxiously attached interactions tend to be manifested in car
individuals are prone to jealousy and tend diovascular responses. Attachment anxiety
to be overwhelmed by jealous feelings (e.g., was associated with less pleasant and more
Guerrero, 1998; Sharpsteen & Kirkpatrick, conflictual interpersonal exchanges and,
1997). Furthermore, they tend to report high more important, with heightened ambula
levels of suspicion and to cope with them by tory diastolic and systolic blood pressure
engaging in intensive partner surveillance during interactions with friends. Avoidant
(Guerrero, 1998). attachment was associated with heightened
ambulatory diastolic blood pressure during
conflictual interpersonal interactions. These
Interpersonal Interactions
findings suggest that attachment insecurities
People differing in attachment style seem to amplify stress-related physiological reac
differ in the way they construe and experi tions to daily interpersonal interactions.
ence interpersonal exchanges. In six stud A person’s attachment style also shapes
ies, the Rochester Interaction Record (R IR ; his or her reactions to specific kinds of inter
Reis & Wheeler, 1991) was used to examine personal exchanges. For example, extensive
attachment-style differences in daily inter evidence documents attachment-style differ
personal interactions over the course of 1 to ences in the ways people react to others’ of
2 weeks (Kafetsios & Nezlek, 2 0 0 2 ; Kerns fenses and hurtful behaviors. These studies
& Stevens, 1996; Pierce & Lydon, 20 0 1 ; have consistently linked attachment security
Pietromonaco & Barrett, 1997; Sibley, Fis with functional, constructive expressions
cher, & Liu, 2 0 0 5 ; Tidwell, Reis, & Shaver, of anger (nonhostile protests) and attach
1996). As compared with secure people, ment insecurity with less functional forms
avoidant ones reported lower levels of sat of anger, such as animosity, hostility, venge
isfaction, intimacy, self-disclosure, support ful criticism, or vicious retaliation (e.g., M i
ive behaviors, and positive emotions during kulincer, 1998b; Rholes, Simpson, & Orina,
daily interactions, as well as higher levels of 1999; Shaver, Mikulincer, Lavy, & Cassidy,
negative emotions (e.g., boredom, tension). in press; Simpson, Rholes, & Phillips, 1996).
In addition, Tidwell and colleagues (1996) In addition, more avoidant people tend to
found that more avoidant people interacted be less inclined to forgive a hurtful partner
less often and for shorter times with oppo and more likely to withdraw or seek revenge
5. A tta ch m e n t Styles 71
(Mikulincer, Shaver, & Slav, 20 0 6 ). They attachment security was associated with
also reported more intense feelings of vul stronger endorsement of personal values re
nerability or humiliation, a stronger sense of flecting concern for other people’s welfare.
relationship deterioration, and less empathy In addition, Gillath and colleagues (2005)
and understanding associated with forgiving found that avoidant attachment was nega
the offending partner. tively associated with engagement in various
Mikulincer, Shaver, and Slav (2006) pro altruistic activities, such as caring for older
vided initial evidence that people differing in adults and donating blood. Although at
attachment style also differ in the way they tachment anxiety was not related to overall
react to episodes in which another person involvement in such volunteer activities, it
behaves positively toward them. Compared was associated with more self-enhancing or
with less avoidant people, those scoring self-soothing motives for volunteering (e.g.,
high on avoidance were less disposed to feel to feel better about oneself, to enjoy a sense
gratitude. Moreover, when avoidant people of belonging). Overall, these studies indicate
were asked to recall a time when they felt that attachment insecurities interfere with
grateful to a relationship partner, they tend prosocial feelings and behaviors.
ed to remember more negative experiences,
involving more narcissistic threats (e.g., “I
A ttachm ent Sources o f Self-E steem
felt I was risking my personal freedom”; “I
thought I was giving up my dignity”) and As mentioned earlier, Bowlby (1973) argued
distrust, and less happiness and love. These that children construct a model of them
negative responses reflect avoidant people’s selves while interacting with attachment
unwillingness to depend on or be supported figures in times of need. During episodes
by others or to express emotions, such as of attachment-figure availability, children
gratitude, that can be interpreted as indicat can easily perceive themselves as valuable,
ing relational closeness or interdependence. lovable, and special when they are valued,
Attachment style is also associated with loved, and regarded as special by a caring
a person’s attitudes and behaviors during attachment figure. Moreover, they learn to
episodes in which another person expresses view themselves as active, strong, and com
signs of distress and neediness. Several stud petent, because they can effectively mobilize
ies have shown that attachment security is a parent’s support and restore emotional
associated with higher scores on self-report equanimity. In this way, interactions with
scales tapping responsiveness to a relation responsive others and the resulting sense of
ship partner’s needs (e.g., Feeney, 1996; attachment security become primary sources
Kunce & Shaver, 1994) and more supportive of feelings of self-worth and mastery.
actual behaviors toward a distressed part Adult attachment research consistently
ner (e.g., Fraley & Shaver, 1998; Simpson, shows that attachment security is strongly
Rholes, & Nelligan, 1992). In addition, associated with positive self-representations.
Westmaas and Silver (2001) found that As compared with anxiously attached peo
avoidant attachment was associated with ple, secure ones report higher self-esteem
negative attitudes toward a person who had (e.g., Bartholomew & Horowitz, 1991;
been diagnosed with cancer and that attach Mickelson, Kessler, & Shaver, 1997), view
ment anxiety was associated with high levels themselves as more competent and effica
of distress during an interaction with the ill cious (e.g., Cooper, Shaver, & Collins,
person. M ikulincer and colleagues (2001) 1998), and possess more optimistic ex
and Mikulincer, Shaver, Gillath, and Nitz- pectations about their ability to cope with
berg (2005) found that both dispositional stressful events (e.g., Berant, Mikulincer, &
and situationally augmented attachment Florian, 2 0 0 1 ; Cozzarelli, Sumer, & M ajor;
security were associated with heightened 1998). Attachment security is also associated
empathy and compassion for a suffering in with having a coherent, balanced, and well-
dividual. organized model of self. In a series of stud
There is also evidence that attachment ies, Mikulincer (1995) found that, although
security is associated with prosocial values. participants with secure attachment styles
Mikulincer and colleagues (2003) reported tended to recall more positive than nega
that chronic and contextually augmented tive self-relevant traits, they had ready cog
72 II. IN T E R P E R S O N A L D IS P O S IT IO N S
nitive access to both positive and negative other people of the strength of the avoidant
self-attributes in a Stroop task. In addition, self or the neediness of the anxious self.
they revealed a highly differentiated and
integrated self-organization in trait-sorting
A ttachm ent Sources o f Person Perception
tasks and had relatively small discrepancies
between actual-self representations and self Extensive evidence links attachment se
standards (ideal-self and ought-self repre curity to positive perceptions of relation
sentations). That is, attachment security not ship partners. As compared with insecure
only encourages positive self-appraisals but individuals, securely attached people have
also seems to allow people to tolerate weak more positive views of their romantic part
points in the self and integrate them within a ners (e.g., Collins & Read, 1990), perceive
coherent and overall positive self-structure. their partners as more supportive (e.g., Col
According to attachment theory, both of lins & Read, 1990), and feel more trusting
the secondary attachment strategies (anx and affectionate toward their partners (e.g.,
ious hyperactivation and avoidant deactiva Collins & Read, 1990; Simpson, 1990). At
tion) distort a person’s sense of self-worth, tachment security is also associated with
but in different ways. Whereas hyperacti- positive expectations concerning partner be
vating strategies negatively bias anxious haviors (e.g., Baldwin, Fehr, Keedian, Seidel,
people’s sense of self-esteem, deactivating & Thompson, 1993; Baldwin et al., 1996).
strategies favor defensive processes of self For example, Baldwin and colleagues (1993)
enhancement and self-inflation. On the examined the cognitive accessibility of ex
one hand, anxious hyperactivating strate pectations concerning partner’s behaviors in
gies cause attention to be directed to self a lexical-decision task and found that secure
relevant sources of distress (e.g., thoughts people had poorer access to negative partner
about personal weaknesses) and exacerbate behaviors (e.g., partner being hurtful) than
self-defeating self-presentational tendencies, anxious and avoidant people. Attachment
which involve emphasizing helplessness and security is also associated with more posi
vulnerability as a way of eliciting other peo tive explanations of a relationship partner’s
ple’s compassion and support. On the other behavior (e.g., Collins, 1996; Mikulincer,
hand, avoidant deactivating strategies divert 1998a). Collins (1996) asked participants
attention away from self-relevant sources of to explain hypothetical negative behaviors
distress and encourage the adoption of a self- of a romantic partner and found that more
reliant attitude, which requires exaggeration secure individuals were more likely to attri
of strengths and self-worth. bute partner’s negative behaviors to uninten
In studies of these defensive biases, M i tional, unstable, and highly specific causes
kulincer (1998a) examined the way people and less likely to provide explanations that
differing in attachment style also differ in had negative implications for relationship
their self-appraisals following threatening stability.
and neutral situations. Participants with an In contrast, insecure people tend to de
avoidant attachment style made more posi scribe specific friends and romantic partners
tive self-appraisals following threatening as in negative terms and also hold negative views
compared with neutral situations. In con of humanity in general. For example, Collins
trast, anxiously attached participants re and Read (1990) reported that anxiously at
acted to threat with self-devaluation, mak tached people were more likely to believe that
ing more negative self-appraisals following others are difficult to understand and that
threatening as compared with neutral con they have little control over their lives. These
ditions. Mikulincer noted that introducing authors also found that avoidant individuals
contextual factors that inhibited defensive were less likely than other people to believe
tendencies (e.g., a “bogus pipeline” device that human beings are altruistic, willing to
that measures “true feelings about things”) stand up for their beliefs, or able to control
inhibited avoidant participants’ self-inflation their lives. Subsequent studies have found
response, as well as anxious participants’ that these negative views are also manifested
self-devaluation response. That is, insecure in insecure people’s lack of esteem for and
people’s self-appraisals seemed to be strate acceptance of others (e.g., Luke, M aio, &
gic defensive maneuvers aimed at convincing Carnelley, 2 0 0 4 ; Shaver et al., 1996), doubts
5. A tta ch m e n t Styles 73
about other people’s trustworthiness (e.g., which a child learns that acknowledgment
Cozzarelli, Hoekstra, & Bylsma, 2 0 0 0 ), and and display of emotions are functional steps
lack of respect for relationship partners (Frei toward restoring emotional equanimity and
&C Shaver, 2002). that one can feel comfortable exploring,
Secondary attachment strategies are also acknowledging, and expressing one’s own
likely to bias person perception. Avoidant emotions (Cassidy, 1994). In adult attach
individuals, who want to maintain distance ment research, there is extensive evidence
from others and view themselves as strong that secure people, as compared with less
and perfect, are also likely to view them secure ones, tend to score higher on self-
selves as distinctive, unique, and better report and behavioral measures of emotion
than other people. In contrast, anxiously al expressiveness (e.g., Feeney, 1995; Searle
attached people, who want to be loved and & M eara, 1999) and self-disclosure (e.g.,
accepted, are likely to perceive themselves Keelan, Dion, & Dion, 1998; M ikulincer &
as more like others, especially in sharing Nachshon, 1991). For example, M ikulinc
similar problems, so they can more easily er and Nachshon (1991) content-analyzed
feel connected to others. For example, M i participants’ face-to-face verbal disclosures
kulincer, Orbach, and Iavnieli (1998) found of personal information to another person
that whereas anxious individuals were more and found that secure participants disclosed
likely than their secure counterparts to per more intimate and emotion-laden inform a
ceive others as similar to themselves and to tion than avoidant participants. Moreover,
exhibit a false-consensus bias in both trait using a biographical memory task in which
and opinion descriptions, avoidant individu participants were asked to recall specific,
als were more likely than secure individuals early memories of positive and negative emo
to perceive others as dissimilar to them and tions, Mikulincer and Orbach (1995) found
to exhibit a false-distinctiveness bias. M iku that secure participants had more ready
lincer and colleagues also found that anxious mental access to painful memories of anger,
individuals reacted to threats by generating sadness, and anxiety than avoidant people.
self-descriptions that were more similar to When compared with anxious people, se
their partners’ self-descriptions. Avoidant cure people still had better access to positive
individuals, in contrast, reacted to the same memories of happiness and experienced less
threats by generating self-descriptions that automatic spread of memories of other nega
were less similar to their partners’ and by tive emotional experiences.
forgetting more of the traits they and their According to attachment theory, inter
partners shared. actions with available and supportive at
tachment figures promote and reaffirm
optimistic and hopeful appraisals of person-
Em otion R egulation, C oping with Stress,
environment transactions. During positive
and M ental H ealth
interactions with good attachment figures,
According to attachment theory, interac children gradually become convinced that
tions with available attachment figures and distress is manageable, that external obsta
the resulting sense of attachment security cles can be overcome, and that restoration
provide actual and symbolic contexts in of emotional equanimity is only a matter of
which to learn constructive emotion regula time. As a result, secure people can make
tion strategies. Beyond strengthening a per self-soothing reappraisals of aversive events
son’s confidence in the effectiveness of prox that help them resolve distressing episodes
imity bids and support seeking, episodes of with less strain than is experienced by less
attachment-figure availability facilitate the secure people. Indeed, as compared with
adoption of other constructive regulatory anxious and avoidant people, secure people
strategies mentioned earlier in this chapter: have been consistently found to hold more
acknowledgment and display of distress, optimistic appraisals of stressful events (e.g.,
positively reappraising the distress-eliciting Berant et al., 20 0 1 ; Birnbaum, Orr, M iku
situation, and engaging in instrumental lincer, & Florian, 1997; M ikulincer & Flo
problem solving. rian, 1998). For example, Berant and col
Interactions with emotionally accessible leagues (2001) found that securely attached
and responsive others provide the context in mothers of infants who were diagnosed with
74 II. IN T E R P E R S O N A L D IS P O S IT IO N S
congenital heart defects reported more posi ple who were exposed to the traumas of war
tive appraisals of motherhood-related tasks, or terrorism.
both immediately after the diagnosis and 1 Unlike relatively secure people, those
year later, than anxious or avoidant moth who are avoidant cannot readily engage in
ers. Six years later, the effects of insecure optimal problem solving because this often
mothers on their children with congenital requires opening knowledge structures to
heart defects were evident in both objective new information, admitting frustration and
and projective measures administered to the possible defeat, dealing with uncertainty
then 7-year-old children (Berant, M ikulinc and confusion, and running freely through
er, & Shaver, 2008). one’s memories without attempting to block
Experiences of attachment-figure avail attachment-system activation (Mikulincer,
ability also offer opportunities to learn that 1997). Avoidant people often prefer to dis
one’s own instrumental actions are often sociate their emotions from their thoughts
able to reduce distress. For example, a child and actions, using what Lazarus and Folk-
learns that his or her bids for proximity alter man (1984) called “distancing coping.” This
a partner’s behavior and result in the resto requires suppression of emotion-eliciting
ration of emotional equanimity. As a result, thoughts, repression of painful memories,
security-providing interactions strengthen diversion of attention from emotion-related
a person’s reliance on active, instrumental material, and inhibition of verbal and non
approaches to problem solving. In support verbal expressions of emotion. For anx
of this view, secure people have been found iously attached people, in contrast, negative
to rely on problem-focused strategies while emotions can be congruent with their goal
coping with stressful events (e.g., Lussier, of attachment-system hyperactivation. In
Sabourin, & Turgeon, 1997; Mikulincer & the process of emotion regulation, anxious
Florian, 1998). This constructive approach people tend to engage in effortful attempts
to emotion regulation was illustrated by to generate and intensify emotional states.
Mikulincer (1998b), who found that se These states include every emotion that plays
cure participants’ recollections of personal a role in activating the attachment system—
experiences of anger were characterized by threats, dangers, and negative interactions
adaptive problem-solving actions aimed at with attachment figures. They also include
repairing the relationship with the instigator emotions that emphasize a person’s wounds
of anger. and incompetence, such as sadness, anxiety,
Attachment security promotes what shame, and guilt, because these make it nat
Lazarus (1991) called a “short circuit of ural to insist on attachment figures’ atten
threat,” sidestepping the interfering and tion and care (Cassidy, 1994).
dysfunctional aspects of emotions while re These emotion regulation patterns have
taining their functional, adaptive qualities. now been well documented in empirical
Efficient management of distress results in studies of attachment style and ways of cop
more and longer periods of positive mood, ing with stressful events (see Mikulincer &
thereby rendering mood disorders, malad Florian, 1998; M ikulincer & Shaver, 2007a).
justment, and psychopathology less likely. In these studies, higher avoidance scores are
Indeed, several studies have documented associated with higher scores on measures
positive associations between secure at of coping by distancing, and attachment
tachment and measures of well-being (e.g., anxiety is associated with higher scores on
Berant et al., 2 0 0 1 ; Birnbaum et al., 1997) measures of emotion-focused coping. For
and negative associations between security example, Mikulincer and Orbach (1995) re
and symptoms of depression, anxiety, and ported that attachment avoidance was asso
hostility (e.g., Cooper et al., 1998; Mickel- ciated with a repressive coping style, Feeney
son et al., 1997). Mikulincer, Shaver, and (1995) reported that avoidance was related
Horesh (2006) also found that both disposi to behavioral blunting (seeking distractions
tional measures of attachment security and when dealing with stress), and Mikulincer
contextual manipulations of the sense of at and Florian (1998) found that people who
tachment security are associated with lower classified themselves as anxiously attached
levels of posttraumatic symptoms (e.g., in tended to report more frequent task-related,
trusion of traumatic thoughts) among peo ruminative worries after failing cognitive
5. A tta ch m e n t Styles 75
tasks than were reported by their secure and ment anxiety is associated with global dis
avoidant counterparts. tress, depression, anxiety, eating disorders,
These emotion regulation strategies are substance abuse, conduct disorders, and se
also manifested in the ways people cope vere personality disorders (see Mikulincer
with attachment-related threats. For ex & Shaver, 2007a). However, for avoidance,
ample, Fraley and Shaver (1997) found at- the findings are more complex. On the one
tachment-style differences in the suppression hand, a host of studies yielded no significant
of separation-related thoughts. Participants associations between avoidant attachment
wrote continuously about whatever thoughts and self-report measures of well-being and
and feelings they were experiencing while global distress (see Mikulincer & Shaver,
being asked to suppress thoughts about their 2007a). On the other hand, several studies
romantic partners’ leaving them for some indicate that avoidant attachment is associ
one else. Attachment anxiety was associated ated with a pattern of depression character
with poorer ability to suppress separation- ized by perfectionism, self-punishment, and
related thoughts— more frequent thoughts of self-criticism (e.g., Zuroff & Fitzpatrick,
breakup following the suppression task and 1995), heightened reports of somatic com
higher skin conductance during the task. In plaints (e.g., Mikulincer, Florian, & Weller,
contrast, more avoidant people were better 1993), a hostile view of other people (e.g.,
able than less avoidant individuals not only Mikulincer, 1998b), substance abuse and
to stop thinking about separation but also to conduct disorders (e.g., Brennan & Shaver,
reduce the intensity of their autonomic re 1995; Cooper et al., 1998; Mickelson et al.,
sponses to these thoughts. 1997), and schizoid and avoidant personali
In a series of studies examining the experi ty disorders (e.g., Brennan & Shaver, 1998).
ence and management of death anxiety (e.g., In addition, whereas no consistent asso
Mikulincer & Florian, 2 0 0 0 ; Mikulincer, ciation has been found in community sam
Florian, & Tolmacz, 1990), anxious individ ples between avoidant attachment and emo
uals were found to intensify death concerns tional problems, studies that focus on highly
and to keep death-related thoughts active demanding and distressing events reveal
in memory. In contrast, avoidant individu that avoidance is related to greater reported
als tended to suppress death concerns and to distress. For example, in studies assessing
dissociate their conscious claims from their mothers’ long-term reactions to the births of
unconscious anxiety. Although avoidance infants with congenital heart defects, avoid
was related to low levels of self-reported fear ance, as assessed at the times of the initial
of death, it was also related to heightened diagnoses of the infants’ disorders, was the
death anxiety on projective Thematic Ap most potent predictor of maternal distress 1
perception Test (TAT) stories. and 7 years later (Berant et al., 2001, 2008).
Avoidant people’s dissociative tenden It seems that avoidant attachment may con
cies were also documented by M ikulincer tribute to mental health under fairly normal
(1998b), who found that avoidant individu circumstances characterized by only mild
als, as compared with secure ones, reacted encounters with stressors. Under highly de
to anger-eliciting episodes with lower levels manding conditions, however, deactivating
of self-reported anger and higher levels of strategies seem to collapse, and in such cases
physiological arousal (heart rate). Two other avoidant individuals may exhibit high levels
studies examined access to emotions dur of distress and emotional problems. This
ing the AAI, finding that avoidant people conclusion is supported by two laboratory
expressed fewer negative feelings during the studies (Mikulincer, Dolev, & Shaver, 2004)
interview but displayed higher levels of phys that showed that the addition of a demand
iological arousal (heightened electrodermal ing cognitive task, which had previously
activity; Dozier & Kobak, 1992). been shown to interfere with mental sup
Attachment theorists view insecure peo pression (e.g., Wegner, Erber, & Zanakos,
ple’s modes of emotion regulation as risk 1993), impaired avoidant individuals’ ability
factors that reduce resilience in times of to block the activation of attachment-related
stress and contribute to emotional prob worries. Specifically, under high-load condi
lems and poor adjustment. Indeed, a large tions, avoidant participants resembled their
number of studies have shown that attach anxiously attached counterparts, exhibit
76 II. I N T E R P E R S O N A L D IS P O S IT IO N S
ing high accessibility of separation-related Third, there has always been controversy
thoughts and negative self-representations. about the possible role of genes, in addition
to social experience, in determining adult
attachment patterns. There is now prelimi
Concluding R em arks nary evidence that classifications and scores
on both the AAI (Torgerson, Grova, & Som-
As we hope to have shown in this relatively merstad, 20 0 7 ) and the EC R (Crawford et
brief but jam-packed trip through the large al., 2007) are influenced by genetic factors,
and still exploding adult attachment litera as are classifications based on the Strange
ture, Bowlby and Ainsworth’s theory has Situation (Bakermans-Kranenburg & van
been an extremely rich and seminal source IJzendoorn, 20 0 7 ). The degree of genetic in
of ideas for empirical research in personal fluence remains to be clarified.
ity and social psychology. Despite the many Fourth, measures such as the EC R are re
lines of research we have summarized, the at lated to scores on the “Big Five” personality
tachment field is much broader than we have factors (e.g., Donnellan et al., 2 0 0 8 ; Noftle
indicated, including impressive longitudinal & Shaver, 2 0 0 6 ), and those relations are due
studies running from infancy to adulthood in part to shared genetic influences (Craw
(Grossmann, Grossmann, & Waters, 2005). ford et al., 2007; Donnellan et al., 2008).
The entire field is analyzed in the two editions Attachment anxiety, not surprisingly, is
of the H an d b ook o f A ttachm ent (Cassidy & substantially correlated with Neuroticism,
Shaver, 1 9 9 9 ,2 0 0 8 ). Anyone wishing to gain and avoidance is often significantly nega
a reasonably complete picture of the field has tively correlated with Agreeableness and
a great deal of reading to do. Extraversion. Yet many studies of associa
Although there are many well-replicated tions between attachment styles, or attach
research findings in the various streams of ment-style dimensions, and other variables
attachment research, there are still numer find predicted attachment effects even when
ous controversies and conundrums in the scores on Big Five trait measures are statisti
field. First and foremost is the problem of cally controlled (e.g., Erez, Mikulincer, van
nonconvergent measures of adult attachment IJzendoorn, & Kroonenberg, 2 0 0 8 ; Noftle
styles. For example, a recent review of stud & Shaver, 2 0 0 6 ), so attachment insecurities
ies (Roisman, Holland, et al., 2 0 0 7 ) based and major personality factors are not simply
on both the AAI and self-report measures of redundant.
adult attachment, such as the EC R , found Given these controversies and many as
little convergence between the two kinds yet unaddressed questions about personal
of measures, even though some of the stud ity and relationships, the future of adult at
ies revealed substantial associations (e.g., tachment research seems bright. Bowlby and
Shaver, Belsky, & Brennan, 20 0 0 ). Given Ainsworth’s theory is an example of the util
that both kinds of measures are based on ity of grand theories even in a field that is
the same theory, it is not yet clear why both increasingly guided by discrete, focused re
yield coherent support for the theory with search questions. By putting together several
out being strongly related to each other. key theoretical innovations and research ad
Second, it is still unclear whether categori vances of his era, Bowlby was able to retain
cal or dimensional measures of adult attach some of the insights of Freudian psychoana
ment make the most sense, theoretically lytic theory while building bridges to other
and psychometrically. The AAI uses a cat theories and to empirical research findings.
egorical classification system, but the ECR The same kinds of innovations and advances
and similar self-report measures are based have been repeatedly demonstrated in post-
on continuous dimensions. Roisman, Fra Darwinian biology, which is perhaps the best
ley, and Belsky (2007) recently showed that professional model for empirical psychology.
the AAI, especially the distinction between It seems likely that the broad swath of phe
secure and avoidant attachment, should be nomena addressed by attachment theory—
scored dimensionally, an argument Fraley that is, the formation of personality in the
and Spieker (2003) made earlier with respect crucible of interpersonal relationships and
to Ainsworth’s Strange Situation. the shaping of such relationships by person
5. A tta ch m e n t Styles 77
ality factors— will be repeatedly reconceptu Bowlby, J. (1980). A tta c h m e n t a n d loss: Vol. S ad
n ess a n d d ep r es sio n . New York: Basic Books.
alized in future versions of what is currently
Bowlby, J . (1982). A tta c h m e n t a n d lo ss: Vol. 1. A t
called attachment theory. ta c h m en t. New York: Basic Books. (Original work
published 1969)
Bowlby, J. (1988). A secu re b a s e : C lin ica l a p p lic a tio n s
R eferen ces o f a t ta c h m e n t th eo ry . London: Routledge.
Brassard, A., Shaver, P. R., & Fussier, Y. (200 7 ). At
Ainsworth, M . D. S. (1967). In fa n cy in U gan da: In fa n t tachment, sexual experience, and sexual pressure
c a r e a n d th e g r o w th o f lov e. Baltimore: Joh ns H o p in romantic relationships: A dyadic approach. P er
kins University Press. s o n a l R ela tio n s h ip s, 14, 4 7 5 - 4 9 4 .
Ainsworth, M. D. S., Blehar, M . C., Waters, E., & Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998).
Wall, S. (1978). P attern s o f a t ta c h m e n t: A ssessed Self-report measurement of adult attachment: An in
in th e S tran ge S itu ation a n d a t h o m e . Hillsdale, NJ: tegrative overview. In J. A. Simpson & W. S. Rholes
Erlbaum. (Fids.), A tta c h m e n t th e o r y a n d c lo s e re la tio n sh ip s
Bakermans-Kranenburg, M. )., & van IJzendoorn, M. (pp. 4 6 - 7 6 ) . New York: Guilford Press.
H. ( 2 00 7 ). Genetic vulnerability or differential sus Brennan, K. A., & Shaver, P. R. (1995). Dimensions
ceptibility in child development: T he case of attach of adult attachment, affect regulation, and roman
ment. J o u r n a l o f C h ild P sy c h o lo g y a n d P sychiatry, tic relationship functioning. P erson ality a n d S o cia l
4 8 , 1 1 6 0 -1 1 7 3 . P sy ch o lo g y B u lletin , 2 1, 2 6 7 - 2 8 3 .
Baldwin, M. W. (1992). Relational schemas and the Brennan, K. A., & Shaver, P. R. (1998). Attachment
processing of social information. P sy ch o lo g ica l styles and personality disorders: Their connections
B u lletin , 112, 4 6 1 - 4 8 4 . to each other and to parental divorce, parental
Baldwin, M . W., Fehr, B., Keedian, E., Seidel, M ., & death, and perceptions of parental caregiving. J o u r
Thompson, D. W. (1993). An exploration of the n a l o f P erson ality , 6 6 , 8 3 5 - 8 7 8 .
relational schemata underlying attachment styles: Carnelley, K. B., & Jano ff-Bulman, R. (1992). Opti
Self-report and lexical decision approaches. P erso n mism about love relationships: General vs. specific
ality a n d S o c ia l P sy ch olog y B u lletin , 19, 7 4 6 - 7 5 4 . lessons from one’s personal experiences. J o u r n a l o f
Baldwin, M. W., Keelan, J. P. R ., Fehr, B., Finns, V., S o c ia l a n d P erso n a l R e la tio n s h ip s, 9, 5 - 2 0 .
& Koh Rangarajoo, E. (1996). Social-cognitive Cassidy, J. (1994). Emotion regulation: Influences
conceptualization of attachment working models: of attachment relationships. M o n o g ra p h s o f the
Availability and accessibility effects. J o u r n a l o f P er S o c iety fo r R e se a rch in C h ild D e v e lo p m en t, 59,
so n a lity a n d S o c ia l P sy ch olog y , 73, 9 4 - 1 0 9 . 228-283.
Bartholomew, K., & Horowitz, L. M . (1991). A ttach Cassidy, J. (1999). The nature of the child’s ties. In J.
ment styles among young adults: A test of a four- Cassidy & P. R. Shaver (Eds.), H a n d b o o k o f a t t a c h
category model. J o u r n a l o f P erson ality a n d S o cia l m en t: T h eo ry , research , a n d clin ica l a p p lic a tio n s
P sy ch olog y , 61, 2 2 6 - 2 4 4 . (pp. 3 - 2 0 ) . New York: Guilford Press.
Bartz, J. A., & Lydon, J. E. (20 0 6). Navigating the in Cassidy, J . , & Kobak, R. R. (1988). Avoidance and
terdependence dilemma: Attachment goals and the its relationship with other defensive processes. In
use of communal norms with potential close others. J. Belsky & T. Nezworski (Eds.), C lin ic a l im p lic a
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 91, tion s o f a t ta c h m e n t (pp. 3 0 0 - 3 2 3 ) . Hillsdale, NJ:
77-96. Erlbaum.
Berant, E., Mikulincer, M., & Florian, V. (2001). The Cassidy, J ., & Shaver, P. R. (Eds.). (1999). H a n d b o o k
association of mothers’ attachment style and their o f a t ta c h m e n t: T h eo ry , research , a n d c lin ica l a p p li
psychological reactions to the diagnosis of infant’s c a tio n s. New York: Guilford Press.
congenital heart disease. J o u r n a l o f S o c ia l a n d C lin Cassidy, J ., & Shaver, P. R. (Eds.) (20 08 ). H a n d b o o k
ic a l P sy ch olog y , 2 0 , 2 0 8 - 2 3 2 . o f a tta c h m e n t: T h eo ry , research , a n d c lin ic a l a p p li
Berant, E., Mikulincer, M., & Shaver, P. R. (2 008). c a tio n s (2nd ed.). New York.
Mothers’ attachment style, their mental health, and Collins, N. L. (1996). Working models of attachment:
their children’s emotional vulnerabilities: A seven- Implications for explanation, emotion, and behav
year study of mothers of children with congenital ior. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y ,
heart disease. J o u r n a l o f P erson ality , 76, 3 1 - 6 6 . 71, 8 1 0 - 8 3 2 .
Birnbaum, G. E., Orr, 1., Mikulincer, M ., & Florian, Collins, N. L., & Read, S. J. (1990). Adult attachment,
V. (19 97). When marriage breaks up: Does attach working models, and relationship quality in dating
ment style contribute to coping and mental health? couples. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l
J o u r n a l o f S o c ia l a n d P ers o n a l R e la tio n s h ip s, 14, ogy, 5 8, 6 4 4 - 6 6 3 .
643-654. Cooper, M. L „ Shaver, P. R., & Collins, N. L. (1998).
Boon, S. D., & Griffin, D. W. (1996). The construction Attachment styles, emotion regulation, and adjust
of risk in relationships: The role of framing in deci ment in adolescence. J o u r n a l o f P erso n a lity a n d S o
sions about intimate relationships. P erso n a l R e la c ia l P sy ch olog y , 74, 1 3 8 0 - 1 3 9 7 .
tio n sh ip s, J , 2 9 3 - 3 0 6 . Cozzarelli, C., Hoekstra, S. J ., & Bylsma, W. H.
Bowlby, J. (1956). The growth of independence in the (20 0 0). General versus specific mental models of
young child. R o y a l S o c iety o f H ea lth J o u r n a l, 76, attachment: Are they associated with different out
587-591. comes? P erso n a lity a n d S o c ia l P sy ch o lo g y B u lletin ,
Bowlby, J. (1973). A tta c h m e n t a n d loss: Vol. 2 . S e p a 26, 605-618.
ra tion : A n x iety a n d an ger. New York: Basic Books. Cozzarelli, C., Sumer, N., & Major, B. (1998). M e n
78 II. I N T E R P E R S O N A L D IS P O S IT IO N S
tal models of attachment and coping with abortion. namics in separating couples. J o u r n a l o f P e r s o n a l
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 74, ity a n d S o c ia l P sy ch olog y , 75, 1 1 9 8 - 1 2 1 2 .
453-467. Fraley, R. C., & Spieker, S. J. (2003). Are infant at
Crawford, T. N., Livesley, W. J ., Jang, K. L., Shaver, tachment patterns continuously or categorically
P. R., Cohen, P., & Ganiban, J. (200 7). Insecure distributed? A taxometric analysis of Strange Situ
attachment and personality disorder: A twin study ation behavior. D e v e lo p m e n t a l P sy ch olog y , 3 9 ,
of adults. E u r o p ea n J o u r n a l o f P erson ality , 2 1, 387-404.
191-208. Fraley, R. C., & Waller, N. G. (1998). Adult attach
Crowell, J. A., Fraley, R. C., & Shaver, P. R. (1999). ment patterns: A test of the typological model. In
Measurement of adult attachment. In J. Cassidy & J . A. Simpson & W. S. Rholes (Eds.), A tta c h m e n t
P. R. Shaver (Eds.), H a n d b o o k o f a t ta c h m e n t: T h e th e o r y a n d c lo s e rela tio n sh ip s (pp. 7 7 -1 1 4 ). New
o ry , resea rc h , a n d clin ic a l a p p lic a tio n s (pp. 4 3 4 - York: Guilford Press.
465). New York: Guilford Press. Fraley, R. C., Waller, N. G., & Brennan, K. A. (20 0 0).
Davidovitz, R., Mikulincer, M., Shaver, P. R., Ijzak, An item-response theory analysis of self-report
R., & Popper, M. (200 7). Leaders as attachment measures of adult attachment. J o u r n a l o f P ers o n a l
figures: Their attachment orientations predict ity a n d S o c ia l P sy ch o lo g y , 78, 3 5 0 - 3 6 5 .
leadership-related mental representations and fol Frei, J. R., & Shaver, P. R. (2 002). Respect in close re
lowers’ performance and mental health. J o u r n a l o f lationships: Prototype definition, self-report assess
P erson ality a n d S o c ia l P sy ch olog y , 93, 6 3 2 - 6 5 0 . ment, and initial correlates. P erso n a l R ela tio n s h ip s,
Davila, J ., Karney, B. R., & Bradbury, T. N. (1999). 9, 1 2 1 - 1 3 9 .
Attachment change processes in the early years of Gaines, S. O., Jr., Reis, H. T., Summers, S., Rusbult, C.
marriage. J o u r n a l o f P erson ality a n d S o c ia l P sy E ., Co x, C. L., Wexler, M . O., et al. (1997). Impact
c h o lo g y , 76, 7 8 3 - 8 0 2 . of attachment style on reactions to accommodative
Davis, D., Shaver, P. R., & Vernon, M. L. (200 4). At dilemmas in close relationships. P ers o n a l R e la t io n
tachment style and subjective motivations for sex. s h ip s, 4, 9 3 - 1 1 3 .
P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , 3 0 , Gallo, L. C., & Matthews, K. A. ( 2 0 0 6). Adolescents’
1076-1090. attachment orientation influences ambulatory blood
Donnellan, M . B., Burt, S. A., Levendosky, A. A., & pressure responses to everyday social interactions.
Klump, K. L. (20 08 ). Genes, personality, and at P sy c h o s o m a tic M ed icin e, 6 8 , 2 5 3 - 2 6 1 .
tachment in adults: A multivariate behavioral ge George, C., Kaplan, N., & M a in , M . (1985). T h e
netic analysis. P erson ality a n d S o c ia l P sy ch o lo g y A d u lt A tta c h m e n t In terv iew . Unpublished manu
B u lletin , 3 4 , 3 - 1 6 . script, University of California, Berkeley.
Dozier, M ., & Kobak, R. (1992). Psychophysiology Gillath, O., Shaver, P. R., Mikulincer, M ., Nitzberg, R.
in attachment interviews: Converging evidence for A., Erez, A., & van IJzendoorn, M . H. (2005). At
deactivating strategies. C h ild D ev e lo p m e n t, 6 3, tachment, caregiving, and volunteering: Placing vol
1473-1480. unteerism in an attachment-theoretical framework.
Erez, A., Mikulincer, M ., van IJzendoorn, M. H., & P ers o n a l R ela tio n s h ip s, 1 2 , 4 2 5 —4 4 6 .
Kroonenberg, P. M . (2 008). Attachment, person Grossmann, K. E., Grossmann, K., & Waters, E.
ality, and volunteering: Placing volunteerism in an (Eds.). (2 005). A tta c h m e n t fr o m in fa n cy to a d u lt
attachment-theoretical framework. P erson ality a n d h o o d : T h e m a jo r lo n g itu d in a l stu d ies. New York:
In d iv id u a l D iffer en c es , 4 4 , 6 4 - 7 4 . Guilford Press.
Feeney, J . A. (1994). Attachment style, communication Guerrero, L. K. (1996). Attachment-style differences
patterns, and satisfaction across the life cycle of in intimacy and involvement: A test of the four-
marriage. P ers o n a l R e la tio n s h ip s, 1, 3 3 3 - 3 4 8 . category model. C o m m u n ic a tio n M o n o g r a p h s, 6 3,
Feeney, J. A. (1995). Adult attachment and emotional 269-292.
control. P erso n a l R e la tio n s h ip s, 2 , 1 4 3 -15 9. Guerrero, L. K. (1998). Attachment-style differences in
Feeney, J . A. (1996). Attachment, caregiving, and the experience and expression of romantic jealousy.
marital satisfaction. P erso n a l R e la tio n s h ip s, 3, P ers o n a l R e la tio n s h ip s, 5 , 2 7 3 - 2 9 1 .
401-416. Hazan, C., & Shaver, P. R. (1987). Romantic love
Feeney, J. A., & Noller, P. (1990). Attachment style as conceptualized as an attachment process. J o u r n a l
a predictor of adult romantic relationships. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 5 2 , 5 1 1 - 5 2 4 .
o f P erso n a lity a n d S o c ia l P sy ch o lo g y , 5 8 , 2 8 1 - 2 9 1 . Hazan, C., & Shaver, P. R. (1990). Love and work: An
Fitzpatrick, M. A., Fey, J ., Segrin, C., & Schiff, J. L. attachment-theoretical perspective. J o u r n a l o f P er
(1993). Internal working models of relationships s o n a lity a n d S o c ia l P sy ch olog y , 5 9, 2 7 0 - 2 8 0 .
and marital communication. J o u r n a l o f L a n g u a g e Hesse, E. (2008). The Adult Attachment Interview: Pro
a n d S o c ia l P sy ch olog y , 12, 1 0 3 - 1 3 1 . tocol, method of analysis, and empirical studies. In
Fraley, R. C. (2002). Attachment stability from infancy J. Cassidy & P. R. Shaver (Eds.), H a n d b o o k o f a t
to adulthood: Meta-analysis and dynamic modeling ta ch m en t: T h eory , research , a n d clin ical a p p lica tio n s
of developmental mechanisms. P erso n a lity a n d S o (2nd ed., pp. 5 5 2 - 5 9 8 ) . New York: Guilford Press.
c ia l P sy c h o lo g y R e v iew , 6, 1 2 3 - 1 5 1 . Kafetsios, K., & Nezlek, J . B. (20 02 ). Attachment styles
Fraley, R. C., & Shaver, P. R. (1997). Adult at tach in everyday social interaction. E u r o p ea n J o u r n a l o f
ment and the suppression of unwanted thoughts. S o c ia l P sy ch olog y , 3 2 , 7 1 9 - 7 3 5 .
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 73, Keelan, J. R., Dion, K. K., & Dion, K. L. (1998). At
1080-1091. tachment style and relationship satisfaction: Test of
Fraley, R. C., & Shaver, P. R. (1998). Airport separa a self-disclosure explanation. C a n a d ia n J o u r n a l o f
tions: A naturalistic study of adult attachment dy B e h a v io r a l S cien c e, 3 0 , 2 4 - 3 5 .
5. A tta ch m e n t Styles 79
Kerns, K. A., & Stevens, A. C. (1996). Parent-child tional experiences of anger. J o u r n a l o f P erson ality
attachment in late adolescence: Links to social rela a n d S o c ia l P sy ch olog y , 74, 513 - 5 2 4 .
tions and personality. J o u r n a l o f Youth a n d A d o le s Mikulincer, M., Dolev, T., & Shaver, P. R. (2 004).
cen ce, 25, 3 2 3 -3 4 2 . Attachment-related strategies during thought sup
Kunce, L. J . , & Shaver, P. R. (1994). An attachment- pression: Ironic rebounds and vulnerable self
theoretical approach to caregiving in romantic rela representations. J o u r n a l o f P erso n a lity a n d S o cia l
tionships. In K. Bartholomew & D. Perlman (Eds.), P sy ch olog y , 87, 9 4 0 - 9 5 6 .
A d v a n c es in p e r s o n a l r e la tio n sh ip s: A tta c h m e n t Mikulincer, M ., & Florian, V. (1998). The relation
p r o c e s s e s in a d u lt h o o d (Vol. 5, pp. 2 0 5 - 2 3 7 ) . Lon ship between adult attachment styles and emotional
don: Kingsley. and cognitive reactions to stressful events. In J . A.
La Ciuardia, J. G., Ryan, R. M ., Couchman, C. E., & Simpson & W. S. Rholes (Eds.), A tta c h m e n t th eo r y
Deci, E. L. ( 2 0 00 ). Within-person variation in se a n d c lo s e rela tio n sh ip s (pp. 14 3 -1 6 5 ). New York:
curity of attachment: A self-determination theory Guilford Press.
perspective on attachment, need fulfillment, and Mikulincer, M ., & Florian, V. (1999). T he association
well-being. J o u r n a l o f P erso n a lity a n d S o c ia l P sy between spouses’ self-reports of attachment styles
c h o lo g y , 79, 3 6 7 - 3 8 4 . and representations of family dynamics. F am ily
Lazarus, R. S. (1991). E m o tio n a n d a d a p ta tio n . New P ro cess, 3 8 , 6 9 - 83.
York: Oxford University Press. Mikulincer, M ., & Florian, V. (200 0 ). Flxploring indi
Lazarus, R. S., & Eolkman, S. (1984). S tress, a p p r a is vidual differences in reactions to mortality salience:
a l, a n d cop in g . New York: Springer. Does attachment style regulate terror management
Levy, M. B., & Davis, K. E. (1988). Love styles and mechanisms? J o u r n a l o f P erson ality a n d S o c ia l P sy
attachment styles compared: Their relations to each ch o lo g y , 79, 2 6 0 - 2 7 3 .
other and to various relationship characteristics. Mikulincer, M ., Florian, V., Cowan, P. A., & Cowan,
J o u r n a l o f S o c ia l a n d P ers o n a l R e la tio n s h ip s, 5, C. P. (2 002). Attachment security in couple rela
439-471. tionships: A systemic model and its implications for
Luke, M. A., Mai o, G. R ., & Carnelley, K. B. (2004). family dynamics. F am ily P rocess, 4 1, 4 0 5 - 4 3 4 .
Attachment models of the self and others: Relations Mikulincer, M ., Florian, V., & Tolmacz, R. (1990).
with self-esteem, humanity esteem, and parental Attachment styles and fear of personal death: A
treatment. P ers o n a l R e la tio n s h ip s, 11, 2 8 1 - 3 0 3 . case study of affect regulation. J o u r n a l o f P ers o n a l
I.ussier, Y., Sabourin, S., & Turgeon, C. (1997). Co p ity a n d S o c ia l P sy ch olog y , 5 8, 2 7 3 - 2 8 0 .
ing strategies as moderators of the relationship be Mikulincer, M ., Florian, V., & Weller, A. (1993). At
tween attachment and marital adjustment. J o u r n a l tachment styles, coping strategies, and posttrau-
o f S o c ia l a n d P erso n a l R e la tio n s h ip s, 14, 7 7 7 - 7 9 1 . matic psychological distress: The impact of the Gulf
M a i n , M . (19 90). Cross-cultural studies of attachment War in Israel. J o u r n a l o f P erson ality a n d S o c ia l P sy
organization: Recent studies, changing method ch o lo g y , 6 4 , 8 1 7 - 8 2 6 .
ologies, and the concept of conditional strategies. Mikulincer, M ., Gillath, O., Halevy, V., Avihou, N.,
H u m an D ev e lo p m e n t, 3 3 , 4 8 - 6 1 . Avidan, S., & Eshkoli, N. (2001). Attachment theo
M ain , M ., Kaplan, N., & Cassidy, J . (1985). Secu ry and reactions to others’ needs: Evidence that acti
rity in infancy, childhood, and adulthood: A move vation of the sense of attachment security promotes
to the level of representation. M o n o g r a p h s o f th e empathic responses. J o u r n a l o f P erson ality a n d S o
S o c iety f o r R e se a rc h in C h ild D ev e lo p m e n t, 5 0, c ia l P sy ch olog y , 81, 1 2 0 5 - 1 2 2 4 .
66-104. Mikulincer, M ., Gillath, O., Sapir-Lavid, Y., Yaakobi,
Main, M ., & Solomon, J. (1990). Procedures for iden E., Arias, K., Tal-Aloni, L., et al. (2003). Attachment
tifying infants as disorganized/disoriented during theory and concern for others’ welfare: Evidence
the Ainsworth Strange Situation. In M. T. Green that activation of the sense of secure base promotes
berg, D. Cicchetti, & M. Cummings (Eds.), A t ta c h endorsement of self-transcendence values. B a sic a n d
m en t in th e p r e s c h o o l y ea rs: T h eo r y , research , a n d A p p lie d S o c ia l P sy ch olog y , 2 5 , 2 9 9 - 3 1 2 .
in terv en tio n (pp. 1 2 1 -1 6 0 ). Chicago: University of Mikulincer, M ., & Nachshon, O. (1991). Attachment
Chicago Press. styles and patterns of self-disclosure. J o u r n a l o f
Mickelson, K. D., Kessler, R. C., & Shaver, P. R. P erson ality a n d S o c ia l P sy ch olog y , 61, 3 2 1 - 3 3 1 .
(19 97). Adult attachment in a nationally represen Mikulincer, M ., & Orbach, I. (1995). Attachment
tative sample. J o u r n a l o f P erson ality a n d S o c ia l styles and repressive defensiveness: The accessibility
P sy ch olog y , 73, 1 0 9 2 - 1 1 0 6 . and architecture of affective memories. J o u r n a l o f
Mikulincer, M . (1995). Attachment style and the men P erso n a lity a n d S o c ia l P sy ch o lo g y , 6 8 , 9 1 7 - 9 2 5 .
tal representation of the self. J o u r n a l o f P erson ality Mikulincer, M ., Orbach, L, & Iavnieli, D. (1998).
a n d S o c ia l P sy ch olog y , 69, 1 2 0 3 - 1 2 1 5 . Adult attachment style and affect regulation: Stra
Mikulincer, M . (1997). Adult attachment style and in tegic variations in subjective self-other similarity.
formation processing: Individual differences in cu J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 75,
riosity and cognitive closure. J o u r n a l o f P erson ality 436-448.
a n d S o c ia l P sy ch olog y , 7 2 , 1 2 1 7 - 1 2 3 0 . Mikulincer, M ., & Shaver, P. R. (200 4). Security-based
Mikulincer, M . (1998a). Adult attachment style self-representations in adulthood: Contents and
and affect regulation: Strategic variations in self processes. In W. S. Rholes & J. A. Simpson (Eds.),
appraisals. J o u r n a l o f P erso n a lity a n d S o c ia l Psy A du lt a t ta c h m e n t: T h eo ry , research , a n d c lin ica l
c h o lo g y , 75, 4 2 0 - 4 3 5 . im p lic a tio n s (pp. 1 5 9 - 1 9 5 ). New York: Guilford
Mikulincer, M . (1998b). Adult attachment style and Press.
individual differences in functional versus dysfunc Mikulincer, M., & Shaver, P. R. (2007a). A tta c h m e n t
80 II. I N T E R P E R S O N A L D IS P O S IT IO N S
in a d u lt h o o d : S tru ctu re, d y n a m ics, a n d chan ge. Schachner, D. A., Sc Shaver, P. R. (20 04 ). Attachment
New York: Guilford Press. dimensions and motives for sex. P ers o n a l R e la t io n
Mikulincer, M ., & Shaver, P. R. (2 007b). A behavioral sh ip s, 11, 1 7 9 - 1 9 5 .
systems perspective on the psychodynamics of at Scharfe, E., & Bartholomew, K. (1995). Acco mmod a
tachment and sexuality. In D. Diamond, S. J. Blatt, tion and attachment representations in young co u
& J . D. Lichtenberg (Eds.), A tta c h m e n t a n d s e x u a l ples. J o u r n a l o f S o c ia l a n d P erso n a l R ela tio n s h ip s,
ity (pp. 5 1 - 7 8 ). New York: Analytic Press. 12, 3 8 9 - 4 0 1 .
Mikulincer, M ., Shaver, P. R., Gillath, O., Sc Nitzberg, Searle, B., Sc M eara, N. M . (1999). Affective dimen
R. A. (2 005). Attachment, caregiving, and altruism: sions of attachment styles: Exploring self-reported
Boosting attachment security increases compassion attachment style, gender, and emotional experience
and helping. J o u r n a l o f P erson ality a n d S o c ia l P sy among college students. J o u r n a l o f C o u n selin g P sy
c h o lo g y , 89, 8 1 7 - 8 3 9 . c h o lo g y , 4 6 , 1 4 7 -1 58 .
Mikulincer, M., Shaver, P. R., & Horesh, N. (2006). Sharpsteen, D. J ., Sc Kirkpatrick, L. A. (1997). R o
Attachment bases of emotion regulation and post- mantic jealousy and adult romantic attachment.
traumatic adjustment. In D. K. Snyder, J . A. Simp J o u r n a l o f P erson ality a n d S o c ia l P sy ch o lo g y , 7 2,
son, Sc J . N. Hughes (Eds.), E m o tio n reg u latio n 627-640.
in fa m ilies : P ath w a y s to d y sfu n ctio n a n d h ea lth Shaver, P. R., Belsky, J ., Sc Brennan, K. A. ( 2 00 0). The
(pp. 7 7 - 9 9 ) . Washington, DC : American Psycho Adult Attachment Interview and self-reports of ro
logical Association. mantic attachment: Associations across domains
Mikulincer, M ., Shaver, P. R., Sc Slav, K. ( 2 0 0 6). At and methods. P erso n a l R e la tio n s h ip s, 7, 2 5 - 4 3 .
tachment, mental representations of others, and Shaver, P. R., & Brennan, K. A. (1992). Attachment
gratitude and forgiveness in romantic relationships. styles and the “ Big Five” personality traits: Their
In M . Mikulincer Sc G. S. Goo dm an (Eds.), D y connections with each other and with romantic
n a m ic s o f ro m a n tic lo v e : A tta c h m e n t, careg iv in g , relationship outcomes. P erso n a lity a n d S o c ia l P sy
a n d se x (pp. 1 9 0 - 2 1 5 ) . New York: Guilford Press. ch o lo g y B u lletin , 18, 5 3 6 - 5 4 5 .
Noftle, E. E., Sc Shaver, P. R. (20 06 ). Attachment di Shaver, P. R., Sc Clark, C. L. (1994). The psychodynam
mensions and the Big Five personality traits: Asso ics of adult romantic attachment. In J. M. Masling
ciations and comparative ability to predict relation & R. F. Bornstein (Eds.), E m p ir ic a l p er sp e c tiv e s on
ship quality. J o u r n a l o f R e se a rc h in P erson ality , 4 0 , o b je c t rela tio n s t h e o r ie s (pp. 1 0 5 - 1 5 6 ). Washing
179-208. ton, DC : American Psychological Association.
Overall, N. C., Fletcher, G. J . O., & Friesen, M . D. Shaver, P. R., Hazan, C., Sc Bradshaw, D. (1988). Love
(2003). Mapping the intimate relationship mind: as attachment: T he integration of three behavioral
Comparisons between three models of attachment systems. In R. J. Sternberg & M. Barnes (Eds.), T h e
representations. P erso n a lity a n d S o c ia l P sy ch o lo g y p s y c h o lo g y o f lo v e (pp. 6 8 - 9 9 ) . New Haven, CT:
B u lletin , 29(12 ), 1 4 7 9 - 1 4 9 3 . Yale University Press.
Pierce, T., & Lydon, J . (2001). Global and specific re Shaver, P. R., Mikulincer, M ., Lavy, S., Sc Cassidy, J.
lational models in the experience of social interac (in press). Understanding and altering hurt feelings:
tions. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , An attachment-theoretical perspective on the gener
80, 613-631. ation and regulation of emotions. In A. L. Vangelisti
Pietromonaco, P. R ., Sc Barrett, L. F. (1997). Working (Ed.), F eelin g h u rt in c lo s e re la tio n sh ip s. New York:
models of attachment and daily social interactions. Cambridge University Press.
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 73, Shaver, P. R ., Papalia, D., Clark, C. L., Koski, L. R.,
1409-1423. Tidwell, M ., & Nalbone, D. (1996). Androgyny and
Pietromonaco, P. R., & Carnelley, K. B. (1994). Gender attachment security: Two related models of optimal
and working models of attachment: Consequences personality. P erson ality a n d S o c ia l P sy ch olog y B u l
for perceptions of self and romantic relationships. letin , 2 2 , 5 8 2 - 5 9 7 .
P erso n a l R e la tio n s h ip s, I, 6 3 - 8 2 . Sibley, C. G., Fischer, R., Sc Liu, J. H. (20 05 ). Reliabil
Reis, H. T., & Wheeler, L. (1991). Studying social in ity and validity of the revised Experiences in Close
teraction with the Rochester Interaction Record. In Relationships (ECR-R) self-report measure of adult
M . P. Zanna (Ed.), A d v a n ce s in e x p e r im e n ta l s o c ia l romantic attachment. P erso n a lity a n d S o c ia l P sy
p s y c h o lo g y (Vol. 2 4 , pp. 2 7 0 - 3 1 8 ) . San Diego, CA: c h o lo g y B u lletin , 3 1, 1 5 2 4 - 1 5 3 6 .
Academic Press. Simpson, J. A. (1990). Influence of attachment styles
Rholes, W. S., Simpson, J. A., & Orina, M. (1999). At on romantic relationships. J o u r n a l o f P erson ality
tachment and anger in an anxiety-provoking situa a n d S o c ia l P sy ch olog y , 5 9, 9 7 1 - 9 8 0 .
tion. J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , Simpson, J. A., Rholes, W. S., & Nelligan, J. S. (1992).
76, 9 4 0 - 9 5 7 . Support seeking and support giving within couples
Roisman, G. I., Fraley, R. C., Sc Belsky, J. ( 2 0 07 ). A in an anxiety-provoking situation: The role of at
taxometric study of the Adult Attachment Inter tachment styles. J o u r n a l o f P erso n a lity a n d S o c ia l
view. D e v e lo p m e n ta l P sy ch olog y , 4 3 , 6 7 5 - 6 8 6 . P sy ch olog y , 6 2 , 4 3 4 - 4 4 6 .
Roisman, G. I., Holland, A., Fortuna, K., Fraley, R. Simpson, J . A., Rholes, W. S., Sc Phillips, D. (1996).
C., Clausell, E., Sc Clarke, A. ( 2 0 07 ). The Adult At Conflict in close relationships: An attachment per
tachment Interview and self-reports of attachment spective. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l
style: An empirical rapprochement. J o u r n a l o f P er o g y , 71, 8 9 9 - 9 1 4 .
so n a lity a n d S o c ia l P sy ch o lo g y , 9 2 , 6 7 8 - 6 9 7 . Sroufe, L. A., Sc Waters, E. (1 977). Attachment as an
Schachner, D. A., Sc Shaver, P. R. (200 2 ). Attachment organizational construct. C h ild D e v e lo p m e n t, 4 8 ,
style and human mate poaching. N ew R e v ie w o f 1 1 8 4 -1 19 9.
S o c ia l P sy ch olog y , 1, 1 2 2 - 1 2 9 . Tidwell, M . C. O., Reis, H. T., Sc Shaver, P. R. (1996).
5. A tta ch m e n t Styles 81
Attachment, attractiveness, and social interaction: Waters, H. S., Rodrigues, L. M ., & Ridgeway, D.
A diary study. J o u r n a l o f P erso n a lity a n d S o cia l (1998). Cognitive underpinnings of narrative at
P sy ch olog y , 71, 7 2 9 - 7 4 5 . tachment assessment. J o u r n a l o f E x p e r im e n ta l
Torgerson, A. M., Grova, B. K., & Soinmerstad, R. C h ild P sy ch olog y , 71, 2 1 1 - 2 3 4 .
(2 0 07 ). A pilot study of attachment patterns in Waters, H. S., & Waters, E. (200 6 ). The attachment
adult twins. A tta c h m e n t a n d H u m an D e v e lo p working models concept: Among other things, we
m en t, 9, 1 2 7 - 1 3 8 . build script-like representations of secure base ex
Tracy, J. L., Shaver, P. R ., Albino, A. W., & Cooper, periences. A tta c h m e n t a n d H u m an D ev e lo p m e n t,
M. L. (20 03 ). Attachment styles and adolescent sex 8, 1 8 5 - 1 9 7 .
uality. In P. Florsheim (Fid.), A d o le s c e n t r o m a n c e Wegner, D. M ., Erber, R., & Zanak os, S. (1993). Ironic
a n d s e x u a l b e h a v io r : T h eo r y , research , a n d p r a c ti processes in the mental control of mood and mood-
c a l im p lic a tio n s (pp. 1 3 7 -15 9). Mahwah, NJ: Erl related thoughts. J o u r n a l o f P erson ality a n d S o c ia l
baum. P sy ch olog y , 65, 1 0 9 3 - 1 1 0 4 .
Tucker, J. S., & Anders, S. L. (1998). Adult attachment Westmaas, J . , &c Silver, R. C. (2001). T he role of attach
style and nonverbal closeness in dating couples. ment in responses to victims of life crises. J o u r n a l o f
J o u r n a l o f N o n v e r b a l B eh av ior, 2 2 , 1 0 9 - 1 2 4 . P erson ality a n d S o c ia l P sy ch olog y , 8 0 , 4 2 5 - 4 3 8 .
Waters, E., Merrick, S., Treboux, D., Crowell, J ., & Zuroff, D. C., & Fitzpatrick, D. K. (1995). Depressive
Albersheim, L. ( 2 00 0 ). Attachment security in in personality styles: Implications for adult attach
fancy and early adulthood: A twenty-year longitudi ment. P erson ality a n d In d iv id u a l D iffe r e n c e s , 18,
nal study. C h ild D ev e lo p m en t, 71, 6 8 4 - 6 8 9 . 253-265.
CHAPTER 6
Interpersonal Dependency
R o b e r t F. B o r n s t e i n
82
6. In terp erson al D ep en d en cy 83
oral stage (e.g., preoccupation with activities cultures (e.g., India, Japan) have tended to
of the mouth, reliance on food and eating as be more tolerant of dependency in adults
a strategy for coping with anxiety). than are more individualistic cultures (e.g.,
Empirical support for the classical psycho America, Great Britain), wherein depen
analytic model of dependency was mixed (see dency is associated with immaturity, frailty,
Bornstein, 1996), and gradually this perspec and dysfunction (Johnson, 1993; Yamagu-
tive was supplanted by an object relations chi, 2004).
model wherein dependency was conceptual Combining key elements of extant theo
ized as resulting from the internalization of retical frameworks, Bornstein (1992, 1993,
a mental representation of the self as weak 1996, 2005) delineated an interactionist
and ineffectual (Blatt, 1974). Retrospective model wherein interpersonal dependency
and prospective studies of parent-child in is conceptualized in terms of four primary
teractions confirmed that those parenting components: (1) cognitive (i.e., a perception
styles that cause children to perceive them of oneself as powerless and ineffectual cou
selves as powerless and vulnerable are in fact pled with the belief that others are compara
associated with high levels of interpersonal tively powerful and potent); (2) m otivational
dependency later in life (Baker, Capron, & (i.e., a strong desire to obtain and maintain
Azorloza, 1996; Blatt & Homann, 1992). relationships with potential protectors and
Specifically, overprotective and authoritar caregivers); (3) affective (i.e., fear of aban
ian parenting, alone or in combination, are donment, fear of negative evaluation by fig
associated with the development of a de ures of authority); and (4) behavioral (i.e., use
pendent personality, in part because of the of relationship-facilitating self-presentation
impact these two parenting styles have on strategies to strengthen ties to others and
the child’s sense of self. Overprotective par preclude abandonment and rejection). The
enting teaches children that they are fragile links among these four components of de
and weak and must look outward to others pendency are illustrated in Figure 6.1.
for protection from a harsh and threatening As Figure 6.1 shows, three variables (par
environment. Authoritarian parenting, by enting style, gender role socialization, and
contrast, teaches the child that the way to cultural norms regarding achievement and
get by in life is to accede passively to others’ relatedness) are central to the etiology of a
demands and expectations (see Bornstein, dependent personality style, leading to the
1993, 2 0 0 5 , for detailed reviews of studies construction of a “helpless self-concept.”
in this area). This helpless self-concept is the linchpin of a
Behavioral and social learning models dependent personality orientation-—the psy
called psychologists’ attention to the role chological mechanism from which all other
that learning— including observational manifestations of dependency originate.
learning— may play in the etiology and dy First, a perception of oneself as powerless
namics of dependency-related responding. and ineffectual helps create the motivational
As Ainsworth (1969) pointed out, intermit component of dependency: If one views one
tent reinforcement of dependency-related self as weak and ineffectual, then one’s desire
behavior will propagate this behavior over to curry favor with potential caregivers and
time and across situation; as Bandura (1977) protectors will increase. These dependency-
noted, modeling— including symbolic mod related motivations in turn give rise to
eling— can facilitate this learning/reinforce dependency-related behaviors (e.g., ingratia
ment process. Building on these initial social tion, supplication) and to affective responses
learning models, later researchers showed that reflect the dependent person’s core be
that traditional gender role socialization liefs about the self. Finally, as the feedback
practices may help account for the higher lev loop in the right half of Figure 6.1 indicates,
els of overt dependent behavior exhibited by dependency-related affective responses actu
women relative to men insofar as dependent ally reinforce the dependent person’s percep
responding is discouraged more strongly in tion of the self as powerless and ineffectual.
boys than in girls in most Western societies Thus, when a dependency-related affective
(Cross, Bacon, & M orris, 200 0 ). Analyses response (e.g., fear of abandonment by a val
of cultural variations in dependency fur ued other) occurs, the helpless self-concept is
ther indicated that traditionally sociocentric primed (i.e., brought into working memory),
84 II. IN T E R P E R S O N A L D IS P O S IT IO N S
dominated by the belief that there will al vidually the gender of 10 poets after reading
ways be some kind person— a representative brief poem excerpts; the experimenter then
of the mother, of course— to care for them compared the two participants’ judgments
and give them everything they need. This and selected three poems on which they had
optimistic belief condemns them to inactiv disagreed. The experimenter asked the two
ity . . . they make no kind of effort, and in participants to discuss these three poems for
some cases they even disdain to undertake 10 minutes and come to a consensus deci
a breadwinning occupation.” Twenty years sion regarding the gender of the poets.
later Fromm (1947, p. 62) extended this In line with previous results in this area,
characterization of the dependent person, Bornstein and colleagues (1987) expect
noting that these individuals “are dependent ed that the dependent participants would
not only on authorities for knowledge and change their opinions in the majority of
help, but on people in general for any kind of dyads, but in fact the opposite occurred:
support. They feel lost when alone because In 35 of 50 dyads (70% ) the nondependent
they feel that they cannot do anything with participant yielded to the initial opinion of
out help. It is characteristic of these people the dependent participant on at least two of
that their first thought is to find somebody the three poems. Postexperiment interviews
else to give them needed information rather provided some insight regarding the psycho
than make even the slightest effort on their logical processes that led to this unexpected
own.” pattern: A majority of dependent partici
Given these views, it is not surprising that pants indicated that they chose not to alter
throughout much of the 20th century social their initial opinions because they wanted to
research emphasized the passive aspects of impress the experimenter (who— in contrast
dependency, documenting links between de to the typical Asch paradigm— was aware
pendency and suggestibility (Jakubczak & of the participant’s initial opinion before
Walters, 1959; Tribich & Messer, 1974), help the discussions took place). In other words,
seeking (Diener, 1967; Shilkret & Masling, when confronted with choosing between
1981), interpersonal yielding in an Asch- impressing a figure of authority by holding
type paradigm (Kagan & Mussen, 1956; their ground or accommodating a peer by
Masling, Weiss, & Rothschild, 1968), and yielding, the dependent participants opted to
compliance with the perceived expectations stand by their initial opinions and impress
of experimenters (Weiss, 1969) and profes the authority figure.
sors (Masling, O ’Neill, & Jayne, 1981). Even
today researchers tend to focus primarily on
C on text-D riv en Variability in Responding
the passive, acquiescent features of interper
sonal dependency (e.g., Leising, Sporberg, & Following Bornstein and colleagues’ (1987)
Rehbein, 2 0 0 6 ; Vittengl, Clark, & Jarrett, study, researchers became increasingly in
2003). terested in identifying contextual cues that
help shape dependency-related behavior. A
study by Bornstein, Riggs, Hill, and Cala
F rom Pervasive Passivity
brese (1996) was among the first to docu
to G o al-D riv en Activity
ment some of these cues. In Bornstein and
When Bornstein, Masling, and Poynton colleagues’ investigation, same-sex pairs of
(1987) conducted a modified replication of college students were brought to the labo
Masling and colleagues’ (1968) yielding ex ratory and told they were taking part in a
periment, an unexpected pattern emerged. In study of the personality-creativity link.
Bornstein and colleagues’ study, dependent Each pair consisted of one dependent and
and nondependent undergraduates were se one nondependent student, classified using
lected using the ROD scale. Same-sex pairs Hirschfeld and colleagues’ (1977) IDI. The
consisting of one dependent and one nonde two students were told that because they
pendent student were constructed, and par had obtained similar personality profiles in
ticipants were informed that they were tak an earlier testing (actually the dependency
ing part in a study of the decision-making prescreening), they were expected to obtain
process. They were asked to determine indi comparable creativity scores.
88 II. IN T E R P E R S O N A L D IS P O S IT IO N S
H alf the participants were told that their with authoritarian behavior and negative
creativity test data would be seen only by the performance feedback but provided positive
other student (the no-authority condition); feedback under conditions of low daughter
the remaining participants were told their competence. When Thompson and Zuroff
tests would be reviewed by two psychology (1999) replicated this study with m other-
professors who would contact them later son pairs, a similar pattern emerged, with
in the semester to discuss their results (the dependent mothers providing the most posi
authority condition). Participants were then tive feedback to sons who displayed average
given several opportunities to engage in be competence and low autonomy. Apparently
haviors they believed would enhance or un dependent mothers are threatened by compe
dermine their test performance (e.g., choos tent and autonomous behaviors in their sons
ing to do many or few practice items before and daughters and respond to these behav
taking the test, choosing to listen to relaxing iors by subtly undermining their offspring’s
or distracting music while being tested). confidence through negative feedback.
The results of the experiment were clear:
Dependent students “self-handicapped”
A n Interactionist Perspective on
(i.e., did few practice items, chose distract
D ependency
ing background music) in the no-authority
condition, because their primary goal in this These findings, taken together, confirm that
situation was to be liked by the peer. How dependency-related responding is proactive,
ever, dependent students “self-enhanced” goal-driven, and guided by beliefs and ex
(i.e., did many practice items, chose relax pectations regarding the self, other people,
ing background music) in the authority and self-other interactions. Thus the behav
condition, because their primary goal had ior of dependent persons varies considerably
changed: Now, impressing the professors from situation to situation, but the depen
became more important than getting along dent person’s underlying cognitions and mo
with a peer. Nondependent students’ behav tives remain constant. With this in mind, it
ior was unaffected by authority condition. is not surprising that dependent college stu
These findings illustrate the predictable dents who believe they performed well on a
variability in dependency-related behavior major-specific aptitude test choose to wait
and confirm that this variability is largely a significantly longer than high-performing
function of the dependent person’s percep nondependent college students to go over
tions of interpersonal risks and opportuni their test results with one of their major
ties. With no authority figure present, being professors (approximately 15 minutes for
liked by a peer was paramount, but once a the dependent students versus 8 minutes for
figure of authority entered into the equation, nondependent students). These waiting-time
impressing this person became more impor differences increase when the dependent stu
tant than getting along with a peer. Thus dent’s helpless self-concept is activated via
dependent students exhibited a very ratio a series of subliminal lexical primes (Born
nal social influence strategy: They chose to stein, 2 0 0 6 b , Experiment 1). However, when
curry favor with the person best able to offer participants are informed that the professor
protection and support over the long term. who is to go over their test results with them
Using a very different paradigm, Thom p will be leaving the college at the end of the
son and Zuroff (1998, 1999) assessed semester (and therefore cannot offer future
context-driven variability in mothers’ re help and support), dependent-nondependent
sponses to their adolescent sons and daugh waiting-time differences disappear (Born
ters. In their first investigation Thompson stein, 2 0 0 6 b , Experiment 2).
and Zuroff (1998) divided a sample of m oth Other examples of goal-driven “active de
ers into dependent and nondependent groups, pendency” emerge in the medical and aca
then provided each mother false feedback demic arenas. For example, studies indicate
regarding her daughter’s problem-solving that dependent women show shorter laten
skill (competence) and desire to partner with cies than nondependent women in seeking
her mother on a problem-solving task (au medical help following detection of a serious
tonomy). Dependent mothers responded to medical symptom (e.g., a possible lump in the
their daughters’ autonomy and competence breast), in part because the dependent women
6. In terp erso n al D ep end ency 89
are more comfortable seeking help from noted that high levels of interpersonal de
potential caregivers (Greenberg &c Fisher, pendency “contribute to an escalating cycle
1977). Dependent patients also adhere more of coercive control regulated by changes in
conscientiously than nondependent patients emotional distance. Although coercive tac
to medical and psychotherapeutic treatment tics may engender short-term behavioral
regimens (Fisher, W inne, & Ley, 1993; Pol- compliance or intense emotional reunion, a
drugo & Forti, 1988). Other investigations frequently coerced partner is likely to with
indicate that dependent college students are draw emotionally . . . in the long run. As
more willing than nondependent students the batterer’s emotional vulnerabilities are
to seek advice from professors and advisors further activated, he may engage in more
when they are having difficulty with class intense, frequent, and diverse coercive be
material. As a result, dependent college stu havior.”
dents have significantly higher grade point
averages than nondependent college students
with similar demographic backgrounds and Conclusion
comparable Scholastic Aptitude Test (SAT)
scores (Bornstein & Kennedy, 1994). In some ways the evolution of research on
These findings should not be taken to sug interpersonal dependency has paralleled the
gest that all active manifestations of depen broader changes taking place in social psy
dency lead to positive outcomes. On the con chology during the past 50 years. W hat was
trary, dependent elementary school students once conceptualized as a personality pattern
who make frequent contact with the teacher that manifested itself consistently across
are perceived by classmates as being clingy different contexts and settings has come to
and demanding, and these students tend be seen in a more nuanced way, as a set of
to score low on peer ratings of sociometric traits that may be expressed very differently
status and high on self-report measures of depending on the opportunities and con
loneliness (Mahon, 1982; Overholser, 1992; straints characterizing different situations.
Wiggins &c Winder, 1961). Other studies W hat was once conceptualized primarily in
suggest that dependency-related insecurity terms of expressed behavior has come to be
can lead to difficulties in friendships and ro understood in terms of the synergistic inter
mantic relationships and increased conflict play of underlying cognitive, motivational,
with college roommates (Mongrain, Lub and affective processes. And like many vari
bers, &C Struthers, 2 0 0 4 ; Mongrain, Vettese, ables in social psychology that were initially
Shuster, & Kendal, 1998). Dependent psychi conceptualized as reflecting flaws or deficits
atric patients tend to have a higher number in functioning (e.g., high self-monitoring,
of “pseudo-emergencies” than nondepen external locus of control), interpersonal de
dent patients (Emery & Lesher, 1982) and pendency has come to be seen as a personal
to overuse medical and consultative services ity style that can impair adjustment in cer
when hospitalized (O’Neill & Bornstein, tain ways but enhance it in others.
2001), a pattern also displayed by dependent Two trends characterize research on inter
nursing home residents (Bakes, 1996). personal dependency today. First, research
In addition, studies consistently show that ers have begun to explore the possibility that
highly dependent men are at significantly in there are trait-like individual differences in
creased risk for perpetrating partner abuse, the degree to which people express under
in part because these men are fearful of lying dependency needs in adaptive (versus
being abandoned by their partner (Born maladaptive) ways. The concept of healthy
stein, 2 0 0 6 a ; Holtzworth-M onroe, Stuart, dependency overlaps with several other
& Hutchinson, 1997; Kane, Staiger, & Ric- constructs in psychology, sociology, and
ciardelli, 2 0 0 0 ). As a result they tend to medicine, including compensatory depen
overperceive abandonment risk, becoming dency (Bakes, 1996), connectedness (Clark
jealous of even casual contacts between their & Ladd, 2 0 0 0 ), and mature dependency
partner and other men (Babcock, Costa, (Baumeister & Leary, 1995). Research on
Green, & Eckhardt, 20 0 4 ). Murphy, Meyer, healthy dependency is still in its infancy, but
and O ’Leary (1994, p. 734) described this studies suggest that in contrast to unhealthy
dependency-abuse dynamic well when they dependency (which is characterized by in
90 II. IN T E R P E R S O N A L D IS P O S IT IO N S
tense, unmodulated dependency strivings the Proximal Antecedents to Violent Episodes scale.
J o u r n a l o f F am ily P sy ch o lo g y , 18, 4 3 3 - 4 4 2 .
exhibited indiscriminately across a broad
Baker, J . D., Capron, E. W., 8c Azorloza, J . (1996).
range of situations), healthy dependency is Family environment characteristics of persons with
characterized by dependency strivings that— histrionic and dependent personality disorders.
even when strong— are exhibited selectively J o u r n a l o f P erson ality D iso r d er s, 10, 8 2 - 8 7 .
(i.e., in some contexts but not others) and Baldwin, M. W., 8c Sinclair, L. (1996). Self-esteem and
“ if . . . then” contingencies of interpersonal accep
flexibly (i.e., in situation-appropriate ways). tance. J o u r n a l o f P erson ality a n d S o c ia l P sy c h o l
In general, people with a healthy dependent ogy, 71, 1 1 3 0 - 1 1 4 1 .
personality orientation show greater in Baltes, M. M. (1996). T h e m an y fa c e s o f d e p e n d e n c y
sight into their dependency needs than do in o ld ag e. Cambridge, UK: Cambridge University
Press.
unhealthy dependent persons, better social
Bandura, A. (19 77 ). Self-efficacy: Toward a unifying
skills, more effective impulse control, greater theory of behavior change. P sy c h o lo g ic a l R ev iew ,
cognitive complexity, and a more mature de 84, 191-215.
fense and coping style (see Bornstein, 2 0 0 5 , Baumeister, R. F., 8c Leary, M. R. (1995). T he need
and Pincus & W ilson, 2 0 0 1 , for reviews of to belong: Desire for interpersonal attachment as a
fundamental human motivation. P s y c h o lo g ic a l B u l
research in this area). letin , 117, 4 9 7 - 5 2 9 .
Second, researchers have devoted increas Blatt, S. J. (1974). Levels of object representation in
ing attention to exploring the mental rep anaclitic and introjective depression. P s y c h o a n a
resentations and information processing lytic S tu dy o f th e C h ild , 2 9 , 1 0 7 -1 5 7.
Blatt, S. J., 8c Hom ann, E. (1992). Parent-child inter
dynamics associated with a dependent per
action in the etiology of dependent and self-critical
sonality orientation. In the former realm, depression. C lin ica l P sy ch o lo g y R ev iew , 1 2 , 4 7 -
researchers have documented features of the 91.
dependent person’s self-concept (Mongrain, Bornstein, R. F. (1992). The dependent personality:
1998), representations of significant others Developmental, social, and clinical perspectives.
P sy c h o lo g ic a l B u lletin , 1 1 2 , 3 - 2 3 .
(Pincus & W ilson, 2001), and internal work Bornstein, R. F. (1993). T h e d e p e n d e n t p er so n a lity .
ing models of self-other interactions (Meyer New York: Guilford Press.
&C Pilkonis, 2005). In the latter realm, re Bornstein, R. F. (1996). Beyond orality: Toward an ob
searchers have assessed the impact of sublim ject relations/interactionist reconceptualization of
the etiology and dynamics of dependency. P sy c h o
inal lexical priming on dependency-related
a n a ly tic P sy ch olog y , 13, 1 7 7 - 2 0 3 .
interpersonal Stroop latencies (Bornstein et Bornstein, R. F. (1998a). Implicit and self-attributed
al., 20 0 5 ), the impact of self-relevant per dependency needs: Differential relationships to lab
sonality trait feedback (both accurate and oratory and field measures of help-seeking. J o u r n a l
false) on perceptions of dependency-related o f P erson ality a n d S o c ia l P sy ch olog y , 75, 7 7 8 - 7 8 7 .
Bornstein, R. F. (1998b). Implicit and self-attributed
Rorschach imagery (Bornstein, 2 0 0 7 ), and
dependency needs in dependent and histrionic per
cognitive distortions associated with posi sonality disorders. J o u r n a l o f P erson ality A ssess
tively and negatively toned experiences in m en t, 71, 1-14.
close relationships (Mongrain et al., 1998). Bornstein, R. F. (1998c). Interpersonal dependency
and physical illness: A meta-analytic review of ret
Given the impact of dependency-related cog
rospective and prospective studies. J o u r n a l o f R e
nitions on the motivational, affective, and se a r c h in P erson ality , 3 2 , 4 8 0 - 4 9 7 .
behavioral sequelae of interpersonal depen Bornstein, R. F. (1999). Criterion validity of objective
dency, continued exploration of these cogni and projective dependency tests: A meta-analytic
tive features is needed. assessment of behavioral prediction. P sy c h o lo g ic a l
A ssessm en t, 11, 4 8 - 5 7 .
Bornstein, R. F. (2002). A process dissociation ap
proach to objective-projective test score interrela
R eferen ces tionships. J o u r n a l o f P erson ality A s se ss m e n t, 78,
47-68.
Abraham, K. (19 27). The influence of oral erotism on Bornstein, R. F. (2 005). T h e d e p e n d e n t p a tien t: A
character formation. In C. A. D. Bryan 8c A. Stra- p r a c t it io n e r ’s g u id e. Washington, DC: American
chey (Eds.), S e le c te d p a p e r s o n p sy c h o -a n a ly s is Psychological Association.
(pp. 3 9 3 - 4 0 6 ) . London: Hogarth Press. Bornstein, R. F. (2 006a). The complex relationship be
Ainsworth, M . D. S. (1969). Object relations, depen tween dependency and domestic violence: Converg
dency, and attachment: A theoretical review of the ing psychological factors and social forces. A m er i
infant-mother relationship. C h ild D e v e lo p m e n t , can P sy ch olog ist, 61, 5 9 5 - 6 0 6 .
40, 969-1025. Bornstein, R. F. (200 6b). Self-schema priming and de
Babco ck, J. C., Costa, D. M ., Green, C. E., 8c Eck- sire for test performance feedback: Further evalua
hardt, C. I. (200 4). W hat situations induce intimate tion of a cognitive/interactionist model of interper
partner violence?: A reliability and validity study of sonal dependency. S e lf a n d Id en tity , 5, 1 1 0 - 1 2 6 .
6. In terp erso n al D ep end ency 91
Bornstein, R. F. (20 0 7). Might the Rorschach be a pro between willingness to adopt the sick role and at
jective test after all?: Social projection of an unde titudes toward women. J o u r n a l o f C h r o n ic D isea se,
sired trait alters Rorschach Oral Dependency scores. 30, 29-37.
J o u r n a l o f P erson ality A ssessm en t, 8 8 , 3 5 4 - 3 6 7 . Hirschfeld, R. M. A., Klerman, G. I,., Gough, H. G.,
Bornstein, R. F., Bowers, K. S., & Bonner, S. (1996a). Barrett, J., Korchin, S. J ., & Chodoff, P. (1 977). A
Effects of induced mood states on objective and pro measure of interpersonal dependency. J o u r n a l o f
jective dependency scores. J o u r n a l o f P erson ality P erson ality A ssessm en t, 4 1, 6 1 0 - 6 1 8 .
A ssessm en t, 67, 3 2 4 - 3 4 0 . Holtzworth-Monroe, A., Stuart, G. L., 8c Hutchin
Bornstein, R. F., Bowers, K. S., & Bonner, S. (1996b). son, G. (1997). Violent versus nonviolent husbands:
Relationships of objective and projective dependen Differences in attachment patterns, dependency,
cy scores to sex role orientation in college student and jealousy. J o u r n a l o f F am ily P sy ch olog y , 11,
participants. J o u r n a l o f P erson ality A ssessm en t, 314-331.
6 6, 5 5 5 - 5 6 8 . Hunsley, J., 8c Bailey, J . M . (1999). The clinical util
Bornstein, R. F., 8c Kennedy, T. D. (1994). Interper ity of the Rorschach: Unfulfilled promises and an
sonal dependency and academic performance. J o u r uncertain future. P s y c h o lo g ic a l A ssessm en t, 11,
n a l o f P erson ality D iso r d er s, 8 , 2 4 0 - 2 4 8 . 266-277.
Bornstein, R. F., I.anguirand, M. A., Geiselman, K. J., Jaku bczak, L. F., Sc Walters, R. H. (1959). Suggestibil
Creighton, J. A., West, M . A., Gallagher, H. A., et ity as dependency behavior. J o u r n a l o f A b n o r m a l
al. (2003). Construct validity of the Relationship a n d S o c ia l P sy ch olog y , 59, 1 02-107.
Profile Test: A self-report measure of dependency- Johnson, F. A. (1993). D e p e n d e n c y a n d J a p a n e s e s o
detachment./oHrwa/ o f P erson ality A ssessm en t, 8 0, cia liz a tio n . New York: New York University Press.
64-74. Kagan, J . , & Mussen, P. (1956). Dependency themes
Bornstein, R. F., Masling, J. M ., 8c Poynton, F. G. on the TA T and group conformity. J o u r n a l o f C o n
(1987). Orality as a factor in interpersonal yielding. su ltin g P sy ch olog y , 2 0 , 2 9 - 3 2 .
P sy c h o a n a ly tic P sy c h o lo g y , 4 , 161 -1 70 . Kane, T. A., Staiger, P. K., & Ricciardelli, L. A. (2000).
Bornstein, R. F., Ng, H. M ., Gallagher, H. A., Kloss, Male domestic violence: Attitudes, aggression, and
D. M ., 8c Regier, N. G. (2005). Contrasting effects interpersonal dependency. J o u r n a l o f I n te r p e r s o n a l
of self-schema priming on lexical decisions and in V io len ce, 15, 16-29.
terpersonal Stroop task performance: Evidence for Kraepelin, E. (1913). P sy ch iatrie: E in leh rb u ch .
a cognitive/interactionist model of interpersonal de Leipzig, Germany: Barth.
pendency. J o u r n a l o f P erson ality , 7 3 , 7 3 2 - 7 6 1 . Leising, D., Sporberg, D., & Rehbein, D. (2006).
Bornstein, R. F., Riggs, J . M., Hill, E. L., 8c Calabrese, Characteristic interpersonal behavior in dependent
C. (1996). Activity, passivity, self-denigration, and and avoidant personality disorder can be observed
self-promotion: Toward an interactionist model of within very short interaction sequences. J o u r n a l o f
interpersonal dependency. J o u r n a l o f P erson ality , P erson ality D iso rd ers, 2 0 , 3 1 9 - 3 3 0 .
6 4 ,6 3 7 -6 7 3 . Mahon , N. FI. (1982). The relationship of self
Clark, K. E., 8c Ladd, G. W. ( 2 0 00 ). Connectedness disclosure, interpersonal dependency, and life
and autonomy support in parent-child relation changes to loneliness in young adults. N u rsing R e
ships: Links to children’s socioemotional orientation sea rch , 3 1, 3 4 3 - 3 4 7 .
and peer relationships. D e v e lo p m e n ta l P sy ch olog y , Masling, J . M ., O ’Neill, R. M ., 8c Jayne, C. (1981).
36, 485-498. Orality and latency of volunteering to serve as ex
Collins, N. L., & Read, S. J. (1990). Adult attachment, perimental subjects. J o u r n a l o f P erson ality A ssess
working models, and relationship quality in dating m en t, 4 5, 2 0 - 2 2 .
couples. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o l Masling, J. M., Rabie, L., 8c Blondheim, S. H. (1967).
og y , 5 8 , 6 4 4 - 6 6 3 . Obesity, level of aspiration, and Rorschach and TAT
Cross, S. F.., Bacon, P. L., & Morris, M. L. (20 00 ). measures of oral dependence. J o u r n a l o f C o n su ltin g
The relational-interdependent self-construal and re P sy ch olog y , 31, 2 3 3 - 2 3 9 .
lationships. J o u r n a l o f P erson ality a n d S o c ia l P sy Masling, J. M ., Weiss, L. R., 8c Rothschild, B. (1968).
c h o lo g y , 78, 7 9 1 - 8 0 8 . Relationships of oral imagery to yielding behavior
Diener, R. G. (1967). Prediction of dependent behavior and birth order. Jo u r n a l o f C o n su ltin g a n d C lin ic a l
in specified situations from psychological tests. P sy P sy ch olog y , 3 2 , 8 9 - 9 1 .
c h o lo g ic a l R e p o rts , 2 0 , 1 0 3 - 1 0 8 . McClelland, D. C., Koestner, R., &c Weinberger, J.
Emery, G., & Lesher, E. (1982). Treatment of depres (1989). How do self-attributed and implicit motives
sion in older adults: Personality considerations. P sy differ? P sy c h o lo g ic a l R ev iew , 9 6 , 6 9 0 - 7 0 2 .
ch o th er a p y , 19, 5 0 0 - 5 0 5 . Meyer, B., & Pilkonis, P. A. (2 005). An attachment
Fisher, P., Winne, P. H., 8c Ley, R. G. (1993). Group model of personality disorders. In M . F. I.enzen-
therapy for adult women survivors of child sexual weger 8c J . F. Clarkin (Fids.), M a jo r t h e o r ie s o f
abuse: Differentiation of completers versus drop p e r s o n a lity d is o r d e r (2nd ed., pp. 2 3 1 - 2 8 1 ) . New
outs. P sy ch o th era p y , 3 0 , 6 1 6 - 6 2 4 . York: Guilford Press.
Freud, S. (1959). Character and anal erotism. In J. Mischel, W., Shoda, Y., 8c Mendoza-Denton, R. (2002).
Strachey (Fid. 8c Trans.), T h e s ta n d a r d e d itio n o f th e Situation-behavior profiles as a locus of consistency
c o m p le t e p s y c h o lo g ic a l w o r k s o f S ig m u n d F reu d in personality. C u rren t D ire ctio n s in P sy ch o lo g ica l
(Vol. 9, pp. 167-176). London: Hogarth Press. S cien ce , 11, 5 0 - 5 4 .
(Original work published 1908) Mongrain, M. (1998). Parental representations and
Fromm, E. (1947). M an f o r h im self. New York: Rine support-seeking behaviors related to dependency
hart. and self-criticism, jo u r n a l o f P erson ality , 6 6 , 151—
Greenberg, R. P., 8c Fisher, S. (1 977). The relationship 173.
92 II. IN T E R P E R S O N A L D IS P O S IT IO N S
Mongrain, M ., Lubbers, R., 8c Struthers, W. (2004). dence and dependent behavior. J o u r n a l o f P ers o n a l
The power of love: Mediation of rejection in room ity A ssessm en t, 4 5, 1 2 5 - 1 2 9 .
mate relationships of dependents and self-critics. Sperling, M . B., 8c Berman, W. H. (1991). An attach
P erso n a lity a n d S o c ia l P sy ch o lo g y B u lletin , 3 0 , ment classification of desperate love. J o u r n a l o f P er
94-105. s o n a lity A ssessm en t, 5 6 , 4 5 - 5 5 .
Mongrain, M ., Vettese, L. C., Shuster, B., 8c Kendal, Thompson, R ., 8c Zuroff, D. C. (1998). Dependent and
N. (1998). Perceptual biases, affect, and behavior self-critical mothers’ responses to adolescent auton
in the relationships of dependents and self-critics. omy and competence. P ers o n a lity a n d In d iv id u a l
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 75, D iffe r e n c e s , 2 4 , 3 1 1 - 3 2 4 .
230-241. Thompson, R ., 8c Zuroff, D. C. (1999). Dependent
Murphy, C. M ., Meyer, S. L., 8c O ’Leary, K. D. (1994). and self-critical mothers’ responses to adolescent
Dependency characteristics of partner assaultive sons’ autonomy and competence. J o u r n a l o f Y outh
men. J o u r n a l o f A b n o r m a l P sy ch olog y , 103, 7 2 9 - a n d A d o le s c e n c e , 2 8 , 3 6 5 - 3 8 4 .
735. Tribich, D., 8c Messer, S. (1974). Psychoanalytic ch ar
O ’Neill, R. M., 8c Bornstein, R. F. (2001). The depen acter type and status of authority as determiners of
dent patient in a psychiatric inpatient setting: R e suggestibility. J o u r n a l o f C o n su ltin g a n d C lin ica l
lationship of interpersonal dependency to consulta P sy ch olog y , 4 2 , 8 4 2 - 8 4 8 .
tion and medication frequencies. J o u r n a l o f C lin ica l Tyrer, P., Mitchard, S., Methuen, C., 8c Ranger, M.
P sy ch olog y , 57, 2 8 9 - 2 9 8 . (2003). Treatment-rejecting and treatment-seeking
Overholser, J. C. (1992). Interpersonal dependency personality disorders: Type R and Type S. J o u r n a l
and social loss. P erson ality a n d In d iv id u a l D iffe r o f P erson ality D iso rd ers, 17, 2 6 3 - 2 6 8 .
en c e s, 13, 17-23. Vittengl, J. R., Clark, L. A., 8c Ja rrett, R. B. (2003).
Pincus, A. L., 8c Gurtman, M. B. (1995). T he three Interpersonal problems, personality pathology, and
faces of interpersonal dependency: Structural anal social adjustment after cognitive therapy for depres
ysis of self-report dependency measures. J o u r n a l o f sion. P sy c h o lo g ic a l A ssessm en t, 15, 2 9 - 4 0 .
P erso n a lity a n d S o c ia l P sy ch o lo g y , 69, 7 4 4 - 7 5 8 . Weiss, L. R. (1969). Effects of subject, experimenter,
Pincus, A. L., 8c Wilson, K. R. (2001). Interpersonal and task variables on compliance with the experi
variability in dependent personality. J o u r n a l o f P er menter’s expectation. J o u r n a l o f P ro jectiv e T ec h
so n a lity , 69, 2 2 3 - 2 5 1 . n iq u es a n d P erso n a lity A ssessm en t, 3 3 , 2 4 7 - 2 5 6 .
Poldrugo, F., 8c Forti, B. (1988). Personality disorders Wiggins, J. S., 8c Winder, C. L. (1961). T he Peer
and alcoholism treatment outcome. D ru g a n d A lc o Nomination Inventory. P s y c h o lo g ic a l R e p o rts , 9,
h o l D e p e n d e n c e , 2 1 , 171-1 7 6. 643-677.
Schneider, K. (1923). D ie p s y c h o p a th is c h e n p er so n - Yamaguchi, S. (2004). Further clarifications of the
lic h k e ite n . Vienna, Austria: Deuticke. concept of a m a e in relation to dependence and at
Shilkret, C. J ., 8c Masling, J . M . (1981). Oral depen tachment. H u m an D e v e lo p m e n t, 47, 2 8 - 3 3 .
CHAPTER 7
^ m- '•■■■ <&
Machiavellianism
D a n ie l N . J o n e s
D e l r o y L . Pa u l h u s
93
94 II. IN T E R P E R S O N A L D IS P O S IT IO N S
This view has required some adjustment Glenwick, 2 0 0 1 ; Paal & Bereczkei, 2007;
based on recent work wherein Machs were Wastell & Booth, 2003) and emotion recog
asked about their motivations. Compared to nition (Simon, Francis, & Lombardo, 1990)
low M achs, high Machs gave high priority have shown consistent negative correlations
to money, power, and competition (Stewart with Machiavellianism.
& Stewart, 2 0 0 6 ) and relatively low priority In sum, the assumption that Machs have
to community building, self-love, and family superior mental abilities— whether it be IQ,
concerns (McHoskey, 1999). Machs admit EQ , or mind reading— is not supported by
ted to a focus on unmitigated achievement the data. Indeed, one should be cautious
and winning at any cost (Ryckman, Thorn about concluding from M achs’ willingness
ton, & Butler, 1994). Note that this distinc to manipulate others that they are naturally
tive motivational profile does not necessarily skilled at the task. Instead, we argue here
conflict with the original view of Machs as that any manipulative abilities that M achia
purely instrumental: After all, money seek vellians possess derive from superior impulse
ing and power seeking tend to maximize in regulation rather than any special cognitive
strumental benefits in the long run. ability.
social strategies and thus fool some people the three subscales instead of a global mea
some of the time; but repeat offenses lead to sure of locus of control. Further clarification
resentment and social exclusion over time. would be provided by a measure that distin
To date, no empirical evidence supports guished perceptions of control by oneself (“I
these claims. Furthermore, as explained sub can co n tro l. . . ”) from perceptions of control
sequently, we dispute the idea that M achia among others (“People can control . . . ”).
vellians prefer short-term over long-term
strategies.
Worldviews
One might expect a positive association be
T h e ir Personalities
tween Machiavellianism and authoritarian
and Psychological A djustm ent
ism because a condescending attitude toward
Self-Monitoring outgroups is central to both constructs. The
1992 review, however, concluded that over
A personality construct sharing many fea all associations are weak. The exception was
tures of Machiavellianism is self-monitoring a positive correlation between authoritarian
(Snyder, 1974). Although both constructs in ism and the Mach IV Moral Views subscale,
volve social manipulation, Machiavellianism which taps tough-mindedness. That link is
also harbors the darker features of cynical understandable because intolerance of per
worldviews and amorality. In the original sonal weakness is an element of the authori
publication of the self-monitoring scale, Sny tarian personality (Christie, 1991).
der (1974) emphasized their distinctiveness, Since then, the only direct study failed to
and subsequent research confirmed that the find an overall association between M achia
two traits correlate only in the .2 0 -.3 3 range vellianism and authoritarianism, although
(Bolino & Turnley, 2 0 0 3 ; Fehr et al., 1992; both measures predicted the willingness to
Leone & Corte, 1994). volunteer for a study on “prison life,” as well
as endorsement of pragmatic sociopoliti
Locus o f Control cal views (Carnahan & M cFarland, 2007).
Indirect research also indicates links with
The 1992 review indicated (counterintui specific aspects of conservatism and authori
tively) that Machiavellians have an external tarianism (Christie, 1991). For example,
locus of control; that is, they feel that exter Machiavellianism has been linked to tradi
nal forces control people’s behavior and out tional attitudes toward women in the work
comes. More recent studies have reported place (Valentine & Fleischman, 2003). As
the same pattern (Gable 8c Dangello, 1994; noted later, Machs also score relatively low
O ’Connor & M orrison, 2 0 0 1 ; Yong, 1994). on communal values (Trapnell & Paulhus,
Along with Paulhus (1983), we consider that in press; Watson & Morris, 1994). In sum,
conclusion to be misleading. None of these the worldview of Machs is one of pragmatic
studies partitioned perceived control into tough-mindedness.
its three spheres of engagement— personal,
interpersonal, and sociopolitical. Paulhus
showed that these three aspects of perceived Mental Health
control have quite different relations with An analysis of the links between psycho
Machiavellianism. M achs’ apparent exter pathology and Machiavellianism must first
nal locus of control derives entirely from the acknowledge the distinction between Axis I
sociopolitical factor: Machs are simply en and Axis II disorders. We defer our discus
dorsing their cynical view of others’ compe sion of Axis II (personality disorders) to a
tence (see also McHoskey et al., 1999). That subsequent section and deal here with Axis
is, they perceive other people as weak and as I disorders, primarily mood and anxiety dis
having little control over their situations. orders.
In contrast, Machiavellians score quite The 1992 review indicated a consistent
high on measures of interpersonal control. positive association between Machiavel
In this sphere, Machs believe that they can lianism and anxiety. Even Christie and Geis
manipulate others to get what they want. (1970) were suspicious that this counterin
We encourage other researchers to include tuitive association was artifactual, resulting
96 II. IN T E R P E R S O N A L D IS P O S IT IO N S
from the willingness of Machs to disclose cialty (Diehl, Kumar, Gateley, Appleby, &
negative feelings. Wrightsman (1991) agreed O ’Keefe, 20 0 6 ). The latter finding is consis
that high anxiety was at odds with the con tent with the view that, even in helping pro
cept of Machiavellianism, especially their fessions, career choices of high Machs are
detachment in situations of interpersonal motivated by financial goals. Some commen
conflict. More recent research has failed to tators have raised the possibility of a reverse
resolve this paradox, with some studies find causal direction: Certain careers may re
ing no correlation (Allsopp, Eysenck, & Ey ward manipulative behavior, thereby induc
senck, 1991; M cN am ara, Durso, & Harris, ing workers to become more Machiavellian.
200 7 ; Paulhus & W illiams, 2 002) and oth For example, success in some professions is
ers finding a positive correlation (Jakobow- determined by reporting the misbehavior of
itz & Egan, 2 0 0 6 ; Ram anaiah, Byravan, & coworkers (e.g., Girodo, 1998; Macrosson
Detwiler, 1994). & Hemphill, 2001).
The findings on guilt are also inconsistent:
Some research indicates that Machs are more Career Success
guilt prone (Drake, 1995), whereas others re
port that Machs are less guilt prone (Wastell We define career success as effective per
& Booth, 2003). Scattered research indicates formance by a worker in the role assigned
some positive correlations with other forms by the employer. The 1992 review found no
of psychopathology, for example, depression overall evidence that Machiavellianism fa
(Bakir, Yilmaz, & Yavas, 1996), paranoia cilitates such career success. However, more
(Christoffersen & Stamp, 1995), alexithymia recent research using behavioral outcomes
(Wastell & Booth, 2003), socially prescribed indicates a clear pattern. Machs appear to
perfectionism (Sherry, Hewitt, Besser, Flett, have an advantage in unstructured organi
& Klein, 2 0 0 6 ), and low self-esteem (Val zations (Gable, Hollon, & Dangello, 1992;
entine &C Fleischman, 2 0 0 3 ; Yong, 1994). Shultz, 1993). They thrive when they have
Overall links between Mach scales and psy more decision power, fewer rules, and less
chopathology measures appear to be weak managerial supervision. In highly structured
and sample-specific. organizations, high Machs actually perform
Interpersonal adjustment (Axis IV) con worse than low Machs (O ’Connor & M orri
cerns whether individuals have harmonious son, 2 0 0 1 ; Shultz, 1993; Sparks, 1994). Our
relations with other people. Although clear confidence in these conclusions is encour
ly relevant to the psychological adjustment aged by the fact that concrete measures of job
of Machiavellians, the Axis IV diagnosis is, success were used in several of these studies.
once again, mixed. On the one hand, Machs In general, the research on career success is
sometimes harm those around them, as we consistent with the original notion of lati
describe later. On the other hand, as noted, tude for im provisation. As Christie and Geis
Machs can earn liking and respect under se (1970) determined in laboratory research,
Machs remain cool, exploit interpersonal
lect circumstances and time frames (presum
ably when they deem it to be in their interest) relationships, bend the rules, and improvise.
(Hawley, 2006). When this flexibility is constrained, Machs
are likely to incur problems.
A self-report study by Ricks and Fraedrich
C areer Issues (1999) exemplified the tradeoff in the job
success of Machiavellians: High Machs re
Career Choice
ported higher sales volume but also reported
The 1992 review concluded that Machs select significantly lower approval rates from their
occupations that are more business oriented supervisors. By one criterion, Machs are a
and less helping oriented. However, research success; by another criterion, they are not.
shows that Machiavellianism is unrelated to Other research with self-report measures
specialty choice in medical students (Moore, of job success has extended to a wider vari
Katz, & Holder, 1995) and nursing students ety of occupations. Aziz and colleagues re
(Moore & Katz, 1995). Other research on lated success to a new measure they called
medical students finds that Machs are less the Machiavellian Behavior Scale (Mach-
likely to opt for general practice as a spe B). The M ach-B correlated positively with
7. M achiavellian ism 97
managers also report a greater willingness avellianism lies in quadrant 2 of the circum
to use coercive power (Corzine & Hozier, plex, indicating that high Machs are high on
2005). agency and low on communion (Gurtman,
Similarly, children who report bullying 1991, 1992; Wiggins & Broughton, 1991).
(either as perpetrators or victims) score high Work by Locke and colleagues yielded a
er on Machiavellianism (Andreou, 2 0 0 0 , composite variable called self-construal that
2004). Machs may be responding strategi indexes a relative preference for communion
cally to being bullied by bullying others. Or over agency. As expected, self-construal
they may report being bullied to garner ben falls diagonally opposite Machiavellianism
efits from authorities. Alternatively, Machs in circumplex space (Locke & Christensen,
may be more willing to admit to the negative 20 0 7 ), confirming a key suspicion regarding
experiences of both bullying and being bul Machs: They are not simply out to achieve
lied. but rather are out to achieve at the expense
In the one study where both self-report of (or at least without regard for) others.
and behavioral measures were collected on
the same children, a paradox emerged. M a
The Big Five
chiavellianism was positively correlated with
misbehavior on children’s self-reports but Because it is currently the predominant per
not on adult ratings (Loftus & Glenwick, sonality taxonomy, relations of Machiavel
2001). It is unclear whether Machiavellian lianism with the Big Five “supertraits” are
children are exaggerating their misbehavior of interest (Costa & M cCrae, 1992). The
or whether they are successful in inhibiting clearest correlates are low Conscientious
it when adults are present. ness and low Agreeableness (Jakobwitz &
Egan, 2 0 0 6 ; Paulhus & Williams, 2002).
Interestingly, research indicates that Mach
Summary
correlates more highly (and negatively) with
Machiavellian misbehavior is well docu a sixth factor of personality (H onesty-
mented in nonaggressive varieties, namely, Humility) than with any of the Big Five (Lee
cheating, lying, and betrayal. By contrast, & Ashton, 2005).
there is no evidence for overt aggression in
behavioral studies of Machiavellian adults.
The Dark Triad
Three overlapping personality variables—
New D irections Machiavellianism, narcissism, subclinical
psychopathy— have come to be known as the
Situating M achiavellianism
“Dark Triad” of personality: They were so
in Personality Space
named because individuals with these traits
The growing consensus on two structural share a tendency to be callous, selfish, and
models— the Big Five and the interpersonal malevolent in their interpersonal dealings
circumplex— has helped clarify the location (Paulhus & W illiams, 2002). Also overlap
of Machiavellianism in broader personality ping is the P-scale from Eysenck’s P-E-N in
space. Those two models help interpret M a ventory (Allsop et al., 1991), which appears
chiavellianism with respect to fundamental to be conceptually equivalent to subclinical
personality axes, as well as elucidating its psychopathy (Williams & Paulhus, 2004).
overlap with other personality variables. The distinctiveness of the Dark Triad
was disputed by McHoskey and colleagues
(McHoskey, 1995, 2001a; McHoskey et
Interpersonal Circumplex
al., 1998): They argued that, in nonclinical
The interpersonal circumplex is framed in samples such as students, the three variables
terms of two independent axes— agency and are equivalent. Their arguments posed a sig
communion (Wiggins, 1991). Agency refers nificant threat to the discriminant validity of
to the motivation to succeed and individuate the Mach construct. Subsequently, Paulhus
oneself; communion refers to the motivation and colleagues published a series of articles
to merge with others and support the group. confirming their overlap but establishing
Several studies have established that M achi sufficient discriminant validity to recom
7. M achiavellian ism 101
mend measuring all three variables in Mach not everyone in an ecology can be coopera
research (e.g., Paulhus &c W illiams, 2002). tive because the advantage of being a high
The authors argued that a failure to include Mach is too great. However, there are two
the other two Dark Triad members renders reasons why not everyone in an ecology can
ambiguous any research on one member be a high Mach. The first is that low Machs
alone. would have the advantage of building strong
social relationships and cooperative allianc
es, and the second is that high Machs would
Evolutionary O rigins
simply be cheating each other and little
The growing influence of evolutionary psy would be gained. Thus there are at least two
chology has provoked discussion of the good reasons that preclude the full spread of
ancestral origins of Machiavellianism. Al Machiavellianism. One is that Machs have a
though it includes arguments for the advan serious disadvantage in forming cooperative
tages of prosocial traits (such as altruism, alliances that depend on trust. The second
compassion, and cooperation), the hallmark reason is that Machiavellian tendencies will
of evolutionary theory is the notion of the show marginal returns: At some point, high
“selfish gene” (Dawkins, 1989). Contrary to Machs would be trying (unsuccessfully) to
many observers’ intuition, it is not paradoxi cheat each other, and no advantage ensues
cal to include both prosocial and antisocial (Mealey, 1995).
tendencies within the behavioral repertoire
of our species (Krueger, Hicks, & M cGue,
Differential Reproductive Strategies
2001 ).
The natural selection of selfishness would The advantages of high and low Machiavel
naturally foster Machiavellian personalities. lianism should correspond to different repro
In ancestral times, those who exploited op ductive strategies. The opportunism ascribed
portunities to cheat, steal, and manipulate to Machiavellians implies that they focus on
others to achieve their goals would have the short term (Wilson et al., 1996). Such an
outreproduced those who did not. Indeed, opportunistic strategy is especially beneficial
this adaptive advantage been referred to in in unstable environments (Figueredo et al.,
the literature as M achiavellian intelligence 20 05), in which repeated interactions with
(Byrne & Whiten, 1988): The term is often the same individuals are rare. In the words
used interchangeably with terms such as s o of Wilson and colleagues (1996), “advan
cial intelligence, everyday politics, social tages of cooperation are usually long term,
astuteness, political intelligence, practical whereas the advantages of exploitation are
intelligence, em otion al intelligence , and usually only short term” (p. 287).
interpersonal intelligence, all of which al Instead, we agree with Hawley (2006) that
lude to cognitive abilities involving skill at the behavioral repertoire of Machiavellians
adapting to social complexities. Such skills, is “bistrategic,” that is, it includes both co
including the ability to manipulate others, operation and coercion. However, we place
would enhance the control of resources such special emphasis on the fact that neither
as food, shelter, and sex (Hawley, 2 0 0 6 ). In long-term nor short-term cooperative tactics
sum, the term M achiavellian intelligence in Machiavellians reflect true cooperation;
(more than the related terms) implies that instead, such behaviors are in the service of
the skillful manipulation of others conferred malevolence.
a significant evolutionary advantage. O f special concern to evolutionary psy
If Machiavellianism is adaptive, it seems chology are sexual strategies. The data are
that all members of our species should ex clear that high Machiavellians tend to be
hibit that inclination. Instead, we see sub more promiscuous than low Machs (Lin
stantial variation. The explanation may be ton & Wiener, 20 0 1 ; McHoskey, 2001b;
found in arguments put forth by Mealey Schmitt, 2 0 0 4 ; Paulhus &c W illiams, 2002).
(1995). She agreed that antisocial traits Recently, more detailed analyses have parti
such as Machiavellianism and psychopathy tioned promiscuous behaviors and attitudes
may reflect an adaptive reproductive strat (Webster & Bryan, 2007). Exploiting that
egy but argued further that antisocial traits distinction, Jones and Paulhus (2008) found
are frequency dependent. In other words, that Machiavellianism correlated only with
102 II. I N T E R P E R S O N A L D IS P O S IT IO N S
the attitude component. The lack of corre These arguments rest on the assumption
lation with promiscuous behavior suggests that manipulation is more effective for the
that high Machs are no less discerning than gender that prefers promiscuity than the one
low Machs in their actual sexual activities. that prefers investment and commitment.
Such findings are another indication that We dispute that assumption and suggest that
Machiavellians are not solely short term in female Machiavellianism manifests itself in
orientation. a manner consistent with the female repro
ductive agenda.
Sex Differences
D evelopm ental O rigins
Evolutionary psychologists emphasize the
different reproductive challenges faced by Researchers have addressed how M achia
men and women. Because women bear the vellianism develops in individual children.
greater parental burden, they have evolved Christie and Geis (1970) speculated on the
to be more long-term-oriented in their re issue but conducted little developmental
productive strategies than are men (Buss & research. To encourage such research, they
Schmitt, 1993). The short-term reproductive developed the “Kiddie M ach” Scale, which
strategies characteristic of men should pre has been widely used. That version assesses
dict dismissive attachment styles and high Machiavellianism in children by tailoring
levels of mating effort. the language to their level. For example, it
The research on Machiavellianism sup includes the item “The best way to get along
ports the gender difference in short-term re with people is to tell them things that make
productive strategies (Figueredo et al., 2005). them happy” instead of the Mach IV word
Most samples show higher Mach scores in ing, “The best way to handle people is to tell
men than in women (Christie & Geis, 1970) them what they want to hear.”
and in young than in older adults (e.g., Raw- That scale has been used in the rekin
was & Singhapakdi, 1998). These trends dling of research on the topic of M achiavel
suggest that Machiavellianism promotes lianism in children (Repacholi & Slaugh
sexual activity. Individuals who seek mul ter, 2003). In M cllw ain’s (2003) review of
tiple short-term sexual opportunities (e.g., that research, she concluded that the young
those unrestricted in sociosexuality) would Machiavellian is characterized by mistrust,
benefit from manipulative tendencies and a cynicism, and affective blunting. Lack of
lack of empathy. empathy, in particular, plays a causal role
Further research confirms that M achia in determining a young M achiavellian’s be
vellianism confers a special reproductive ad havior.
vantage on men. Linton and Wiener (2001) A factor analysis by Sutton and Keogh
showed that high Mach men reported higher (2001) revealed three factors in the Kiddie
rates of possible conceptions than low Mach Mach Scale: lack of faith in human nature,
men. One possible explanation is that M a dishonesty, and distrust. Only lack of faith
chiavellians are likely to deceive, coerce, and in human nature correlated with age, sug
manipulate partners into sex (Jones, Harms, gesting that cynicism increases over time.
& Paulhus, 2008). Mach is positively asso The authors also suggested that, initially,
ciated with a variety of deceptive and self- children may not differentiate manipulative
serving tactics in romantic relationships that from prosocial behavior. In other words,
include feigning love, intoxicating partners, they see doing and saying things to make
divulging intimate secrets, infidelity, and other people “happy” as commendable rath
coercion (McHoskey, 2001b). The fact that er than dishonest or unethical.
these associations were more pronounced for As noted earlier, some writers had an
men than for women led McHoskey (2001b) ticipated that Machiavellians would have
to conclude that biological sex moderates an advanced theory of mind. Instead, the
the effect of Machiavellianism on sexual research showed no relation with theory of
behavior. Insofar as men are more likely to mind but a growing negativity among those
benefit from short-term opportunistic repro scoring high on the Kiddie-Mach. As a re
ductive strategies, this interaction is predict sult, Kiddie-Machs receive ambivalent reac
able from evolutionary psychology (Buss & tions from others, even in preschool years
Schm itt, 1993). (Repacholi et. al., 2003).
7. M achiavellian ism 103
A radically different conclusion has been 1970). Subsequent evidence has supported
drawn by Hawley (2006). In her view, M a those three notions. As noted, Machs seem
chiavellian children are received well by to thrive in business situations with a high
their peers and indeed are socially compe latitude for improvisation (Shultz, 1993),
tent in almost every respect. The difference but they perform worse in other situations,
in her conclusion may derive from the dif such as when latitude for improvisation
ferent methodology employed. Rather than is impeded (Sparks, 1994). Even after suc
measure children with Kiddie-Mach, she di cessful manipulations, Machs may suffer a
rectly observed the behavior of socially com decrement in reputation that reduces future
petent versus socially inept children. Those opportunities (Wilson et al., 1996).
who use both coercive and prosocial strate The source of evaluation may influence
gies (i.e., bistrategic controllers) were labeled whether Machs are judged as successful or
Machiavellian (Hawley, 2003). not. When evaluated by a supervisor, Machs
Only recently has a behavioral genetics seem to evoke negative evaluations, but they
study permitted insight into possible genetic concomitantly report and record higher lev
and environmental causes. In addition to a els of sales in certain jobs (Ricks & Frae-
genetic component in common with narcis drich, 1999).
sism and psychopathy, Machiavellianism
shows a substantial shared-environment
Communal Goals
component (Vernon, Villani, Vickers, &
Harris, 2008). The latter implicates social Surprising to the intuitions of some commen
ization mechanisms, such as parental model tators, Machs may be just as generous and
ing or an overreaction to harsh or unpredict helpful as others, depending on the situation.
able family environments. For example, Bereczkei, Birkas, and Kerekes
A few other studies point to possible (2007) found that Machs volunteer less than
genetic-environment interactions. By late low Machs unless their volunteering is made
adolescence, M ach scores of sons correlated public, thus promoting a strong reputation
positively with parents’ Mach scores, sup (Bereczkei, Birkas, 8c Kerekes, 2009).
porting a modeling hypothesis (Ojha, 2007). Group members may prefer high Machs
Daughters in father-absent families report for roles that help the group deal with en
higher levels of Machiavellianism, but not emies and opponents (Wilson et al., 1998).
toward family members (Barber, 1998). A classic example is the preference for a
Adding complexity, there is evidence that Machiavellian as president of the United
children’s Mach scores initially oppose, but States (Deluga, 20 0 1 ; Simonton, 1986). On
later come to match, parental scores (Gold, the other hand, Machs are less favored as
Christie, 8c Friedman, 1976). friends, confidants, and business partners
(Wilson et al., 1998).
Length of interaction also plays a role. As
M achiavellianism as a Personality Tradeoff
noted by Fehr and colleagues (1992), high
A repeated theme in this chapter is the no Machs are more liked in short-term encoun
tion that Machiavellianism harbors both ters (such as when participants are viewing
adaptive and maladaptive qualities. Key to a videotape) (Ickes, Reidhead, 8c Patterson,
understanding this tradeoff is the distinction 1986). However, when individuals simulate
between agentic and communal notions of the experience of engaging with a high Mach
adaptiveness. Adaptiveness for agentic goals (such as by reading a first-person story), they
concerns the promotion of personal achieve judge Machs more negatively (Wilson et al.,
ment, whereas adaptiveness for communal 1998).
goals concerns the benefits to one’s group.
M achiavellianism R efined:
Agentic Goals R eturning to Its Roots
A consistent theme in the literature has been On the whole, our review of the literature
that Machs thrive best in contexts that (1) has sustained the construct validity of M a
afford face-to-face interaction, (2) allow chiavellianism as measured by the Christie
latitude for improvisation, and (3) involve and Geis (1970) instruments. There is sub
emotional distractions (Christie 8c Geis, stantial confirmation of M achs’ cynical
104 II. IN T E R P E R S O N A L D IS P O S IT IO N S
worldview, pragmatic ethics, and use of du we have begun work on a refined measure,
plicitous tactics. Furthermore, the apparent dubbed Mach VI (Jones & Paulhus, 2008).
exceptions noted throughout this chapter fit Preliminary research indicates that the Mach
a coherent pattern. VI does show the necessary properties to tap
Disconcerting, however, are reports of a more strategic form of Machiavellianism.
positive associations of Mach IV with im
pulsivity (Marusic et al., 1995). Certainly,
impulsive hostility may represent an evolu- N ote
tionarily viable strategy, but the appropriate
label for that personality type is subclinical 1. Pronounced “ m ack,” these labels are not to be
psychopathy (Paulhus & W illiams, 2002). confused with “ m aw k,” as in M ach 4 (four times
the speed of sound).
Psychopaths and Machiavellians do share
similar antisocial tendencies (Mealey, 1995),
but the original theory— from Machiavelli
R eferen ces
(1513) to Christie and Geis (1970)— speci
fied clearly that Machs are cool and strategic Allsopp, J ., Eysenck, H. J ., & Eysenck, S. B. G. (1991).
rather than hostile and impulsive. Machiavellianism as a component in psychoticism
To support our case, we draw attention and extraversion. P erso n a lity a n d In d iv id u a l D if
fe r e n c e s , 1 2, 2 9 - 4 1 .
to a relatively unmined source regarding
Andreou, E. ( 2 00 0 ). Bully/victim problems and their
manipulative strategies, namely, Sun-tzu’s association with psychological constructs in 8 to
Art o f War (1998). His writings anticipated 12-year-old Greek schoolchildren. A g g ressiv e B e
those of Machiavelli by nearly 2 ,0 0 0 years h av ior, 2 6 , 4 9 - 5 6 .
yet have been singularly overlooked. Most Andreou, E. (200 4). Bully/victim problems and their
association with Machiavellianism and self-efficacy
relevant to our current point is the special
in Greek primary school children. B ritish J o u r n a l o f
emphasis that Sun-tzu placed on the cool E d u c a tio n a l P sy ch olog y , 74, 2 9 7 - 3 0 9 .
preparation required to effect successful po Austin, E. J ., Farrelly, D., Black, C., & Moore, H.
litical and military outcomes. In sum, the (2 0 07 ). Emotional intelligence, Machiavellianism
emphasis on cool strategy in all key theo and emotional manipulation: Does El have a dark
side? P erson ality a n d In d iv id u a l D iffe r e n c e s , 4 3,
retical sources is not entirely consistent with 179-189.
current measures of Machiavellianism (see Aziz, A. (2004). Machiavellianism scores and self-
also Hawley, 2006). rated performance of automobile salespersons. Psy
Our conclusion is that the Mach IV needs c h o lo g ic a l R e p o r ts , 9 4 , 4 6 4 - 4 6 6 .
refinement to better represent this strategic Aziz, A. (2005). Relationship between Machiavellian
ism scores and performance of real estate salesper
element. An improved scale would confirm sons. P s y c h o lo g ic a l R e p o r ts , 9 6 , 2 3 5 - 2 3 8 .
that (1) Machs are less impulsive than psy Aziz, A., May, K., & Crotts, J . C . (2002). Relations of
chopaths and no more impulsive than non- Machiavellian behavior with sales performance of
M achs, (2) Machs manipulate in the long stockbrokers. P sy c h o lo g ic a l R e p o r ts , 9 0, 4 5 1 - 4 6 0 .
Bakir, B., Yilmaz, R., & Yavas, S. (1996). Relating de
term as well as the short term, and (3) Machs
pressive symptoms to Machiavellianism in a Turk
engage in aggression (including revenge) only ish sample. P sy c h o lo g ic a l R e p o r ts , 7 8, 1 0 1 1 -1 01 4.
to the degree that it is deemed profitable. In Barber, N. (1998). Sex differences in disposition t o
short, Machs should be strategic, as well as wards kin, security of adult attachment, and socio
tactical. sexuality as a function of parental divorce. E v o lu
tion a n d H u m an B eh a v io r, 19, 1 2 5 - 1 3 2 .
Strategic Machiavellians should be willing
Bass, K., Barnett, T., & Brown, G. (1999). Individual
to forgo short-term benefits to achieve long difference variables, ethical judgments, and ethical
term benefits. One prediction is that M achi behavioral intentions. B u sin ess E th ics Q u arterly , 9,
avellians (as opposed to psychopaths) should 183-205.
pay close attention to their reputations. As Becker, J . A., 8c O ’Hair, H. D. ( 2 0 07 ). Machiavellians’
motives in organizational citizenship behavior.
Machiavelli suggested, the generation and
J o u r n a l o f A p p lied C o m m u n ic a tio n R e se a rch , 35,
maintenance of a favorable or menacing rep 246-267.
utation can reap benefits across a sustained Bereczkei, T. Birkas, B., 8c Kerekes, Z. (20 09 ). T h e
period of time. Although key theoretical p r e s e n c e o f o th er s, p r o s o c ia l traits, M a c h ia v e llia n
sources emphasize its importance, reputa ism : A p e r s o n a lity x s itu a tio n a p p r o a c h . M an u
script submitted for publication.
tion propagation has been overlooked by Bereczkei, T., Birkas, B., 8c Kerekes, Z. (20 07 ). Pub
allowing impulsive content to contaminate lic charity offer as a proximate factor of evolved
the Mach IV scale. To rectify this deficit, reputation-building strategy: An experimental anal
7. M achiavellian ism 105
Hawley, P. H. (2003). Prosocial and coercive configu Lopes, J . , Sc Fletcher, C. (200 4). Fairness of impres
rations of resource control in early adolescence: A sion management in employment interviews: A
case for the well-adapted Machiavellian. M e r r ill- cross-country study of the role of equity and M a
P a lm e r Q u arterly , 49, 2 7 9 - 3 0 9 . chiavellianism. S o c ia l B e h a v io r a n d P erson ality ,
Hawley, P. H. ( 2 0 06 ). Evolution and personality: A 32, 747-768.
new look at Machiavellianism. In D. Mroczek & Macrosson, W. D. K., Sc Hemphill, D. J. (2001). M a
T. Little (Eds.), H a n d b o o k o f p e r s o n a lity d e v e lo p chiavellianism in Belbin team roles. J o u r n a l o f
m en t (pp. 147-161). M ah w ah , NJ: Erlbaum. M a n a g e ria l P sy ch o lo g y , 16, 3 5 5 - 3 6 3 .
Hegarty, H. W. (1995). Effects of group norms and M artin , M . W., Anderson, C. M ., Sc Thweatt, K. S.
learning on unethical decision behavior. P sy c h o (1998). Aggressive communication traits and their
lo g ic a l R e p o rts , 76, 5 9 3 - 5 9 4 . relationships with the cognitive flexibility scale and
Ickes, W., Reidhead, S., & Patterson, M . (1986). M a the communication flexibility scale. J o u r n a l o f S o
chiavellianism and self-monitoring: As different as c ia l B e h a v io r a n d P erson ality , 13, 5 3 1 - 5 4 0 .
“me” and “you.” S o c ia l C o g n itio n , 4, 5 8 - 7 4 . Marusic, I., Bratko, D., & Zarevski, P. (1995). Self-
Jakobw itz, S., & Egan, V. (20 06 ). T he dark triad and reliance and some personality traits: Sex differ
normal personality traits. P erso n a lity a n d In d iv id ences. P erson ality a n d In d iv id u a l D iffe r e n c e s , 19,
u a l D iffe r e n c e s , 4 0 , 3 3 1 - 3 3 9 . 941-943.
Jones, D. N., Harms, P. D., 8c Paulhus, D. L. (2 008). McHoskey, J. W. (1995). Narcissism and Machiavel
T h e s e x u a l p r o file o f th e D ark T riad . Manuscript in lianism. P sy c h o lo g ic a l R e p o r ts , 77, 7 5 5 - 7 5 9 .
preparation. McHoskey, J. W. (1999). Machiavellianism, intrinsic
Jones, D. N., Sc Paulhus, D. L. ( 2 0 0 8 , February). A versus extrinsic goals and social interest: A self-
m ea su re o f strateg ic M a c h ia v e llia n is m : M a ch VI. determination theory analysis. M o tiv a tio n a n d
Paper presented at the meeting of the Society for Per E m o tio n , 2 3 , 2 6 7 - 2 8 3 .
sonality and Social Psychology, Albuquerque, N M . McHoskey, J. W. (2001a). Machiavellianism and per
Jones, G. E., Sc Kavanagh, M . J. (1996). An experi sonality dysfunction. P erson ality a n d In d iv id u a l
mental examination of the effects of individual and D iffe r e n c e s , 3 1 , 7 9 1 - 7 9 8 .
situational factors on unethical behavioral inten McHoskey, J . W. (2001b). Machiavellianism and sexu
tions in the workplace. J o u r n a l o f B u sin ess E th ics, ality: On the moderating role of biological sex. P er
IS , 5 1 1 - 5 2 3 . so n a lity a n d In d iv id u a l D iffe r e n c e s , 3 1, 7 7 9 - 7 8 9 .
Kashy, D. A., 8c DePaulo, B. M . (1996). W ho lies? McHoskey, J. W., Hicks, B., Betris, T., Szyarto, C.,
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o lo g y , 70, Worzel, W., Kelly, K., et al. (1999). Machiavellian
1037-1051. ism, adjustment, and ethics. P sy c h o lo g ic a l R e p o rts ,
Krueger, R. F., Hicks, B. M ., & McGue, M . (2001). 8 5, 1 3 8 - 1 4 2 .
Altruism and antisocial behavior: Independent ten McHoskey, J . W., Worzel, W., & Szyarto, C. (1998).
dencies, unique personality correlates, distinct eti Machiavellianism and psychopathy. J o u r n a l o f P er
ologies. P s y c h o lo g ic a l S c ien ce, 1 2, 3 9 7 - 4 0 2 . so n a lity a n d S o c ia l P sy ch o lo g y , 74, 1 9 2 - 2 1 0 .
Kumar, K., Sc Beyerlein, M. (1991). Construction and Mc llw ain, D. (2 003). Bypassing empathy: A M a c hia
validation of an instrument measuring ingratiatory vellian theory of mind and sneaky power. In B.
behaviors in organizational settings. J o u r n a l o f A p Repacholi Sc V. Slaughter (Eds.), In d iv id u a l d if
p lie d P sy ch olog y , 76, 6 1 9 - 6 2 7 . fe r e n c e s in th e o r y o f m in d (pp. 1 3 - 3 8 ) . New York:
Leary, M. R ., Knight, P. D., Sc Barnes, B. D. (1986). Psychology Press.
Ethical ideologies of the Machiavellian. P erson ality M cN am ara, P., Durso, R., Sc Harris, E. ( 2 0 07 ). “M a
a n d S o c ia l P sy c h o lo g y B u lletin , 12, 7 5 - 8 0 . chiavellianism” and frontal dysfunction: Evidence
Lee, K., & Ashton, M . C. (20 05 ). Psychopathy, M a from Parkinson’s disease. C o g n itiv e N e u r o p s y c h ia
chiavellianism, and narcissism in the Five-Factor try, 12, 2 8 5 - 3 0 0 .
Model and the H E X A C O model of personality Mealey, L. (1995). The sociobiology of sociopathy:
structure. P erson ality a n d In d iv id u a l D iffer en c es , An integrated evolutionary model. B e h a v io r a l a n d
38, 1571-1582. B rain S cien ces, 18, 5 2 3 - 5 9 9 .
Leone, C., Sc Corte, V. (1994). Concern for self Meyer, H. D. (1992). Norms and self-interest in ulti
presentation and self-congruence: Self-monitoring, matum bargaining: The prince’s prudence. J o u r n a l
Machiavellianism and social conflicts. S o c ia l B e o f E c o n o m ic P sy ch o lo g y , 13, 2 1 5 - 2 3 2 .
h a v io r a n d P erson ality , 2 2 , 3 0 5 - 3 1 2 . Moore, S., Sc Katz, B. (1995). Machiavellianism scores
Linton, D. K., Sc Wiener, N. I. (2001). Personality and of nursing faculty and students. P sy c h o lo g ic a l R e
potential conceptions: Mating success in a modern p o r ts , 77, 3 8 3 - 3 8 6 .
western male sample. P erso n a lity a n d In d iv id u a l M oore, S., Katz, B., Sc Holder, J . (1995). Machiavel
D iffe r e n c e s , 3 1 , 6 7 5 - 6 8 8 . lianism and medical career choices. P sy c h o lo g ic a l
Liu, C. C. (2 008). The relationship between M achia R e p o r ts , 76, 8 0 3 - 8 0 7 .
vellianism and knowledge-sharing willingness. Mudrack, P. E. (1993). An investigation into the accept
J o u r n a l o f B u sin ess P sy ch o lo g y , 2 2 , 2 3 3 - 2 4 0 . ability of workplace behaviors of a dubious ethical
Locke, K. D., Sc Christensen, L. (200 7). Re-construing nature. J o u r n a l o f B u sin ess E th ics, 1 2, 5 1 7 - 5 2 4 .
the relational-interdependent self-construal and its Mudrack, P. E., Sc M ason, E. S. (1995). More on ac
relationship with self-consistency. J o u r n a l o f R e ceptability of workplace behaviors of a dubious eth
se a r c h in P erson ality , 41, 3 8 9 - 4 0 2 . ical nature. P s y c h o lo g ic a l R e p o r ts , 76, 6 3 9 - 6 4 8 .
Loftus, S. T., Sc Glenwick, D. S. (2001). Machiavel Mudrack, P. E., Mason , E. S., Sc Stepanski, K. M.
lianism and empathy in an adolescent residential (1999). Equity sensitivity and business ethics. J o u r
psychiatric population. R e s id e n tia l T rea tm en t fo r n a l o f O c c u p a tio n a l a n d O rg a n iz a tio n a l P sy ch o l
C h ild ren a n d Y outh, 19, 3 9 - 5 7 . og y , 7 2, 5 3 9 - 5 6 0 .
7. M achiavellian ism 107
Musser, S. J . , & Orke, K. A. (1992). Ethical value sys Machiavellianism and economic opportunism. J o u r
tems: A typology. J o u r n a l o f A p p lie d B eh a v io r a l n a l o f A p p lie d S o c ia l P sy ch olog y , 37, 1 1 8 1 - 1 1 9 0 .
S c ien c e, 2 8 , 3 4 8 - 3 6 2 . Schmitt, D. P. (2004). The Big Five related to risky sex
Nathanson, C ., & Paulhus, D. L. ( 2 0 0 6 , June). B ey o n d ual behaviour across 10 world regions: Differential
fo r g iv e n es s: D issectin g th e s e q u e n c e o f r e a c tio n s to personality associations of sexual promiscuity and
in te r p e r s o n a l tran sg ressio n s. Poster presented at relationship infidelity. E u r o p ea n J o u r n a l o f P erso n
the meeting of the Association for Psychological ality, 18, 3 0 1 - 3 1 9 .
Science, New York. Shen, D., & Dickenson, M . A. (2001). Consumers’ ac
Nathanson, C., Paulhus, D. L., & Williams, K. M. ceptance of unethical clothing consumption activi
( 2 0 06 ). Predictors of a behavioral measure of scho ties: Influence of cultural identification, ethnicity,
lastic cheating: Personality, and competence, but and Machiavellianism. C loth in g a n d T ex tiles R e
not demographics. C o n te m p o r a r y E d u c a tio n a l P sy se a r c h J o u r n a l, 19, 7 6 - 8 7 .
c h o lo g y , 3 1 , 9 7 - 1 2 2 . Shepperd, J . A., & Socherman, R. E. (1997). On the
Newcomb, A. F., Bukowski, W. M ., & Pattee, L. manipulative behavior of low Machiavellians:
(1993). Children’s peer relations: A meta-analytic Feigning incompetence to “sandbag” an opponent.
review of popular, rejected, neglected, controver J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 73,
sial, and average sociometric status. P sy ch o lo g ica l 1448-1459.
B u lletin , 113 , 9 9 - 1 2 8 . Sherry, S. B., Hewitt, P. L., Besser, A., Flett, G. L.,
O ’Connor, E. M ., & Simms, C. M . (1990). Self & Klein, C. ( 2 0 06 ). Machiavellianism, trait per
revelation as manipulation: The effects of sex and fectionism, and perfectionistic self-presentation.
Machiavellianism on self-disclosure. S o c ia l B e h a v P erso n a lity a n d In d iv id u a l D iffer en c es , 4 0 , 8 2 9 -
io r a n d P erson ality , 18, 9 5 - 1 0 0 . 839.
O ’Connor, W. E., & Morrison, T. G. (2001). A co m Shultz, C. J., II. (1993). Situational and dispositional
parison of situational and dispositional predictors predictors of performance: A test of the hypoth
of perceptions of organizational politics. J o u r n a l o f esized Machiavellianism x structure interaction
P sy ch olog y , 13 5 , 3 0 1 - 3 1 2 . among salespersons. J o u r n a l o f A p p lied S o c ia l P sy
Ojha, FI. (200 7). Parent-child interaction and M ac hia ch o lo g y , 2 3 , 4 7 8 - 4 9 8 .
vellian orientation. J o u r n a l o f th e In d ia n A c a d e m y Simon, L. J., Francis, P. I,., & Lombardo, J. P. (1990).
o f A p p lie d P sy ch olog y , 3 3, 2 8 5 - 2 8 9 . Sex, sex-role, and Machiavellianism as correlates of
Paal, T., & Bereczkei, T. (20 0 7). Adult theory of mind, decoding ability. P erce p tu a l a n d M o to r S kills, 71,
cooperation, Machiavellianism: The effect of min- 243-247.
dreading on social relations. P erso n a lity a n d In d i Simonton, D. K. (1986). Presidential personality: Bio
v id u a l D iffe r e n c e s , 4 3 , 5 4 1 - 5 5 1 . graphical use of the Gough Adjective Check List.
Paulhus, D. L. (1983). Sphere-specific measures of per J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 51,
ceived control. J o u r n a l o f P erso n a lity a n d S o cia l 1 4 9 -1 6 0.
P sy ch olog y , 4 4 , 1 2 5 3 - 1 2 6 5 . Singhapakdi, A., & Vitell, S. J. (1991). Selected fac
Paulhus, D. L., & Williams, K. M. (2002). The Dark tors influencing marketers’ deontological norms.
Triad of personality: Narcissism, Machiavellianism, J o u r n a l o f th e A c a d e m y o f M arketin g S cien ce, 19,
and psychopathy. Jo u r n a l o f R e se a rc h in P ers o n a l 37-42.
ity, 3 6 , 5 5 6 - 5 6 3 . ' Siu, W. S., & Tam, K. C. (1995). Machiavellianism and
Raman aiah, N. V., Byravan, A., & Detwiler, F. R. J. Chinese banking executives in Hong Kong. I n te r n a
(1994). Revised N E O Personality Inventory profiles t io n a l J o u r n a l o f B a n k M arketin g , 13, 15 -2 1 .
of Machiavellian and non-Machiavellian people. Snyder, M. (1974). Self-monitoring of expressive be
P s y c h o lo g ic a l R e p o rts , 75, 9 3 7 - 9 3 8 . havior. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l
Rawwas, M . Y. A., & Singhapakdi, A. (1998). Do co n og y , 3 0 , 5 2 6 - 5 3 7 .
sumers’ ethical beliefs vary with age?: A substantia Sparks, J. R. (1994). Machiavellianism and personal
tion of Kohlberg’s typology in marketing. J o u r n a l success in marketing: The moderating role of lati
o f M arketin g T h e o r y a n d P ractice, 6, 2 6 - 3 8 . tude for improvisation. J o u r n a l o f th e A c a d e m y o f
Repacholi, B., & Slaughter, V. (2003). In d iv id u a l d if M a rk etin g S cien ce, 2 2 , 3 9 3 - 4 0 0 .
fe r e n c e s in th e o r y o f m in d. New York: Psychology Stewart, A. F.., &c Stewart, F^. A. (20 0 6). The prefer
Press. ence to excel and its relationship to selected person
Repacholi, B., Slaughter, V., Pritchard, M ., & Gibbs, ality variables. J o u r n a l o f In d iv id u a l P sy ch olog y ,
V. (2003). Theory of mind, Machiavellianism, and 62, 27 0 -284.
social functioning in childhood. In B. Repacholi Suman, B. )., Singh, S., & Ashok, K. (200 0 ). M achia
& V. Slaughter (Eds.), In d iv id u a l d iffe r e n c e s in vellians: Their manifest need patterns. P sy ch o-
th e o r y o f m in d (pp. 6 7 - 9 7 ) . New York: Psychology L in g u a , 3 0 , 2 1 - 2 4 .
Press. Sun-tzu (1998). T h e a r t o f war. (Y. Shilling & J. J. I..
Ricks, J ., & Fraedrich, J . (1999). The paradox of M a Duvvendak, Trans.). New York: Wordsworth.
chiavellianism: Machiavellianism may make for Sutton, J., & Keogh, F’ . (2001). Components of M a
productive sales but poor management reviews. chiavellian beliefs in children: Relationships with
J o u r n a l o f B u sin ess E th ics, 2 0 , 1 9 7 - 2 0 5 . personality. P erso n a lity a n d In d iv id u a l D iffe r e n c
Ryckman, R. M ., Thornton, B., &c Butler, J. C. (1994). es, 3 0 , 137-148.
Personality correlates of the hypercompetitive at Trapnell, P. D., &C Paulhus, D. L. (in press). Agentic
titude scale: Validity tests of Horney’s theory of and communal values. J o u r n a l o f P erson ality As-
neurosis. J o u r n a l o f P erso n a lity A ssessm en t, 62 , sessm en t.
84-94. Valentine, S., & Fleischman, G. (2003). The impact of
Sakalaki, M ., Richardson, C., & Thepaut, Y. (200 7). self-esteem, Machiavellianism, and social capital
108 II. I N T E R P E R S O N A L D IS P O S IT IO N S
Gender Identity
W en dy W ood
A l ic e H . E a g l y
hat individual differences in gender & Eagly, 2 0 0 8 ; Eagly, Wood, & Diekman,
W are important to study? Because gen
der refers to the cultural meanings ascribed
2 0 0 0 ; Eagly, Wood, & Johannesen-Schmidt,
2004). These typical role occupancies pro
to male and female social categories in soci duce gender roles, which are defined as so
eties, psychologists have focused on whether cially shared expectations for men’s and
individuals define themselves in terms of women’s behavior. As gender roles are ac
these cultural meanings. We use the term cepted by individuals, they are internalized
gender identity to refer to these masculine into their self-concepts. People differ in the
and feminine self-definitions. Individuals extent to which they accept these norma
differ in gender identity within each sex, tive expectations about men and women as
and men and women differ on the average. personally self-defining and thereby differ in
Gender identity is only one of many possible the extent to which they incorporate cultural
social identities, with each identity repre gender into their personal identities.
senting one’s psychological relationship to a The content of gender roles reflects the
particular social category in which one has characteristics that facilitate sex-typical
membership (e.g., race, social class, religion; tasks in a given society. To the extent that
see Frable, 1997; Sherif, 1982). women more than men occupy roles that
Psychologists’ conviction that gender iden involve domestic activities and communal
tity is important has given rise to a wide range behavior (e.g., nurturing children, providing
of constructs that represent culturally based service to others), the psychological attri
masculine and feminine self-definitions. In butes that facilitate these role behaviors form
this chapter, we organize these constructs the basis for shared gender-role expectations
in terms of three facets of masculinity and for women and for feminine gender identity.
femininity: representations of oneself as (1) To the extent that men more than women
possessing gender-typed personality traits occupy roles that involve economically pro
and interests, (2) having male-typical ver ductive activities and directive behavior
sus female-typical relationships to others, (e.g., resource acquisition, managing large
and (3) being a member of the category of organizations), the psychological attributes
women or men, as that category is defined that facilitate these role behaviors form the
within a given society. basis for shared gender-role expectations for
From a social-role perspective, gender men and for masculine gender identity. Yet,
identity reflects the different placement of gender roles have origins in multiple biologi
men and women into societal roles (Diekman cal and cultural factors (see Wood & Eagly,
109
110 II. I N T E R P E R S O N A L D IS P O S IT IO N S
gender identity that is relevant to the behav By tailoring measures of gender identity to
ior under investigation. the domain of interest, researchers increase
This issue of matching gender identity their chances of finding meaningful effects
constructs to relevant behaviors is crucial, of gender identification.
above and beyond the more mundane issue Beyond the recommendation to match con
faced in all individual-difference research of tent domains across identity measures and
choosing measures to maximize reliability predicted behaviors, the compatibility prin
and validity (e.g., M arsh, 1987). To appro ciple implies that good prediction follows
priately link gender identity to behavior, it from assessing the identity measure and the
is necessary to understand some elementary behavioral measure at the same level of gen
principles of the prediction of behavior from erality. For example, if self-assessments on
psychological dispositions. the culturally masculine trait of assertiveness
were used to predict behavior, the behavioral
measure ideally would include not merely a
T h e Principle o f Compatibility
single assertive behavior such as speaking up
The choice of measures of gender identity at meetings but rather a wide range of asser
should be guided by the principle o f co m tive behaviors selected from a wide range of
patibility (Ajzen, 2 0 0 5 ; Eagly & Chaiken, settings. A single behavior such as speaking
1993), which stipulates that identity mea up at a meeting is an imperfect representa
sures are more likely to predict responses if tion of assertiveness because there are many
they are in the same content domain as the reasons why an otherwise assertive person
measure. This principle, initially developed might not engage in this behavior, especially
for enhancing the prediction of behaviors at a particular meeting. Because individual
from attitudes (Ajzen & Fishbein, 1977), behaviors are multiply determined, correla
also is important for predicting behaviors tions between measures of assertiveness in
from personality traits (e.g., Epstein, 1980). general and any single behavior are gener
The key insight for successful prediction of ally low.
behavior from attitudes, personality, or any When research matches dispositional and
other disposition is that prediction is en behavioral measures at the same level of gen
hanced by matching the content of the be erality, substantial correlations can emerge
havioral measure to the content of the dis between the two measures (Epstein, 1980).
positional measure. Therefore, measures of However, because most research on gender
gender identity would successfully predict identity has used general measures of identi
behaviors in the domain of the disposition. ty and related these to only one or a few spe
Based on the compatibility principle, cific behaviors of interest, most correlations
gender identity measures that assess self- in the literature we review are relatively low.
reported masculine and feminine personal Correlations would be higher if researchers
ity traits would best predict corresponding had related general identity measures to ag
behaviors, such as communal behaviors of gregated indexes of relevant behaviors. Al
taking care of others or agentic behaviors of ternatively, researchers could improve pre
assertiveness. Identity measures that assess diction by designing more specific measures
gender-typed vocations and interests would of identity. For example, a narrowly defined
best predict related behaviors, such as gen- feminine quality such as belief in one’s so
der-typed hobbies and occupations. Identity cial sensitivity could be related to relatively
measures that assess sex-typical social re specific responses, such as the ability to infer
lational preferences would best predict the others’ feelings in a variety of settings.
kinds of relationships that women and men
form with others, including their embedded
D irect and Indirect M easures
ness in dyadic pairs or larger hierarchical
structures in which they relate to multiple Although most measures of gender iden
others. Identity measures that assess mem tity involve direct self-ratings on relevant
bership in the socially defined categories of response scales, identity can be tapped
men and women would best predict group- through measures that are much less direct.
related judgments such as preference for one’s Dual-process theories in social psychology
own sex and prejudice against the other sex. provide a way to understand differences be
112 II. I N T E R P E R S O N A L D IS P O S IT IO N S
tween these measurement approaches (see (e.g., McGuire & Padawer-Singer, 1976)
Chaiken & Trope, 1999; Smith & DeCoster, and content analyses of self-descriptive pho
2000 ). Direct rating scales tap propositional tographs (e.g., Clancy & Dollinger, 1993).
knowledge about oneself through verbal These and other indirect measures are de
ized judgments of gendered identity: “I am signed to estimate gender identity without
warm” or “I identify with women.” Such a direct verbal report and often without
measures require that people have some participants’ awareness that this identity is
conscious awareness of their gender identity being assessed.
and are able and willing to report on it using There are several reasons to expect some
the given scale format. In contrast, indirect divergence between direct and indirect mea
measures tap spontaneous aspects of gender sures of gender identity. One is that direct
identity that may or may not be accessible ratings are farther dow nstream in the pro
to conscious, verbal description (see Smith cesses of judgment and thus subject to more
& DeCoster, 20 0 0 ). Furthermore, indirect deliberation than the more spontaneous, au
measures may rely on associative processing tomatic associations tapped by indirect mea
systems that reflect the cumulation of expe sures (Fazio & Olson, 2003). As a result,
riences over time. responses to direct measures can be more
Indirect measures of gender identity often influenced by pressures to appear socially
assess respondents’ reaction times in making desirable than are responses to indirect mea
identity-relevant judgments. Such reaction sures (Greenwald, Poehlman, Uhlmann, &
times can reflect the strength of associative Banaji, in press). Another reason for diver
connections between oneself and culturally gence is that indirect and direct measures
feminine and masculine traits or between may not represent the same content. This
oneself and male and female groups. Stron occurs when researchers use different bases
ger identity, represented by closer associa to select direct and indirect measures by, for
tions between oneself and gender concepts, example, deriving one measure but not the
should produce faster reactions. For ex other from gender stereotypes. In the fol
ample, in a gender prim ing task, exposure lowing sections, we consider these and other
to the prime of m e or they is followed by issues in analyzing the merits of direct and
the participant classifying a gendered word indirect measures of our three facets of gen
(e.g., lady, fishing rod-, van Well, Kolk, & der identity.
Oei, 2 0 0 7 ) into categories of person versus
object. People with a strong gender identity
presumably have gender primed by the word Individual Differences
m e and therefore are relatively fast in mak in Self-D escribed Personal Traits
ing such categorizations. Another indirect and Attributes
measure, the Implicit Association Test (IAT),
G end er Identity as Bipolar M asculinity—
assesses the strength of association between
F em in in ity in Heterogeneous D om ains
self and aspects of gender identity through
the speed of responding when categorizing Modern measures of gender identity origi
the self (vs. others) as masculine or feminine nated in Terman and M iles’s (1936) test of
(Greenwald & Banaji, 1995). The resulting masculinity and femininity. This measure
IAT scores can form a bipolar dimension, is composed of items that elicited m axi
reflecting the ease of associating masculine mally different responses from women and
versus feminine traits with the self compared men. The resulting collection of items is a
with others. IAT scores can also form unipo heterogeneous lot that includes word asso
lar scales. According to meta-analytic esti ciations, associations to inkblots, interest
mates, if direct self-ratings and IAT indirect items, introversion-extraversion items, and
measures are assessed in compatible ways self-judgments of overall masculinity and
so that both, for example, compare mascu femininity. For example, femininity scores
line and feminine gender identity, then cor increased with liking “nursing,” “babies,”
relations of around .30 consistently emerge and “charades,” whereas masculinity in
(Hofmann, Gawronski, Gschwendner, Le, creased with disliking these. This method of
& Schmitt, 2005). Additional indirect mea item selection and scoring placed masculin
sures include open-ended self-descriptions ity and femininity as two ends of a single bi
8. G en der Id en tity 113
polar continuum. Other psychologists then construct, were not very strongly related to
followed this approach of selecting test items one another. Constantinople’s critique and
that strongly differentiated between women her accusation that masculinity and femi
and men and labeling the resulting scales as ninity are “among the muddiest concepts in
measures of masculinity and femininity (see the psychologist’s vocabulary” (p. 390) cata
reviews by Lippa, 2 0 0 1 , 2005). lyzed development of a different framework
This tradition has been continued more for assessing gender identity.
recently in measures of identification that
tap female-typical or male-typical interests.
G end er Identity as Separate M asculine and
Favoring this approach, Lippa (1991; Lippa
F em in in e D im ensions o f Personality Traits
& Connelly, 1990) developed a method of
gender diagnosticity in which women and In the new framework spurred by Constan
men rate their preferences for occupations, tinople’s (1973) and others’ critiques, mas
hobbies, and everyday activities. These rat culinity and femininity appear as two sepa
ings then allow the computation of the pat rate dimensions. Drawing scale items from
tern of preferences that maximally discrimi the cultural stereotypes of the personality
nates between the male and female raters (in traits of women and men, Bern (1974) pro
terms of a weighted combination of items posed the Bern Sex Role Inventory (BSRI),
that constitutes a discrim inant function). which represents masculinity and femininity
Respondents’ gender identities are then de as separate, orthogonal dimensions. These
termined by comparing their scores with items were selected because the personality
this male-typical versus female-typical pat traits they represented were more stereotypi
tern of preferences. cal of one sex than the other and more fa
Lippa’s gender diagnosticity measure of vorably evaluated in that sex. The measure
gender identity, like the Terman and Miles thus assesses self-defined personality traits
(1936) measure, is based on items that that are either masculine (e.g., self-reliance,
maximally discriminate between women’s assertiveness, forcefulness) or feminine (e.g.,
and men’s self-reports. However, it differs affection, sympathy, warmth). Among the
in its narrower focus on interests and in its four quadrants that resulted, two defined
calibration of what distinguishes the sexes respondents considered sex typed by Bern:
within each sample of respondents. This (1) those high on masculinity and low on
method of computing a gender discriminant femininity, labeled m asculine , and (2) those
function has been applied to other types of high on femininity and low on masculinity,
items as well (e.g., Burke & Tully, 1977). labeled feminine. The two remaining quad
As would be expected from the compatibil rants defined respondents considered not
ity principle, Lippa’s (1991) measure relates sex typed: (1) those high on both masculin
especially well to occupational preferences, ity and femininity, labeled androgynous,
with more masculine respondents preferring and (2) those low on both masculinity and
occupations that deal mainly with things femininity, labeled undifferentiated. This
and more feminine respondents preferring two-dimensional scheme decoupled gender
occupations that deal mainly with people identity from its earlier bipolar framing and
(Lippa, 1998, 2005). thereby represented identities in all combi
Constantinople (1973) provided an early nations of high and low masculinity and
critique of these kinds of measures of mas femininity.
culinity and femininity. She complained In a related project, Spence and Helm-
about the empirically derived selection of reich (Spence & Helmreich, 1978; Spence,
items, especially the motley types of con Helmreich, & Stapp, 1974) developed the
tent in Terman and M iles’s (1936) measure Personal Attributes Questionnaire (PAQ),
and similar broad-spectrum measures. She which also defined gender identity in terms
demonstrated that statistical analyses of of two separate dimensions of personality
such items often revealed multiple dimen attributes that are stereotypical of women
sions, not a single bipolar dimension. An or of men. Spence argued that the PAQ and
other criticism was that different versions the BSRI are measures not of culturally de
of masculinity-femininity scales, which fined masculinity and femininity but of con
presumably assessed the same psychological stellations of socially desirable personality
114 II. IN T E R P E R S O N A L D IS P O S IT IO N S
traits defined by either instrumentality (e.g., sures such as the BSRI and PAQ that rep
decisiveness, competitiveness, activity) or resent agentic and communal traits? These
expressiveness (e.g., kindness, helpfulness, measures assess only one aspect of the
understanding). Alternatively, in the termi gender-related qualities that form the basis
nology introduced by Bakan (1966) and fa of people’s sense of maleness or femaleness
vored by many gender researchers, the two and therefore should relate to behaviors only
dimensions of the BSRI and PAQ gained the within the relevant domain. Empirical sup
labels of agency and com m union. Although port comes from a meta-analysis by Taylor
some researchers have found that the items and Hall (1982), in which people who were
that make up the BSRI and the PAQ scales high on the masculine dimension of the BSRI
are not necessarily internally consistent (e.g., or PAQ engaged in more agentic behaviors
M arsh, 1987), these two-dimensional gen than those who were low, and people who
der identity measures have remained very were high on the feminine dimension engaged
popular in research. in more communal behaviors that those who
Subsequent elaborations of the PAQ in were low. Furthermore, prediction from the
cluded scales designed to capture nega BSRI or PAQ to behaviors in domains other
tive aspects of instrumentality (e.g., being than communion and agency were generally
domineering, overbearing) and expressive weak and inconsistent (Spence & Buckner,
ness (e.g., being whiny, passive) (Helmre- 1995). Despite widespread use of these mea
ich, Spence, & Wilhelm, 1981). Addition sures in psychological research, investigators
ally, Athenstaedt (2003) extended the two have only occasionally recognized that their
dimensions to include items assessing their predictive power is circumscribed to com
behavioral expressions. munal and agentic behaviors. Following the
These two-dimensional personality-based principle of compatibility, we expect in ad
measures of gender identity may seem anom dition that these identity measures will have
alous in view of modern personality theory, maximum impact when a study’s behavioral
which has converged on a five-dimensional measures are at the same level of generality.
organization of traits known as the Big Five Thus identity measures defined in terms of
(Wiggins, 1996): Extraversion, A greeable broadly formulated personality traits, such
ness, Conscientiousness, N euroticism , and as the BSRI and the PAQ, will more effec
O penness to experience. Although the BSRI tively predict aggregated indexes of multiple
and PAQ scales correlate with some of the Big agentic and communal behaviors than any
Five traits, more fine-grained analyses have single behavior.
revealed that each Big Five trait is made up
of separate components, and sex differences
Indirect M easures o f Personal Traits
are not always consistent in magnitude or
and Attributes
direction across the components that make
up the broader traits (Costa, Terracciano, Indirect measures of gender identity as
& M cCrae, 2001). Given such complexities, sess more automatic and spontaneous self
agency and communion are not readily re descriptions. The IAT is the most popular
configured in terms of the Big Five. Instead, indirect measure of traits and attributes.
the agency-communion scheme provides an For example, Greenwald and Farnham’s
alternative organization of personality traits (2000) respondents categorized self-related
to the Big Five, and this two-dimensional pronouns (me, I) or other-related pronouns
organization is particularly useful in study (them, it) with communal (warm, tender) or
ing gender because of its match to gender agentic (competitive, aggressive) attributes.
stereotypes and roles. Attesting to the value The resulting IAT scores were formed into
of this two-dimensional scheme, social- a bipolar dimension of masculinity versus
psychological researchers on impression for femininity, reflecting the ease of associating
mation and stereotyping have often favored masculine versus feminine traits with one
two dimensions that are construed in terms self compared with others. As would be ex
of some version of these agentic and com pected from the compatibility principle, this
munal families of traits (e.g., Judd, James- bipolar measure of gender identity was posi
Hawkins, Yzerbyt, & Kashima, 2005). tively related to the PAQ and BSRI scales
Based on the compatibility principle, what when they were computed as bipolar scales.
behaviors are likely to be predicted by mea Nonetheless, the IAT also can potentially
8. G en der Id en tity 115
be scored to reflect separate unipolar mas sense of self that highlights individuals’
culinity and femininity dimensions, and in unique abilities and attributes (Markus &
this form it should be associated with direct Kitayama, 1991; but see Oyserman, Coon,
measures of the compatible masculinity or & Kemmelmeier, 2002).
femininity subscale. It is unknown whether In a landmark article summarizing re
the indirect and direct forms of such mas search indicating that men and women dif
culinity and femininity measures predict be fer in such self-construals, Cross and M ad
haviors differently. son (1997) argued that women describe
themselves more in terms of relationships
with others, whereas men describe them
Individual Differences selves more in terms of separateness from
in Interdependent Self-Construal others. For example, women’s greater in
terdependence is evident in their sensitivity
Gender identity also includes beliefs about to others’ nonverbal cues, emotional empa
oneself in social relationships, often labeled thy, and capacity to adopt others’ cognitive
self-construal. This aspect of gender iden perspectives. Furthermore, women’s self
tity reflects the social contexts within which esteem tends to depend on their ability to
men and women carry out sex-typical activi maintain relationships with others, whereas
ties in a society. To the extent that women men’s depends more on maintaining inde
more than men occupy roles that encourage pendence from others. In addition, women
close, interdependent relations with others, attend more to close relationships and like
feminine gender roles are likely to include to discuss them with others, whereas men
self-construals that emphasize connections prefer to discuss less personal topics such as
to intimate others. Comparably, to the ex sports and politics. Cross and Madson inter
tent that men more than women in a society preted these (and other) findings as evidence
occupy roles that encourage independent ac for women’s greater interdependence and
tion and/or action within larger collectives, men’s greater independence (see also review
masculine gender roles are likely to include of work on self-construals by Cross et al.,
self-construals of autonomy from others Chapter 35, this volume).
and/or positions within larger collectives. Subsequent work challenged this charac
The interdependent aspects of gender iden terization of men as less dependent on so
tity focus not on individuals’ possession of cial relations than women. Reasoning that
personality attributes such as communion or all people have a need to belong, Baumeister
agency but on the ways that men and women and Sommer (1997) argued that the sexes
define themselves in relationships with inti express this dependency differently, with
mate others and with social groups. women more likely to form close relation
Initial work on this aspect of gender iden ships with intimate others and men more
tity focused on the degree to which men likely to form relationships within larger col
and women regard themselves as separate lectives and groups in which they can assert
from or connected to other people. That power and dominance (see also Baumeister
is, women were thought to have a relatively & Leary, 1995). Thus women’s sense of their
interdependent self-definition, in which val interdependence with others is relational, or
ued and important others are included in oriented to committed, close relations with
self-representations, and men to have a more others, whereas men’s sense of interdepen
independent self-definition, in which the dence is collective, or oriented toward larger
self is autonomous and distinct from others social groups (Gabriel & Gardner, 1999;
(Cross & Madson, 1997; Josephs, M arkus, Gardner & Gabriel, 2004).
& Tafarodi, 1992; see Cross, Hardin, & Relational and collective gender identities
Gercek Swing, Chapter 35, this volume). Sex align with Brewer and Gardner’s (1996) anal
differences would thereby align with cross- ysis of the forms of interdependent selves. Ac
cultural differences in self-construals in the cording to these researchers, whether people
form of East Asian cultures’ promotion of an construe themselves in terms of their inter
interdependent sense of self that highlights personal relationships or in terms of larger,
relationships, group memberships, and har more interpersonal collectives determines
mony with others, as opposed to Western various aspects of self-functioning (e.g., sa
ized cultures’ promotion of an independent lient components of self, primary social mo
116 II. I N T E R P E R S O N A L D IS P O S IT IO N S
tives). In addition, relational and collective investigation across five cultures revealed no
interdependence represent not only charac sex differences in self-reported individual
teristic ways of interpreting the self in rela ism as reflected in assertiveness and acting
tion to others but also temporarily activated independently in group contexts. Similarly,
states in which circumstances prime the ap Nario-Redmond, Biernat, Eidelman, and
propriate identities and lead to perceptions Palenskie (2004) found an absence of sex
of self as interlinked in relationships with differences in U.S. college students’ ratings
intimate others or in larger groups. of the importance of personal identity items
Evidence that men and women differ in pertaining to independence (e.g., rebellious
their chronic levels of interdependent self- ness, creativity). Also, Kashima and Hardie’s
construal comes from research that has com (2000) relational, individual, and collective
pared relational and collective forms of de self-aspects scale revealed no sex differences
pendence. Arguing that Cross and Madson’s among Australian college students in the
(1997) review had focused primarily on the prominence of individual aspects of the self.
relational aspect of interdependence, Gabriel In line with sex differences in the form
and Gardner (1999) provided a variety of ev of interdependence, women typically report
idence that men are more collectively inter higher relational dependence than men.
dependent than women. For example, when For example, across U.S. college student
asked to give spontaneous self-descriptions, samples, women consistently scored higher
men were more likely than women to list than men on a scale designed to assess Re-
group memberships (e.g., fraternity mem lationally Interdependent Self-Construal
ber, black man), whereas women were more (RISC scale) with items such as, “My close
likely than men to list specific close relation relationships are an important reflection of
ships (e.g., friend, happily married; Gabriel who I am” (Cross, Bacon, &c Morris, 2 0 0 0 ;
& Gardner, 1999, Study 1). In addition, Gabriel & Gardner, 1999; Gore, Cross, &
when asked to recall and describe a happy or M orris, 2006). Also, across five cultures,
sad emotional event, men were more likely Kashima and colleagues (1995) found that
than women to report an experience in the women scored consistently higher than men
context of a collective (e.g., with a fraternity on measures of closeness in emotional rela
or sorority), whereas women were more like tions with others.
ly than men to report an experience with a Direct measures have provided less con
close other (e.g., friend, family member; G a clusive evidence of sex differences in collec
briel & Gardner, 1999, Study 3). tive interdependence. Gabriel and Gardner
(1999, Study 2) found college men more col
lectively dependent than women on a version
D irect M easures
of the RISC designed to tap collective iden
o f Interdependent Self-C onstrual
tity (e.g., “The groups I belong to are an im
Direct tests of the independent, relational, portant reflection of who I am”). However,
and collective self-concepts of women and other research has not obtained analogous
men usually elicit self-reports on scales as findings (e.g., Kashima & Hardie, 2 0 0 0 ), in
sessing the importance or descriptiveness cluding studies in which respondents rated
of each of these self aspects. The research the importance of various group identi
published thus far suggests that the sexes ties to their own self-definitions (Luhtanen
differ primarily in the extent to which they & Crocker, 1992; Nario-Redmond et al.,
construe the self in relations with close oth 2004). This apparent inconsistency in find
ers. Accordingly, relational interdependence ings might be resolved through a compre
holds most promise as a facet of gender iden hensive meta-analytic review of the relevant
tity. research findings.
Our review of the literature suggested that
men and women do not differ consistently
Indirect M easures o f Interdependence
in their overall levels of independence. Thus
few sex differences have been reported on Indirect measures that assess whether peo
measures assessing self-descriptions as inde ple spontaneously mention individual others
pendent as opposed to interdependent. For or collectives when describing themselves
example, Kashima and colleagues’ (1995) also attest to women’s greater relational
8. G ender Id en tity 117
categories. Alternatively, such measures can Tudor, & Nelson, 1991). It consists of a set
refer to ideal men and women and thereby of Venn-like diagrams with varying degrees
reflect identification with injunctive catego of overlap between two circles. When it is
ries. In addition, although some measures used to assess strength of collective gender
separately assess identification with each identity, respondents pick the diagram that
gender, measures often assess identification best depicts the extent of overlap between
with a gender ingroup in opposition to the themselves and their gender group.
gender outgroup (Turner et al., 1987). Another spontaneous measure is whether
people mention gender in response to an
open-ended request to list self-descriptive
Direct Measures o f Collective Gender Identity
attributes (e.g., McCrae & Costa, 1988;
A popular measure of identification with de McGuire 8c Padawer-Singer, 1976). The
scriptive group categories is the im portance spontaneous mention of gender categories
o f identity subscale of Luhtanen and Crock might reflect the chronic salience of gender,
er’s (1992) collective self-esteem scale. When as well as situationally induced salience, as
adapted to gender groups, this measure con indicated by the finding that children in the
sists of four items that assess the importance minority sex in a classroom mentioned their
of being a woman or a man to one’s self- sex more often in describing their physical
image. Other measures elicit self-reports of selves (McGuire & Padawer-Singer, 1976).
how typical respondents are in their gender
group (e.g., Eagan & Perry, 2001).
Collective Gender Identification
To capture the injunctive nature of gender
Predicts Behavior
categories— that is, people’s beliefs about
what is desirable for the sexes— measures Given the logic of the compatibility princi
can specify ideal or desired gender catego ple, collective gender identity should predict
ries. For example, to assess gender ideals, behavior as a group member, such as valuing
Wood and colleagues (1997) assessed how one’s group over other groups (Tajfel, 1982).
important it was for respondents to be simi In particular, evaluative forms of group
lar to the ideal man or woman and to reject identity— that is, identification with what is
the other sex ideal. good about one’s gender group compatible
with ingroup bias. Illustrating this effect,
Wood and colleagues’ (1997) participants
Indirect Measures o f Collective Gender Identity
rated their similarity to the societal ideal for
Reaction-time measures assess gender iden their sex (e.g., the ideal woman) and their
tity indirectly through the speed with which dissimilarity to the ideal for the other sex.
participants associate self (vs. others) with To the extent that participants had a stron
gender categories. The IAT assesses strength ger identification with their own (vs. other)
of gender identity through reaction times to gender ideal, they experienced a boost in self
differentiate self words (e.g., me) from non esteem when imagining themselves acting in
self words (e.g., other) when each is paired gendered ways. Evidence for ingroup bias
with words indicative of gender groups also has been found on measures that mini
(e.g., he, fem ale) (Aidman & Carroll, 2 0 0 3 ; mize the social desirability concerns that can
Greenwald et al., in press). With lexical- limit demonstrations of outgroup prejudice,
decision measures, participants are primed as shown by meta-analytic evidence that col
or not primed with self-constructs, and then lective identity predicted ingroup bias better
their reaction times for recognizing gender- with indirect, IAT measures than with direct
related words are assessed (e.g., w om an, measures (Greenwald et al., in press).
footb all ; van Well et al., 20 0 7 ). As expected Another group behavior associated with
based on the principle of compatibility, the collective identity is self-stereotyping, or the
IAT and lexical-decision measures of collec ascription of group characteristics to the self
tive gender identity are positively correlated, (Turner et al., 1987). Given the logic of the
r(43) = .48 (van Well et al., 2007). compatibility principle, broad measures of
A graphic measure of spontaneous gender identification with a gender group would not
identity assesses the extent to which peo necessarily relate to specific behaviors such
ple include others in the self (Aron, Aron, as ascription of particular gender-typed at
8. G en der Id en tity 119
tributes to oneself. In illustration, Kiefer and taking, and aggression that may injure oth
Sekaquaptewa (2007) found that women ers (Booth, Granger, Mazur, & Kivlighan,
with a strong gender identification were 2006). For example, testosterone levels rise
no more likely than those with a weak one in anticipation of athletic and other compe
to ascribe the stereotypic attribute of poor tition and in response to insults. Higher lev
math performance to themselves. However, els of oxytocin are associated with behaviors
the predicted self-stereotyping was obtained that produce parental bonding, nurturance,
among women with a strong gender-group and intimacy, especially in women (Camp
identity who also believed that math is mas bell, 2008). For example, oxytocin levels rise
culine and not feminine. Thus gender self- in women during childbirth and in response
stereotyping on specific attributes can be to massage and sexual contact. These hor
detected in studies that assess the specific monal processes do not work in isolation but
attributes that people ascribe to gender cat instead combine with gender identities to fa
egories. cilitate performance of relevant behaviors.
Gender self-stereotyping also is found Thus, testosterone is especially relevant for
when measures of gender identity are tai people with masculine gender identities that
lored to the specific domain of interest. For lead them to experience social interactions
example, W itt and Wood (2008) assessed as dominance contests. O xytocin and other
gender identity by asking respondents how neurochemical processes involved in bond
important it was that they acted like a typi ing are especially relevant for people with
cal man or a typical woman with respect feminine identities that lead them to define
to romantic relations (e.g., dating, flirting). social interactions as involving bonding and
Experience-sampling methods revealed that affiliation with close others.
highly identified students interacted with Gender identity, as a component of the
peers of the other sex in typically feminine self-concept, also informs self-regulation.
(or masculine) ways. In general, collective When people self-regulate, they exercise con
gender identity relates to self-stereotyping trol over their behavior to bring the self into
on specific attributes primarily when the line with valued standards. Gender identities
attributes are ones that respondents believe serve as self-regulatory standards when they
characterize their gender group or when descriptively specify how a person of one’s
gender identity is assessed with respect to gender is expected to act or injunctively
the specific domain of interest. specify how a person of one’s gender ideally
would act.
According to self-regulatory theories,
Gender Identity Guides Responding people guide their actions toward valued
standards and goals by a matching process
What are the mechanisms by which the often likened to a cybernetic feedback loop
three types of gender identity we have con (e.g., Carver & Scheier, 2008). With this
sidered influence people’s responses? Gender feedback loop, the regulatory system moni
identity guides behavior through a set of bi tors the extent to which current behavior
ological and psychological mechanisms (see matches self-standards. When people’s be
Wood & Eagly, in press). Biological process havior successfully matches their gender
es include hormonal fluctuations that act as identity, they experience positive emotion
chemical signals that promote actions in line and increased self-esteem; when their behav
with gender identity. Social processes also ior deviates from their gender identity, they
are implicated, given that people use gender experience negative emotion and decreased
identities as self-standards against which to esteem. Therefore, people with a stronger
regulate their behavior. gender identity— of whatever type— would
Biological processes work to promote use this identity as a standard for their own
gender identities as people selectively recruit behavior and experience more of a boost in
hormones and other neurochemical pro positive affect and self-esteem when vicari
cesses to facilitate performance of relevant ously imagining themselves or actually be
behaviors. Higher levels of testosterone are having in ways that are consistent with this
associated with dominance behaviors, es identity (Diekman & Eagly, 2 0 0 8 ; Wood et
pecially those involved in competition, risk- al., 1997). Wood and her colleagues have
120 II. I N T E R P E R S O N A L D IS P O S IT IO N S
demonstrated this mechanism with respect monal processes include the recruitment
to collective gender identity (e.g., W itt & of testosterone and oxytocin to facilitate
Wood, 2008). performance in line with these identities.
Self-regulatory control proceeds not only In support of this analysis, women high in
through motivational signals of affect and masculinity on the BSR I, who perceived
self-esteem but also through enhanced atten themselves as self-directed, action-oriented,
tion to, processing of, and recall of informa and resourceful, were likely to have higher
tion relevant to gender standards. The infor circulating testosterone (Baucom, Besch, &
mation processing consequences of gender Callahan, 1985). Given the evidence that
identity were a cornerstone of Bern’s (1981) testosterone is recruited in the service of role
theory that sex-typed people have a “gener performance, this pattern is consistent with
alized readiness to process information on the idea that agentic women are sensitive to
the basis of the sex-linked associations” held dominance issues in daily life and recruit
in long-term memory (p. 355). Bern argued this hormone as they assert dominance.
that gender identity provides a kind of lens
for processing information relevant to the
self and gender. However, only inconsistent Origins o f Gender Identity
evidence suggests that the agentic and com
munal forms of gender identity assessed by Gender identity is one of a variety of gender-
Bern’s favored measure (i.e., BSRI) guide related constructs that children develop as
gender-related processing across a range of they mature within their society. Gender
stimuli (see Kite & Deaux, 1986). Based identity emerges through the interaction
on the compatibility principle, evidence for over time of social-cognitive learning and
such processing should emerge mainly when biological processes (see Bussey & Bandura,
measures of identity are in the same domain 1999, 2 0 0 4 ; Ruble & M artin, 1998). These
as the measures of information processing. include learning to label oneself as a boy
In support, people who score higher on re or girl and to understand gender constancy
lational forms of gender identity are more (Kohlberg, 1966). More complex learning
likely to attend to and remember informa is involved in the development of cognitive
tion about others’ relationships (e.g., Cross structures that link the self with gendered
et al., 2002). activities, interests, and personality traits
The capacity to engage in regulatory con (M artin & Ruble, 2004). Furthermore, for
trol of gender-linked behavior appears to identity to be expressed in behavior, females
emerge with maturity (Bussey & Bandura, and males must develop the belief that they
1992). Consistent with such a developmen can engage in such behavior (Bussey & Ban
tal effect, 3-year-old children did not an dura, 1999).
ticipate feeling differently about themselves Gender identity develops within a broader
after playing with same-sex or other-sex societal context in which women and men
toys, whereas older children expressed more cooperate in a division of labor and thereby
positive feelings toward playing with same- fill different social roles (Eagly et al., 2000).
sex toys than other-sex ones. Furthermore, The different placement of men and women
suggesting the developmental trajectory of in society organizes the processes by which
regulatory mechanisms, the older but not boys and girls come to possess a gender iden
the younger children’s anticipatory affective tity and thereby are suited to participate in
reactions predicted their subsequent actual sex-typical social roles. Individual differenc
toy preferences. es in gender identity also reflect the unique
In summary, the three facets of gender experiences that people may have within
identity we address in this chapter all plausi their society.
bly influence responding through a common The division of labor within a society in
set of biosocial mechanisms. Through self- fluences gender identity because it influences
regulatory mechanisms, people carry out the perceived costs and benefits of behaviors
behaviors that are consistent with gender for each sex. Women on average perceive that
identities based on gender-stereotypic traits communal behaviors, people-centered inter
and interests, relational closeness to others, ests and vocations, dyadic relational styles,
or collective male and female groups. H or and a collective identity as a woman are
8. G en der Id en tity 121
especially rewarding. Men on average per documented are the consequences of the
ceive that agentic behaviors, thing-centered high levels of prenatal androgen exposure
interests and vocations, independent and/or of children with congenital adrenal hyper
hierarchical relational styles, and a collec plasia (CAH) disorder. Such exposure yields
tive identity as a man are especially reward increased male-typical toy, playmate, and
ing. Gender identity reflects these average activity preferences among girls, although it
perceived utilities of men and women, along has little systematic effect for boys. N one
with the unique perceptions that each indi theless, such exposure does not appear to
vidual may develop. have consistent effects on broader measures
Men’s and women’s understanding of the of gender identity (Meyer-Bahlburg et al.,
costs and benefits related to gender identities 2 0 0 4 ; although see Hines, Brook, & Con
develops in part through the expectations way, 2004).
held by other people. Others tend to reward The general idea that biological, cogni
behaviors consistent with gender roles be tive, and social factors interact to produce
cause such actions validate shared beliefs individual differences in gender identity has
about women and men and promote social been embraced by evolutionary models of
interaction that is easy to follow and under gender. However, evolutionary models dif
stand (see Wood & Eagly, in press). Others fer in how they envision such interactions.
may respond with rejection to deviating be In evolutionary psychology, contemporary
haviors that challenge gender role expecta gender differences are thought to originate
tions. Through the process of responding from the successful ancestral adaptation to
to gender expectations, people may develop the different reproductive demands faced by
gender identities. Because people often un men and women. For example, men devel
derestimate the influence of others (Vorauer oped attributes of aggressiveness and domi
& Miller, 1997), they might observe their nance because these facilitated mating suc
own expectation-consistent behavior and cess in competition with other men (see Buss,
infer that they possess a corresponding dis 2005). In such models, environmental influ
position— a gender identity. Supporting this ences reflect the contingent expression of
reasoning, research on gender-stereotypical sex-typed evolved dispositions that depends
expectations has yielded some of the stron on the features of current environments that
gest evidence of such behavioral confir match evolutionary environments. Nonethe
mation of others’ expectations (Leander, less, these mechanistic interactions take only
Chartrand, & Wood, 2 0 0 9 ; see overview by limited forms given that gender identity and
Deaux & Lafrance, 1998). Further evidence other psychological dispositions are thought
that children are rewarded for acting in to be preexisting in men’s and women’s biol
ways that confirm gender-role expectations ogy and merely selected by current environ
comes from observations of socialization ments.
practices in nonindustrial societies (Barry, Proposing a more dynamic form of inter
Bacon, & Child, 1957). Socialization re action, Wood and Eagly’s (2 0 0 2 , 2007) bio
search in industrialized nations has pro social model treats the psychological attri
vided less evidence of parents’ differential butes of women and men as emergent given
delivery of rewards and punishments, with the evolved characteristics of the sexes, their
the exception of certain sex-typed activi developmental experiences, and their situat
ties and preferences (see Lytton & Romney, ed activity in society. These evolved charac
1991). Nonetheless, sex-typed expectancies teristics include the physical attributes of the
also are communicated through other social sexes and related behaviors, especially wom
channels, such as modeling of conventional en’s childbearing and nursing of infants and
family and occupational roles (Bussey & men’s greater size, speed, and upper-body
Bandura, 1999). strength. The dispositions that characterize
Biological mechanisms, especially hor men and women in a given society thus are
monal processes, also influence the develop flexibly defined by the biosocial interaction.
ment of gender identity. The prenatal hor Consequently, variations in gender identity
mone environment is known to influence and other sex-typed attributes emerge across
the development of some human behaviors cultures, age cohorts, and social roles as
that show sex differences. Perhaps best local conditions interact with the universal
122 II. I N T E R P E R S O N A L D IS P O S IT IO N S
framework provided by men’s and women’s researchers’ main goal is not predicting
evolved characteristics. agentic or communal behaviors, these mea
sures have commonly yielded disappoint
ing results. Because Bern’s and Spence and
Conclusion Helmreich’s scales were narrowly focused
on agentic and communal personality traits,
This chapter is in part a historical piece, they do not predict the broad range of psy
given that the best known developments in chological phenomena that can flow from
gender identity research took place in the gender identity. The cause of the disappoint
1970s. Researchers’ enthusiasm in that peri ing results is not in the scales themselves
od was ignited by Bern’s (1974) gender identi but in researchers’ applications of them.
ty scale, which reflected feminist thinking at Measures can address a variety of facets of
the time about the advantages of andi^Dgyny identity, and they will be most successful
for mental health and behavioral flexibility. at predicting responses that are compatible
Bern’s scale and related individual-difference with the facet assessed by the gender identity
measures also fit into the prevailing tendency scale. The classic measures of gender identity
of many feminists to regard sex differences in terms of communal and agentic personal
in behavior as stemming not from causes in ity effectively predict the specific domains
trinsic to women and men but from learned of communal and agentic responding, but
identity differences between the sexes. other measures predict more satisfactorily
Although interest in androgyny has outside of these domains.
waned, understanding of individual differ As we have presented in this chapter,
ences within gender groups has remained researchers can think about gender iden
an important scientific agenda, and gender tity in ways that reach beyond traits. With
identity remains a viable approach. Because these other treatments of gender identity,
identities represent individuals’ psychologi researchers have available to them a rich va
cal relationships to the social categories in riety of measures. By bringing these other
which they have membership, the study of measures to researchers’ attention, we hope
identities is crucial to understanding how to reinvigorate this important area of inqui
society infiltrates the psychology of the per ry and facilitate prediction of gender-typed
son. Gender identity is surely one of the most behavior in a wide range of domains. As we
important of social identities and thus war have illustrated, measures of gender identity
rants psychologists’ continuing attention. in terms of interests predict vocations and
Research on gender identity since the related leisure activities. Measures of gen
1970s mainly has been a single-note tune,
der identity in terms of construal of oneself
with the majority of research carried forward
in intimate relationships predict reactions
with Bern’s (1974) BSRI scale and the related
of men and women to close others. Finally,
PAQ scale developed by Spence and Helm- measures of collective gender identity pre
reich (1978). For example, in 2 0 0 8 , a search dict ingroup favorability and self-ascription
of PsycINFO turned up 1,748 total citations
of gender-group attributes. The available
for the BSRI, with 2 6 6 of these during the
approaches encompass a range of individual
past 5 years. The comparable figures for the
differences, and the associated measuring
PAQ are 885 overall and 218 during the past
instruments are broadly useful to psycholo
5 years. These statistics far outpace citations
gists, sociologists, and other researchers in
to any other gender-related measure of indi
terested in assessing individual differences in
vidual differences. This continuing popular
gender identity.
ity of the PAQ and BSRI suggests that, when
researchers think about explaining indi
vidual differences among women and men,
A ck n o w led gm en ts
their first (and often only) thought is to turn
to these personality-trait-based measures of T h is chapter w as com pleted while Wendy W ood
gender identity. Depending on researchers’ w as a Fellow at the R adcliffe Institute o f A d
purposes, this choice can be a mistake. vanced Study, H arvard University. We thank
W hat are the consequences of the continu A bigail Stew art for her helpful com m ents on an
ing popularity of the BSRI and PAQ? When earlier d raft o f the chapter.
8. G en der Id en tity 123
Eagan, S. K., 8c Perry, D. G. (2001). Gender identity: Helgeson, V. S. (1994b). Relation of agency and co m
A multidimensional analysis with implications for munion to well-being: Evidence and potential ex
psychosocial adjustment. D e v e lo p m e n t a l P sy c h o l planations. P sy c h o lo g ic a l B u lletin , 116, 4 1 2 - 4 2 8 .
og y , 37, 4 5 1 - 4 6 3 . Helmreich, R. L., Spence, J. T., 8c Wilhelm, J. A.
Eagly, A. H., 8c Chaiken, S. (1993). T h e p s y c h o l (1981). A psychometric analysis of the Personal At
og y o f attitu d es. Orlando, FL: Harcourt Brace Jo- tributes Questionnaire. S ex R o le s, 7, 1 0 9 7 - 1 1 0 8 .
vanovich. Hines, M., Brook, C., 8c Conway, G. S. (20 04 ). A n
Eagly, A. H., Wood, W., & Diekman, A. B. (2 000). drogen and psychosexual development: Core gender
Social role theory of sex differences and similarities: identity, sexual orientation, and recalled childhood
A current appraisal. In T. Eckes 8c H. M. Trautner gender role behavior in women and men with con
(Eds.), D e v e lo p m e n t a l s o c ia l p s y c h o lo g y o f g e n d e r genital adrenal hyperplasia (CAH), jo u r n a l o f S ex
(pp. 1 2 3 —174). Mahwah, NJ: Erlbaum. R e se a rch , 4 1, 7 5 - 8 1 .
Eagly, A. H., Wood, W., 8c Johannesen-Schmidt, M. Hofmann, W., Gawronski, B., Gschwendner, T., Le,
C. (200 4). Social role theory of sex differences and H., 8c Schmitt, M. (2 005). A meta-analysis on the
similarities: Implications for the partner preferences correlation between the Implicit Association Test
of women and men. In A. H. Eagly, A. E. Beall, 8c and explicit self-report measures. P erson ality a n d
R. J . Sternberg (Eds.), P sy ch o lo g y o f g e n d e r (2nd S o c ia l P sy c h o lo g y B u lletin , 3 1, 1 3 6 9 - 1 3 8 5 .
ed., pp. 2 6 9 - 2 9 5 ) . New York: Guilford Press. Josephs, R. A., Markus, H. R ., 8c Tafarodi, R. W.
Epstein, S. (1980). T he stability of behavior: II. Impli (1992). Gender and self-esteem, jo u r n a l o f P ers o n
cations for psychological research. A m er ica n Psy a lity a n d S o c ia l P sy ch olog y , 6 3 , 3 9 1 - 4 0 2 .
c h o lo g is t, 3 5, 7 9 0 - 8 0 6 . Judd, C. M ., James -Hawkins, L., Yzerbyt, V., 8c
Fazio, R. H., 8c Olson, M . A. (2003). Implicit measures Kashima, Y. (200 5). Fundamental dimensions of so
in social cognition research: Their meaning and cial judgment: Understanding the relations between
uses. A n n u al R e v iew o f P sy ch olog y , 5 4 , 2 9 7 - 3 2 7 . judgments of competence and warmth, jo u r n a l o f
Frable, D. E. S. (1997). Gender, racial, ethnic, sexual, P erson ality a n d S o c ia l P sy ch o lo g y , 89, 8 9 9 - 9 1 3 .
and class identities. A n n u al R e v ie w o f P sy ch olog y , Kashima, E. S., 8c Hardie, E. A. ( 2 00 0 ). The develop
48, 139-162. ment and validation of the Relational, Individual,
Gabriel, S., 8c Gardner, W. L. (1999). Are there “his” and Collective self-aspects (RIC) Scale. A sian j o u r
and “hers” types of interdependence?: The implica n a l o f S o c ia l P sy ch olog y , 3, 1 9 - 4 8 .
tions of gender differences in collective versus re Kashima, Y., Yamaguchi, S., Kim , U., Choi, S.-C.,
lational interdependence for affect, behavior, and Gelfand, M. J., 8c Yuki, M . (1995). Culture, gen
cognition, jo u r n a l o f P erson ality a n d S o c ia l P sy der, and self: A perspective from individualism-
c h o lo g y , 77, 6 4 2 - 6 5 5 . collectivism research. J o u r n a l o f P erso n a lity a n d
Gardner, W. L., 8c Gabriel, S. (20 04 ). Gender differ S o c ia l P sy ch olog y , 69, 9 2 5 - 9 3 7 .
ences in relational and collective interdependence: Kiefer, A. K., 8c Sekaquaptewa, D. ( 2 00 7 ). Implicit
Implications for self-views, social behavior, and stereotypes, gender identification, and math-related
subjective well-being. In A. H. Eagly, A. F2. Beall, outcomes: A prospective study of female college stu
& R. J. Sternberg (Eds.), P sy c h o lo g y o f g e n d e r (2nd dents. P sy c h o lo g ic a l S cien ce, 18, 1 3 - 1 8 .
ed., pp. 1 6 9 -1 9 1 ). New York: Guilford Press. Kite, M . E., 8c Deaux, K. (1986). Gender versus cat
Gardner, W. L., Gabriel, S., 8c Hochschild, L. (2002). egory clustering in free recall: A test of gender
When you and I are “we,” you are not threatening: schema theory. R e p r e se n ta tiv e R e se a rc h in S o cia l
T he role of self-expansion in social comparison. P sy ch o lo g y , 16, 3 8 - 4 3 .
jo u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 8 2 , Kohlberg, L. A. (1966). Cognitive-developmental anal
239-251. ysis of children’s sex-role concepts and attitudes. In
Gore, J . S., Cross, S. E., & Morris, M . L. ( 2 00 6). Let’s E. E. Macco by (Ed.), D e v e lo p m e n t o f s e x d iffe r e n c
be friends: Relational self-construal and the devel es (pp. 8 2 - 1 7 3 ). Stanford, CA: Stanford University
opment of intimacy. P erso n a l R e la tio n s h ip s, 13, Press.
83-102. Kuhn, M. H., 8c McPartland, T. S. (1954). An empiri
Grace, S. L., 8c Cramer, K. L. (2003). The elusive na cal investigation of self-attitudes. A m erica n S o c io
ture of self-measurement: T he Self-Construal Scale lo g ic a l R ev iew , 19, 6 8 - 7 6 .
versus the Twenty Statements Test, j o u r n a l o f S o c ia l Leander, N. P., Chartrand, T. L., 8c Wood, W. (2009).
P sy ch olog y , 143, 6 4 9 - 6 6 8 . M in d y o u r m a n n er ism s: E licitin g s te r e o ty p e c o n
Greenwald, A. G., 8c Banaji, M . R. (1995). Implicit fo r m it y th rou g h b e h a v io r a l m im icry . Manuscript
social cognition: Attitudes, self-esteem, and stereo submitted for publication.
types. P s y c h o lo g ic a l R e v iew , 1 0 2 , 4 - 2 7 . I.ippa, R. A. (1991). Some psychometric characteristics
Greenwald, A. G., 8c Farnham, S. D. ( 2 00 0 ). Using of gender diagnosticity measures: Reliability, valid
the Implicit Association Test to measure self-esteem ity, consistency across domains, and relationship to
and self-concept, j o u r n a l o f P erson ality a n d S o cia l the Big Five. J o u r n a l o f P erson ality a n d S o c ia l P sy
P sy ch olog y , 79, 1 0 2 2 - 1 0 3 8 . c h o lo g y , 61, 1 0 0 0 - 1 0 1 1 .
Greenwald, A. G., Poehlman, T. A., Uhlmann, E. L., Lippa, R. A. (1998). Gender-related individual dif
8c Banaji, M . R. (in press). Understanding and using ferences and the structure of vocational interests:
the Implicit Association Test: III. Meta-analysis of T he importance of the people-things dimension.
predictive validity, j o u r n a l o f P erson ality a n d S o J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 74,
c ia l P sy ch olog y . 996-1009.
Helgeson, V. S. (1994a). Prototypes and dimensions Lippa, R. A. (2001). On deconstructing and recon
of masculinity and femininity. S ex R o le s, 3 1, 6 5 3 - structing masculinity-femininity. J o u r n a l o f R e
6 82. s e a r ch in P erson ality , 3 5 , 1 6 8 - 2 0 7 .
8. G en der Id en tity 125
E m o t io n a l D is p o s it io n s
CHAPTER 9
Neuroticism
Thom as A. W id iger
he term neurosis was purportedly first retically or empirically based models of gen
T coined in 1769 by a Scottish doctor,
William Cullen, referring to disorders result
eral personality structure that fail to include
a domain of neuroticism (Digman, 1990).
ing from a “general affection” of the nervous Discussed within this chapter are alterna
system. The first edition of the American tive conceptualizations (and assessments) of
Psychiatric Association’s (1952) D iagnostic neuroticism, its origins, and its important
an d Statistical M anual: M ental D isorders life outcomes.
had a section devoted to “psychoneurotic
disorders”: “The chief characteristic of these
disorders is ‘anxiety’ which may be directly C onceptualization and Assessment
felt and expressed or which may be uncon
sciously and automatically controlled by the In Digman’s (1990) seminal review of the de
utilization of various psychological defense velopment of the Big Five factors of personal
mechanisms” (American Psychiatric Asso ity, he opined that “while fairly good agree
ciation, 1952, p. 31). ment appears to be developing concerning
Neuroticism, as a fundamental trait of the number of necessary dimensions, there
general personality, refers to an enduring is less accord with respect to their meaning”
tendency or disposition to experience nega (p. 422). This difficulty is quite understand
tive emotional states. Individuals who score able, as it is unlikely that any one single word
high on neuroticism are more likely than the can adequately characterize such broad do
average person to experience such feelings as mains of personality functioning. Any single
anxiety, anger, guilt, and depression. They word places more emphasis on a particular
respond poorly to environmental stress, range or aspect of the broad domain at the
are likely to interpret ordinary situations expense of other components or facets. Dig
as threatening, and can experience minor man appeared somewhat reluctant to offer
frustrations as hopelessly overwhelming. his own interpretation of this fourth fac
They are often self-conscious and shy, and tor of the Big Five, but he did indicate that
they may have trouble controlling urges and “Dimension IV is usually referred to as neu
impulses when feeling upset. Neuroticism roticism vs. emotional stability” (Digman,
is now recognized as one of the more reli 1990, p. 422).
ably identified and fundamental domains of The term neuroticism has been favored by
personality functioning and structure (M c Eysenck (1967), Costa and McCrae (1992),
Crae & Costa, 2003). There are few theo Digman (1990), and Zuckerman (2003).
129
130 III. E M O T IO N A L D IS P O S IT IO N S
E m otional instability (or stability) has been of the Tridimensional Personality Question
favored by Guilford (1975) and Goldberg naire, or TPQ ), and the Eysenck Personality
(1993). E m otionality is favored by Lee and Profiler (EPP; the Eysenck Personality Ques
Ashton (2004). Negative em otionality or tionnaire [EPQJ and the Eysenck Personality
negative tem peram ent has been preferred by Inventory [EPI] did not include facet scales).
Watson and Tellegen (1985) and Watson and Considered herein more closely are the facets
Clark (1994). Another conceptualization of angry hostility and aggression, impulsiv-
is harm avoidance, offered by Cloninger ity, emotional instability, and dependency.
(2000). These alternative titles are similar to
one another, and there is consistent empiri
A n gry Hostility and A ggression
cal support to indicate that these constructs,
or the instruments to assess them, are highly Costa and M cCrae (1992) suggest that “the
convergent (Watson, Clark, & Harkness, general tendency to experience negative af
1994; Widiger & Simonsen, 2 0 0 5 ; Zucker- fects such as fear, sadness, embarrassment,
man, 2002). Nevertheless, the alternative anger, guilt and disgust is the core of the
titles do reflect somewhat different concep neuroticism domain” (p. 14). Consistent
tualizations that become particularly evi with this conceptualization, the N EO PI-R
dent when comparing respective assessment includes facet scales for angry hostility,
instruments, especially variation in facet anxiety, and depression, along with self-
scales. consciousness, vulnerability, and impul-
Table 9.1 lists six alternative measures of sivity (see Table 9.1). Watson and Tellegen
neuroticism and how their facet scales are (1985) define this domain of personality as
aligned with one another or, more accurate negative emotionality, negative affectivity,
ly, how weakly they are aligned. Included or negative temperament and, when using
within Table 9.1 are the facet scales from the the Positive and Negative Affect Schedule
N EO Personality Inventory— Revised (NEO (PANAS; Watson, Clark, & Tellegen, 1988),
PI-R) assessment of the five-factor model confine its assessment to such negative af
(FFM ), the Multidimensional Personality fects as hostility and irritability, along with
Questionnaire (M PQ), the Big Five Aspects being afraid, scared, nervous, jittery, guilty,
Scales (BFAS), the H EXA C O Personality ashamed, upset, and/or distressed (surpris
Inventory (H EXA CO -PI), the Temperament ingly, though, no depression, sadness, or sor
and Character Inventory (TC I; an expansion row). The expanded version, the PANAS-X
N ote. NEO PI-R, NEO Personality Inventory— Revised (Costa & McCrae, 1992); EPP, Eysenck Personality Profiler (EPP;
Eysenck et al., 1992); MPQ, Multidimensional Personality Questionnaire (Tellegen, 1982); BFAS, Big Five Aspects Scale
(DeYoung et al., 2007); HEXACO-PI, HEXACO Personality Inventory (Lee 8c Ashton, 2006); TCI, Temperament and
Character Inventory (Cloninger, 2000).
9. N eu ro ticism 131
(Watson & Clark, 1994), includes subscales (2005) in their H EXA CO model of personal
for each of the “basic negative emotions” of ity structure (HE.XACO stands for Honesty,
fear, sadness, and guilt, along with hostility, Emotionality, Extraversion, Agreeableness,
assessed by the trait terms angry, hostile, ir Conscientiousness, and Openness). Within
ritable, scornful, disgusted, and loathing. H EXA CO Emotionality “content related to
An important distinction between the di anger versus even-temper shifts from neu
mensional models of personality of Clark roticism to the new variant of low agree
and Watson (Clark, 2 0 0 5 ; Watson, Gamez, ableness, and content related to sensitivity,
&C Simms, 2005) and Costa and McCrae sentimentality versus toughness shifts from
(1992) is that Clark and Watson do not in agreeableness to the new variant of neuroti
clude a domain of antagonism separate from cism (which we have named emotionality to
negative temperament. Clark and Watson reflect this exchange of content)” (Ashton &
have long advocated instead a three-factor Lee, 2 0 0 5 , p. 1324).
model of general personality structure, con The four H EXA CO -PI facets of emo
sisting of negative affect, positive affect, and tionality are fear, anxiousness, dependency,
constraint. Therefore, much of the antago and sentimentality. Ashton and Lee (2005)
nistic and aggressive behavior in the FFM suggest that these four facets are more con
is included within their negative affectivity sistent with the language structure of this
domain. This is explicit within the M PQ as domain of personality (Saucier & Goldberg,
sessment of negative emotionality (Tellegen, 20 0 1, suggest alternatively that the two
1982), whose subscales include Aggression, primary lexical facets should be irritability
along with Stress Reaction and Alienation and anxiousness/fearfulness). In this regard,
(see Table 9.1). M PQ Aggression includes the Ashton and Lee conceptualization does
being physically aggressive, enjoying upset align more closely with Cloninger’s (2000)
ting and frightening others, enjoying scenes and Eysenck’s (1967) characterization and
of violence (e.g., fights, violent movies), and assessment of harm avoidance and neuroti
victimizing others for one’s own advantage cism, respectively (see Table 9.1). The facet
(Tellegen, 1982; Tellegen & Waller, 1987). scales of Cloninger’s TC I harm avoidance
A conceptual rationale for including an concern fear of uncertainty, shyness, fati
tagonistic, aggressive behavior within the gability, and anticipatory worry. The facet
domain of negative emotionality or neuroti scales of EPP neuroticism concern anxiety,
cism is that the anger could be said to be inferiority, and unhappiness. Neither in
the driving force or motivating energy for cludes anger (let alone aggression).
aggressive behavior, just as positive emo H EXA CO emotionality also fails to in
tionality can be considered to be the driv clude a facet of depressiveness. One might
ing force of extraverted behavior (Watson expect emotionality to include anger and de
& Clark, 1997) and anxiousness the driving pression. Questions have been raised as to
force of self-conscious behavior (Costa & whether such feelings as disgust and embar
M cCrae, 1992). “Hostile people are gener rassment are actually emotions, but anger
ally antagonistic” (Costa & M cCrae, 1992, does appear to be comfortably included
p. 45). However, a separation of antagonis within a domain of emotionality (Ekman,
tic behavior from negative emotionality is 1999; Russell, 2003) and certainly within a
supported by the factor-analytic studies of domain of negative emotionality, along with
language structure (Ashton & Lee, 2001). In anxiousness and depressiveness (Watson &
addition, not all antagonistic behavior will Tellegen, 1985).
be driven by anger (e.g., manipulation, ex Some of the variation in conceptualization
ploitation, arrogance, and deception), and across models can reflect simply the number
even some instances of aggression can occur of factors that are extracted. For instance,
without significant feelings of anger (e.g., in if one reduces the FFM to just three factors,
strumental aggression). it is perhaps not surprising that aggression
A complementary position to shifting ag shifts into a domain of negative affectivity
gression into neuroticism is to shift angry (M arkon, Krueger, & Watson, 2005). In
hostility out of neuroticism and into the this regard, one could say that the three- and
FFM domain of antagonism, which is pre five-factor models are not in conflict. They
cisely what is proposed by Ashton and Lee just reflect the level of a common hierarchi
132 III. E M O T IO N A L D IS P O S IT IO N S
cal model in which one prefers to work and tion refers to the tendency to think and re
study (i.e., two-, three-, four-, five-, or six- flect on the consequences of an act before
factor models). However, even if this is the engaging in that act” (Whiteside & Lynam,
basis for the variation across models, there 20 0 1 , p. 685). N EO PI-R self-discipline re
is still the question of which level of the hi fers to “the ability to begin tasks and carry
erarchy provides the optimal description of them through to completion despite boredom
personality structure. In addition, one must and other distractions” (Costa & M cCrae,
remain cognizant of the substantive differ 1992, p. 18). Whiteside and Lynam refer to
ences in neuroticism across the levels of the this disposition as perseverance. The fourth
hierarchical model. and final variant of impulsivity is N EO PI-R
excitement seeking, involving an enjoyment
in taking risks and engaging in dangerous
Impulsivity
activities, which aligns closely with Zucker-
Costa and McCrae (1992) suggest that neu man’s (2002) sensation seeking. Lynam and
roticism involves more than just a suscep colleagues have developed a measure of these
tibility to psychological distress: “Perhaps four variants of impulsivity and have further
because disruptive emotions interfere with demonstrated that the four variants have
adaptation, men and women high in neu quite different correlations with existing im
roticism are also prone to have irrational pulsivity measures and validators (Lynam &C
ideas [and] to be less able to control their Miller, 2 0 0 4 ; Whiteside, Lynam, Miller, &
impulses” (p. 14); hence their inclusion of a Reynolds, 2005).
facet of impulsivity within the N EO PI-R. In sum, there does appear to be compel
The N EO PI-R assessment of neuroticism ling conceptual and empirical support for
is relatively unique in its inclusion of this including impulsivity within neuroticism,
facet. Nevertheless, studies that report cor although it might not be best considered as
relations of N EO PI-R Impulsivity with (for a core trait. Perhaps it is best understood as
instance) Zuckerm an-Kuhlm an Personality a corollary. It might also be preferable to use
Questionnaire (ZKPQ) Impulsive Sensation the title of urgency when referring to this
Seeking (Zuckerman, 2002) or M PQ Con facet of the N EO PI-R rather than the more
straint (Tellegen, 1982) do tend to report ambiguous or relatively nonspecific title of
substantial correlations with neuroticism as impulsivity (Whiteside & Lynam, 2001).
well (e.g., Aluja, Garcia, & Garcie, 2004).
Impulsivity itself is a rather broad trait or, at
Em otional Instability
least, a variably defined construct that can
refer to quite a number of different behav DeYoung, Quilty, and Peterson (2007),
iors (Depue & Collins, 1999). Whiteside and through analyses of the facet scales of the
Lynam (2001) have, in fact, used the NEO N EO PI-R (Costa & M cCrae, 1992) and
PI-R facet structure to distinguish between the Abridged Big Five Circumplex scales
four different variants of impulsivity, one of from the International Personality Item Pool
which would be correctly placed within the (Goldberg, 1999), identified two facets for
domain of neuroticism: urgency. “It refers to each domain of the FFM . They suggested
the tendency to experience strong impulses, that these 10 facets align well with genetic
frequently under conditions of negative af factors within the N EO PI-R identified by
fect” (Whiteside & Lynam, 2 0 0 1 , p. 685). Jang, Livesley, Angleitner, Reimann, and
Whiteside and Lynam (2001) suggest that Vernon (2002). “Each of the Big Five do
additional variants of impulsivity are low mains, therefore, appears potentially di
premeditation, low perseverance, and sensa visible into two subdomains with distinct
tion seeking. The disposition to refrain from biological sources” (DeYoung et al., 2007,
acting on the spur of the moment and with p. 881). They constructed an inventory to
out regard to consequences is represented by assess these 10 facets, the Big Five Aspects
the N EO Pl-R Conscientiousness facet of Scales (BFAS). The two facets of Neuroti
deliberation, “the tendency to think careful cism are volatility and withdrawal.
ly before acting” (Costa & M cCrae, 1992, The withdrawal facet is not what the name
p. 18). They refer to this variant of impulsiv might imply, social withdrawal. Its items as
ity as a lack of premeditation: “Premedita sess instead traditional features of neuroti-
9. N eu ro ticism 133
cism, such as feeling blue, worrying, feeling characteristically nervous, pessimistic, or ir
threatened, being easily discouraged, being ritable (Miller & Pilkonis, 2006).
afraid, and feeling overwhelmed. Volatility
items within the BFAS include “get upset
D ependency
easily,” “keep my emotions under control,”
“change my mood a lot,” “am a person One of the four facets of H EXA CO emo
whose moods go up and down easily,” and tionality is dependency, defined as a “need
“get easily agitated” (DeYoung et al., 2007, for emotional support from others” (Lee &
p. 887). Such items are clearly referring to Ashton, 2 0 0 4 , p. 334). Dependency is one
an emotional instability (or volatility), con of the personality disorders included within
sistent with Goldberg’s (1993) original char the D iagnostic an d Statistical M anual o f
acterization of this domain as emotional sta M ental D isorders (American Psychiatric
bility versus instability. Association, 2000). The essential feature
The N EO PI-R, in contrast, does not in of dependent personality disorder is a per
clude any facet scales that represent or ex vasive and excessive need to be taken care
plicitly assess this volatility or emotional of that leads to submissive and clinging be
ity instability. The N EO PI-R does include havior and fears of separation (American
scales to assess anxiousness, depressiveness, Psychiatric Association, 2000). Bornstein
and angry hostility, which are the negative and Cecero (2000) and Saulsman and Page
emotions that would be expressed by a per (2004) conducted meta-analyses of studies
son who is emotionally volatile or unstable, correlating measures of the FFM with mea
but the N EO PI-R negative affect scales are sures of dependency, and both reported that
typically interpreted as referring to a dis the highest and most consistent relationship
position to be characteristically or consis was with the domain of neuroticism.
tently anxious, dysphoric, or angry rather However, considering dependency to be a
than being periodically (e.g., inconsistently) facet of neuroticism could represent a confu
emotionally explosive, unstable, or volatile. sion of the fears, needs, and insecurities of
Shedler and Westen (2004) have presented the dependent person with his or her depen
their Shedler-Westen Assessment Procedure dent behavior. Agreeableness, as assessed by
as “an alternative to the five-factor model” the N EO PI-R (Costa &c M cCrae, 1992), is
(p. 1743). Factor analyses of this item pool the tendency to be trusting, straightforward,
have yielded scales that resemble four of the altruistic, compliant, modest, and tender-
five domains of the FFM , but “we also iden minded. Well before ever considering per
tified some useful diagnostic distinctions, sonality disorders as maladaptive variants
such as the distinction between negative af of the FFM , Costa and McCrae (1985) had
fectivity and emotional dysregulation, which stated that “agreeableness can also assume a
increasingly appear to be distinct concepts, pathological form, in which it is usually seen
as reflected in the difference between stable as dependency” (p. 12). Widiger, Trull, Clar-
dysthymia and [the affective instability of] kin, Sanderson, and Costa (2002) include as
borderline personality disorder” (Westen & pects of neuroticism within their FFM con
Shedler, 2007, p. 818). Further support is ceptualization of dependency (particularly
provided by Miller and Pilkonis (2006). the facets of anxiousness, self-consciousness,
Volatility and emotional dysregulation and vulnerability) but, consistent with Costa
probably should be understood as facets and M cCrae, they suggest that “the DSM -IV
of neuroticism, consistent with Goldberg’s diagnostic criteria set [for dependent person
(1993) characterization of this personality ality disorder] includes many explicit exam
domain as a contrast between emotional sta ples of pathological agreeableness, such as
bility and instability (DeYoung et al., 2007; excessive compliance (difficulty expressing
Mullins-Sweatt & Widiger, 200 7 ). Persons disagreement), altruism (volunteering to do
who are emotionally volatile, unstable, or unpleasant things), and modesty (needing
dysregulated will characterize themselves as advice and reassurance from others to make
being higher in levels of anger, anxiousness, everyday decisions)” (p. 96). Pincus (2002)
and depressiveness than the average person. confines his conceptualization of dependen
Nevertheless, there are important distinc cy simply to interpersonal relatedness. Such
tions between being emotionally volatile and trait terms as docile, servile, self-sacrificing,
134 III. E M O T IO N A L D IS P O S IT IO N S
m odest, agreeable, com pliant, clinging, o b e of threat, either physical or social. The ab
dient, gullible, submissive, self-effacing , and sence of a capacity to be anxious would, in
even dependent have long been considered many instances, be maladaptive, as it would
interpersonal in nature, representing various impair the individual’s ability to anticipate
combinations of the fundamental domains or appreciate signs of threat and danger.
of high communion and low agency that Similarly, depression can serve as a signal to
define the interpersonal circumplex (Pincus, others that help is needed to overcome a loss,
2 0 0 2 ). damage, or injury. Persons with this capac
In sum, the self-consciousness, vulner ity to express depressed, sad, or sorrowful
ability, and insecurity of neuroticism may behavior are more likely to elicit the help
contribute to the occurrence of dependent and support from others that are necessary
behavior (Miller & Pilkonis, 2 0 0 6 ), as anx to overcome negative life events.
iousness would contribute to the occurrence Given the lack of obvious clarity about
of self-consciousness (Costa & M cCrae, the precise situations in which any such
1992), as anger would contribute to the oc signal should or should not be emitted and
currence of antagonistic behavior (Clark & the lack of certainty regarding the optimal
Watson, 1999) and negative affects to the magnitude of any signal within any particu
occurrence of urgent impulsivity (Whiteside lar situation, it should not then be too sur
& Lynam, 2001). This dependent behavior prising that there is considerable variation in
may perhaps be best understood as a poten individual differences in the disposition to
tial correlate of neuroticism or emotional respond with anxiety or depression (Buss,
ity rather than as a fundamental or defin 1996; Penke, Denissen, & Miller, 2007).
ing feature of neuroticism, particularly as Evolution would favor variability in a signal
dependent behavior need not necessarily or expression whose adaptive value will vary
always be driven by neuroticism. across time and social, physical contexts.
In sum, there is considerable agreement as There is consistent support for the heri-
to the existence of a domain of neuroticism, tability of neuroticism. The research is gen
emotionality, emotional instability, or nega erally consistent with the routine finding
tive emotionality, however it may be identi that approximately 4 0 - 6 0 % of the vari
fied. And there is considerable convergence ance appears to be genetic, 2 0 -3 0 % con
among alternative measures of this domain. cerns nonshared environmental influences,
Nevertheless, there is no unanimity as to its and shared environmental contribution
precise assessment, and this is quite evident approaches zero, with the remaining vari
when one considers the specific facets of ance left unexplained (Turkheimer, 2000).
respective measures of this domain of gen Yamagata and colleagues (2006) have gone
eral personality functioning. This variation beyond the univariate behavior genetic
in how the domain is assessed can have a methodology to explore multivariate behav
significant impact on research findings, as ior genetic heritability that considers the co
illustrated in the following section, when variation among two or more traits. They
considering the origins, or more precisely considered N EO PI-R data obtained from
the molecular genetics, of neuroticism. three large, independent twin samples from
Canada, Germany, and Japan to determine
whether the facet structure of the NEO PI-R
Origins is consistent with shared genetic variance.
They concluded “that the five factors are
It is not difficult to infer the reasons for a ‘genetically crisp’ ” (Yamagata et al., 2 0 0 6 ,
fitness advantage for emotional stability p. 994). However, it should be noted that
relative to instability, but there are also com one qualification for the domain of neuroti
pelling arguments for the adaptivity or fit cism was that angry hostility, although ge
ness value of emotional instability or, more netically loading primarily on neuroticism,
specifically, anxiousness and depressiveness. did also load on the genetic factor of an
“N orm al” anxiety is an emotion that helps tagonism within the German and Japanese
organisms defend against a wide variety of samples and impulsiveness loaded as highly
threats. Pain is a signal of a threat to physi on the genetic factor of conscientiousness as
cal safety and survival; anxiety is an emo it did for neuroticism within the Canadian
tional pain that can be a comparable signal and Japanese samples.
9. N eu ro ticism 135
Yamagata and colleagues (2006) suggest (2005), showing “a strong dominant effect
ed that one implication of their findings “is of S-H TT-LPR on N EO neuroticism, and a
that molecular genetic studies of personal more modest but nevertheless significant re
ity seeking putative loci would clearly ben cessive effect on TCI/TPQ harm avoidance”
efit from the use of the N EO PI-R” (p. 994). (pp. 8 9 5 -8 9 6 ). This conclusion is consistent
Studies of the heritability of neuroticism with a subsequent study by Schmitz, Hen-
have gone beyond simply bivariate and mul nig, Kuepper, and Reuter (2007), who re
tivariate genetic analyses to explore more ported a significant effect for neuroticism as
precisely molecular genetics. The primary assessed by the N EO Five-Factor Inventory
focus of attention has been a polymorphism (NEO FFI; Costa & M cCrae, 1992) and the
of the serotonin transporter gene 5H TT- EPQ — Revised (F’PQ -R; Eysenck, Barrett,
LPR. This interest has grown from the find Wilson, & Jackson, 1992; Miles & Hempel,
ing that serotonergic systems are considered 200 4), but not by TC I harm avoidance. They
to be integral to emotion regulation and that further explored which specific items of the
drugs that are effective in reducing anxi N EO FFI, EP Q -R , and TC I were most con
ety and depression act largely through the tributory, and they suggested that the signifi
serotonergic system. Two meta-analyses of cant findings are due largely to items assess
molecular genetic research have supported ing depressiveness and stress sensitivity.
the conclusion that there is a meaningfully Findings that the results vary in part due
significant relationship (effect sizes of ap to the measurement instrument, though,
proximately 0 .2 0 ; Cohen, 1992) between should not be at all surprising, given the
neuroticism and short versus long alleles of considerable variation in how each instru
S H T T-LP R , particularly when neuroticism ment assesses neuroticism (see Table 9.1). In
is assessed with the NEO PI-R (Schinka, addition, it is perhaps unrealistic to expect a
Busch, & Robichaux-Keene, 2 0 0 4 ; Sen, very specific genetic association, or at least
Burmeister, & Ghosh, 2004). a strong effect size, for an endophenotype
The specific results from and conclusions as broad as neuroticism (Flint & Munafo,
concerning these meta-analyses, however, 2007). And, finally, the structure of the
have been disputed. A meta-analysis con SH TT-LPR gene has itself been questioned.
ducted by Munafo and colleagues (2003) Although most past studies considered the
also found support for the association of SH TT-LPR gene to have two alleles, one
SHTT-LPR with neuroticism, although not long and one short, there is now research
as strong (an effect size of only 0.11; Cohen, to suggest that it is triallelic (Beitchman
1992). More important, perhaps, Munafo, et al., 2006). It has one short allele, s, and
Clark, and Flint (2005a) subsequently ex two long alleles, Lg and L a (Beitchman et
plored whether the association was instru al., 2006). The s and the Lg alleles tend to
ment specific and concluded on the basis of be relatively low-expressing (they are pres
their new meta-analysis that the association ent less often), and they are associated with
was confined largely to studies using Clon lower brain levels of 5H T (Beitchman et
inger’s (2000) TC I (or TPQ ) harm-avoidance al., 2 0 0 6 ; Hu et al., 20 0 7 ); the L a allele is
scale rather than NEO PI-R neuroticism, a not. Research before this discovery consid
finding in direct contradiction to the meta ered only the presence of the short allele as
analyses of Schinka and colleagues (2004) relevant to risk; cases with the high-risk Lg
and Sen and colleagues (2004). The incon allele were perhaps inaccurately assigned to
sistency appears to reflect (in part) disagree the low-risk comparison group, thus reduc
ments with regard to which studies should ing differences between the groups in gene-
be included (e.g., Munafo et al., 2 005a, based risk status.
excluded studies with psychiatric samples),
how studies should be weighted, and the op
timal genetic comparison group. Subsequent Life O utcom es
reanalyses have not resolved the dispute
(Munafo, Clark, & Flint, 20 0 5 b ; Schinka, Neuroticism, as a disposition to experience
2 0 0 5 ; Sen, Burmeister, & Ghosh, 2005), negative emotional states, to respond poorly
although when using Cohen’s (1992) mea to environmental stress, and to experience
sure of effect size, Munafo and colleagues even minor frustrations as hopelessly over
(2005b) replicated the findings of Schinka whelming, would, not surprisingly, be con
136 III. E M O T IO N A L D IS P O S IT IO N S
sidered to contribute to the etiology of a va Schutte, 2 0 0 6 ; M unafo, Zetteler, & Clark,
riety of negative life outcomes. One of the 20 0 7 ); increased alcohol abuse (Malouff,
more robust and heavily studied findings is Thorsteinsson, Rooke, & Schutte, 20 0 7 );
the association of neuroticism with a wide greater unprotected sexual activity (Hoyle,
variety of psychopathology. M alouff, Thor- Fejfar, & M iller, 20 0 0 ); lower happiness
steinsson, and Schutte (2005) conducted (DeNeve & Cooper, (1998); higher depen
a meta-analysis of the relationship of neu dency (Bornstein & Cecero, 2 0 0 0 ); higher
roticism (as well as other domains of the extrinsic religiosity (Saroglou, 2002); and
FFM) with D SM -IV mental disorders. They higher rates of criminal arrest (Huo-Liang,
reported effect sizes for the relationship of 2006). Beyond these meta-analyses, neuroti
neuroticism with mood disorders to be 1.54, cism has been shown to be associated with
with anxiety disorders to be 1.04, with so higher existential concerns, weaker identity
matoform disorders to be 1.20, with eating integration, dissatisfaction and conflict with
disorders to be 1.29, and with schizophrenia in relationships, dissolution of relationships
to be 1.08 (the effective size for the relation (e.g., divorce), and financial insecurity (Ozer
ship with substance use disorders was 0.54). & Benet-Martinez, 2006). Neuroticism is
An effect size of 0 .8 0 or higher is generally clearly a very robust predictor of negative
understood to indicate a large effect (Cohen, life outcomes.
1992). Cassin and von Ranson (2005) re These relationships between neuroticism
ported comparable results in their meta and negative life outcomes are generally
analysis of the research on neuroticism and understood to be etiological, that is, neu
eating disorders. Van Os and Jones (2001) roticism contributes to the development or
reported in a prospective longitudinal study occurrence of the negative life outcomes.
of 5,3 6 2 participants that persons with high Connor-Smith and Flachsbart (2007) con
neuroticism scores at age 16 were 1.93 times ducted a meta-analysis of the relationship
more likely to meet criteria for schizophrenia of personality and coping mechanisms as
later in life. In a community sample of 2 ,0 8 5 reported in 165 studies and reported that
young adults, Parslow, Jorm , and Christens neuroticism was associated with problem
en (2006) reported a higher likelihood of atic strategies such as wishful thinking,
persons elevated on neuroticism to develop withdrawal, and emotion-focused coping.
posttraumatic stress disorder on subsequent O f course, it is not entirely clear whether
trauma exposure. the poorer coping mechanisms were a re
The robust relationship of neuroticism sult of neuroticism or whether poor coping
with various forms of psychopathology is strategies led to the development of anxious
consistent with other important life out ness, depressiveness, vulnerability, and self-
comes. Neuroticism is associated with the consciousness.
occurrence of a variety of physical illnesses, There are, in fact, a variety of mechanisms
including (but not limited to) hypertension, through which neuroticism could result
cardiovascular disease, and diabetes, as well in negative life outcomes. With respect to
as reduced survival among those with the lowered physical health, neuroticism could
respective illnesses (Smith & MacKenzie, provide a direct effect, through altered au
2 0 0 6 ; Suls & Bunde, 2005). The volume and tonomic regulation of the cardiovascular
robustness of the study of neuroticism and system, immune suppression, and increased
negative life outcomes is suggested simply inflammation associated with higher levels
by the number of meta-analyses that have of negative affectivity, or an indirect effect,
been conducted, including meta-analyses through poorer health habits, increased ex
of the relationship of neuroticism to lower posure to daily stressors, and other life dif
subjective well-being (Steel, Schmidt, & ficulties (Smith & MacKenzie, 2 0 0 6 ; Suls &
Schultz, 200 8 ); lower academic satisfaction Bunde, 2005). The same can occur for the
(Trapmann, Hell, Hirn, & Schuler, 2 0 0 7 ); contribution of neuroticism to the etiology of
lower job satisfaction (Judge, Heller, & psychopathology. Neuroticism can contrib
M ount, 20 0 2 ); lower performance motiva ute both diathesis and stress (Caspi, Roberts,
tion, such as lower self-efficacy and goal & Shiner, 2005), providing a vulnerability to
setting (Judge & Ilies, 2002); lower leader psychopathology through both reactive and
ship (Judge, Bono, Ilies, & Gerhardt, 2002); evocative person-environment transactions.
higher smoking (Malouff, Thorsteinsson, & The former relationship is essentially a di
9. N eu ro ticism 137
rect effect; reacting to events with high levels (pathoplastic relationships), and they can be
of distress, anxiety, and worry provides an associated through a common, underlying
explicit risk for various forms of psychopa etiology (spectrum relationships). The etio
thology, particularly mood and anxiety dis logical, causal relationship is the one of most
orders. The evocative person-environment interest to theorists and researchers, but any
transaction may occur when one’s frequent study concerned with the potential contribu
expressions of upset, worry, and vulnerabil tion of neuroticism to the development of
ity produce negative reactions from others, psychopathology must be fully cognizant of
thus reinforcing and increasing the original the potential impact of the pathoplastic and
distress (i.e., personality as causing stress). spectrum relationships on the obtainment of
There is empirical support, for instance, for significant research findings. Each of these
a relationship of neuroticism to the presence two other forms of relationship of neuroti
and experience of lower levels of social sup cism to negative life outcomes are discussed
port (Kendler, Gardner, & Prescott, 2006). in turn.
The contribution of neuroticism to the de
velopment of physical health problems, fi
Pathoplastic Relationships
nancial difficulties, and dissolution of rela
tionships and other negative life outcomes The influence of neuroticism and psycho
(Ozer & Benet-Martinez, 2 0 0 6 ) would also pathology on the presentation, appearance,
contribute in turn to a considerable amount or expression of one another is typically
of stress, which persons high in neuroticism characterized as a pathoplastic relationship
would have an inherent difficulty emotion (Widiger 8c Smith, 2008). This pathoplastic
ally surmounting. relationship is bidirectional, as psychopa
The contribution of neuroticism to the thology can vary in its appearance depend
etiology of psychopathology has also been ing on a person’s premorbid level of neuroti
studied at a genetic level. M unafo, Clark, cism, and the appearance or presentation
Roberts, and Johnstone (2006), for instance, of neuroticism can similarly be affected by
addressed the question of whether trait neu the presence of a current (or even recently
roticism mediates the putative association experienced) psychopathology. Pathoplas
between the serotonin transporter gene tic relationships are not confined to but are
polymorphism (SHTT-LPR) and lifetime most readily understood as methodological
major depression in adults. In this study, 251 confounds in studies on the etiological rela
participants completed the EPQ and a self- tionship of personality to negative life out
report measure of depression. The SHTT- comes.
LPR genotype was significantly associated For example, a pathoplastic effect of neu
with both neuroticism and lifetime major roticism on the expression or appearance
depression, as was neuroticism with lifetime of ill health is the finding that persons high
major depression. Neuroticism, however, ac in neuroticism complain of more symptoms
counted for 42% of the effect of SHTT-LPR and are more likely to seek treatment (ten
genotype on lifetime major depression, in Have, Oldehinkel, Vollebergh, & Ormel,
dicating possible mediation. Similar results 2005). Objectively, they may be no more
have been reported by Jacobs and colleagues ill than the person low in neuroticism, but
(2006). they are more likely to report the presence
The study of the etiological contribution of symptoms and to seek treatment for them
of neuroticism to psychopathology (and (Chapman, Duberstein, Sorensen, Lyness,
other negative life outcomes) can also be & Emery, 2006). This does not necessarily
hindered, or at least complicated, because imply that they lack a clinically significant
of two other ways in which neuroticism and level of actual ill health, but it is possible
psychopathology can be associated with that the extent of the relationship of neu
one another. Widiger and Smith (2008) roticism to ill health could be due (at least
distinguish between three fundamentally in some part) to the increased reporting of
different ways in which neuroticism and symptoms in persons high in neuroticism.
psychopathology can relate to one another. This concern has been primarily evident in
In addition to a causal relationship, neuroti studies of the relationship of neuroticism to
cism and psychopathology can influence the physical illness (Smith & MacKenzie, 2 0 0 6 ;
presentation or appearance of one another Suls & Bunde, 2005), but perhaps it should
138 III. E M O T IO N A L D IS P O S IT IO N S
disorder, as assessed with a semistructured sessment when the person is suffering from a
interview. None of them had ever received mood (or other comparable) disorder or con
a psychotropic medication, had ever abused dition. Fluctuations in mood secondary to a
drugs, or had ever been in treatment for a mood disorder could reflect actual changes
mental disorder, nor were any of them cur in levels of neuroticism, and mood disorders
rently seeking or desiring treatment for a could perhaps be understood as alterations
mental disorder. In other words, they were in personality functioning, but if one wishes
in many respects above normal in psycho to understand an etiological contribution of
logical functioning. One certainly could not neuroticism to the onset of the mood disor
attribute any subsequent changes in their der, the two conditions do need to be distin
personality traits to the effect of treating a guished from one another.
co-occurring mood disorder. The paroxetine Longitudinal studies that begin prior to
(and placebo) treatment continued for 4 the onset of the psychopathology are per
weeks. Knutson and colleagues reported that haps the most informative approach to
the SSRI administration (relative to placebo) controlling for pathoplastic distortions in
reduced significantly scores on a self-report self-description, and such studies have in
measure of neuroticism. The magnitude of deed reported strong support for the in
change even correlated with plasma levels of creased likelihood of future onsets of major
SSRI within the SSRI treatment group. As depression in persons with elevated neu
concluded by Knutson and colleagues, this roticism, even for persons with no prior his
was a clear “empirical demonstration that tory of diagnosed mood disorder (Fanous,
chronic administration of a selective sero Neale, Aggen, & Kendler, 2007; Kendler,
tonin reuptake blockade can have significant Gatz, Gardner, & Pederse, 20 0 6 ). In cross-
personality and behavioral effects in normal sectional studies, semistructured interview
humans in the absence of baseline depression assessments of neuroticism may be helpful,
or other psychopathology” (p. 378). In sum, particularly to the extent that the interview
normal personality can be altered through ers actively attempt to identify the level of
pharmacology. neuroticism that predated the onset of a
Costa and colleagues (2005) also note current psychopathology. Self-report inven
that it is generally accepted that Alzheimer’s tories do not typically instruct participants
disease, Parkinson’s disease, and traumatic to distinguish between their current psycho
brain injury can produce actual changes pathology and their premorbid personality
to personality functioning. The American functioning, and some may even focus re
Psychiatric Association’s (2000) diagnostic spondents’ attention on their current func
manual does recognize the concept of per tioning. However, there is currently only one
sonality change due to a general medical semistructured interview for the assessment
condition, including even a labile type that of neuroticism (Trull & Widiger, 1997), and
is characterized primarily by affective labil semistructured interviews will not necessar
ity. In fact, the World Health Organization ily be immune to the pathoplastic effects of
(W HO) recognizes within the International mood distortion (Widiger & Samuel, 2005).
Classification o f D iseases (ICD -10; W H O, Another approach would be to obtain as
1992) the concept of personality change sessments of neuroticism from persons who
secondary to severe mental illness, which are closely familiar with the subject of the
would include changes in levels of neuroti assessment. These informants (e.g., spouses,
cism secondary to a mood disorder. friends, or colleagues) will lack the distort
However, to the extent that one considers ing effects of the mood disorder and may
self-report descriptions of neuroticism sec even be able to clearly distinguish the target’s
ondary to a mood disorder to reflect either functioning prior to versus during the onset
fluctuations or actual changes in personal of the current psychopathology. Agreement
ity functioning, it becomes difficult to con between self-descriptions and peer descrip
duct research on the etiological contribution tions of personality traits has generally been
of neuroticism to the mood disorder. They good to excellent when sampling within non-
are no longer distinguished constructs. One clinical populations (McCrae, Stone, Fagan,
should at least not attempt to infer premor- & Costa, 1998), but agreement between
bid personality traits on the basis of an as self-descriptions and informant descriptions
140 III. E M O T IO N A L D IS P O S IT IO N S
of personality within clinical samples has at co-occurrence) among these Axis I mental
times been poor to only adequate (Klonsky, disorders (Clark, 2 0 0 5 ; Krueger, M arkon,
Oltmanns, & Turkheimer, 2002). This dis Patrick, & Iacono, 2 0 0 5 ; Watson, 2005;
agreement may be due in part to the impact Widiger &C Clark, 20 0 0 ). There are many
of the psychopathology on the target’s self instances in which the presence of multiple
description, but there is, as yet, inadequate diagnoses suggests the presence of distinct
research to fully understand the nature and yet comorbid psychopathologies; however,
implications of the disagreement (Ready & perhaps in just as many instances, one has
Clark, 2005). instead the presence of a single, common
underlying psychopathology and diath
esis. “Comorbidity may be trying to show
Spectrum Relationships
us that many current treatments are not so
Psychopathology is typically defined as a much treatments for transient ‘state’ mental
condition that meets the diagnostic crite disorders of affect and anxiety as they are
ria for a mental disorder currently included treatments for core processes, such as nega
within D SM -IV (American Psychiatric As tive affectivity, that span normal and abnor
sociation, 20 0 0 ). Axis I of DSM -IV includes mal variation as well as undergird multiple
such conditions as major depressive disor mental disorders” (Krueger, 2 0 0 2 , p. 44).
der, social phobia, eating disorders, and dys- Krueger and his colleagues have been par
thymia. Axis II includes such conditions as ticularly productive in replicating within a
borderline, avoidant, and dependent person variety of populations the two dimensions
ality disorders. Neuroticism is not included of internalization and externalization iden
as a mental disorder within this authorita tified by Achenbach (1966) many years ago
tive classification of psychopathology, and within childhood psychopathology (Krueger
as a domain of personality functioning that & M arkon, 2 0 0 6 a , 2006b).
is evident within a wide range of the general The broad domain of internalization
population, it is not generally considered by maps well onto the personality temperament
most theorists to be a form of psychopathol of neuroticism (Clark, 20 0 5 ; Krueger 8c
ogy. In sum, neuroticism is considered to be M arkon, 2 0 0 6 b ; Watson, Gamez, & Simms,
a personality trait distinct from psychopa 2005). Kendler, Prescott, Myers, and Neale
thology. Hence it is meaningful to study the (2003) applied multivariate genetic analyses
potential contribution of neuroticism to the to 10 mental disorders assessed in more than
etiology, course, and treatment of various 5 ,6 0 0 members of m ale-m ale and fem ale-
forms of psychopathology. female twin pairs from a population-based
However, an alternative perspective is that registry. They concluded that “the patterns
neuroticism and various forms of psychopa of comorbidity of these disorders (internaliz
thology may actually represent overlapping ing vs. externalizing, and within internaliz
expressions of a common, underlying condi ing, anxious misery vs. fear) is driven large
tion. They may exist along a common spec ly by [common] genetic factors” (p. 936).
trum of functioning. For instance, rather Kahn, Jacobson, Gardner, Prescott, and
than contributing to the etiology of depres Kendler (2005) reported large effect sizes
sion, neuroticism may itself be a form of de for the association of neuroticism with de
pression. Rather than its contributing to the pression, generalized anxiety disorder, and
etiology of a personality disorder, the per panic disorder in a sample of 7,588 twin
sonality disorder may itself be a variant of pairs and found that neuroticism explained
neuroticism. Each of these two hypotheses 2 0 -4 5 % of the comorbidity among depres
will be discussed in turn. sion and the anxiety disorders. Hettema,
Neale, Myers, Prescott, and Kendler (2006)
reported similarly that one-third to two-
Neuroticism on a Spectrum
thirds of the genetic variance in mood (de
with A xis I Psychopathology
pressive) and anxiety disorders was shared
As noted earlier, neuroticism is associ with neuroticism.
ated with a wide variety of Axis I mental To the extent that the relationship of neu
disorders (M alouff et al., 2005). There is roticism to Axis I psychopathology is due
also considerable comorbidity (diagnostic to the presence of a common diathesis, it is
9. N eu ro ticism 141
have the strongest relationship with DSM- (2002), Judge and Ilies (2002), M alouff and
IV personality disorders (Saulsman & Page, colleagues (2 0 0 6 , 2 0 0 7 ), Markon and col
200 4 ), particularly borderline (Widiger, leagues (2005), M unafo and colleagues
2005). (2003, 2005b , 20 0 7 ), O ’Connor (2005),
To the extent that respective DSM -IV O ’Connor and Dyce (1998), Saroglou (2002),
personality disorders are extreme or mal Saulsman and Page (2004), Schinka and col
adaptive variants of the personality trait leagues (2004), Sen and colleagues (2004),
of neuroticism, it may not be particularly Steel and colleagues (2008), and Trapmann
meaningful to study the contribution of neu and colleagues (2007); there are, of course,
roticism to the etiology of these personality many additional meta-analyses beyond the
disorders. How would one understand, for focus of this chapter.
instance, a correlation of the H EXA CO -PI Nevertheless, it is also evident that there is
(Lee & Ashton, 2 0 0 6 ) assessment of emo not, in fact, a complete consensus as to the
tionality with D SM -IV dependent personal optimal conceptualization and assessment
ity disorder when one of the H EXA CO -PI of neuroticism. The alternative formulations
facets of emotionality is explicitly concerned of this domain of personality functioning do
with the assessment of dependency? From appear to involve meaningful differences in
the perspective of Lee and Ashton (2006), conceptualization, and the respective assess
dependency is not a behavior that results ment instruments do yield fundamentally
from neuroticism (or emotionality); it is a different results. Further attention does ap
direct phenotypic expression or manifesta pear to be needed in reaching a consensus
tion of the personality trait of neuroticism. as to the optimal facet structure of neuroti
Some of the symptomatology of a D SM -IV cism, or at least this variation in facet struc
personality disorder can be understood as a ture does need to be appreciated in future
phenotypic expression of an interaction be neuroticism studies.
tween a temperament of negative affectivity The most vibrant area of research is the
and aversive life events (Morey &C Zanarini, contribution of neuroticism to the etiology
2 0 0 0 ; Widiger, 20 0 5 ), but any particular of negative life outcomes. It is suggested
assessment of neuroticism will include the herein, though, that this research will need
assessment of this phenotypic expression, to further consider and distinguish the al
as well as the underlying temperament. In ternative relationships between neuroti
sum, neuroticism, as assessed by any exist cism and these negative life outcomes (i.e.,
ing measure of the FFM , is providing an ex pathoplastic, spectrum, and causal). Cross-
plicit and direct assessment of the respective sectional studies can and do provide quite
personality disorder. informative results, but it is evident that the
most telling findings will be obtained from
longitudinal studies. Neuroticism and psy
Conclusions chopathology, for instance, affect and alter
one another over time in a complex, unfold
It is difficult to imagine there being a per ing interaction. Many vulnerability stud
sonality trait on which there has been a ies have used samples of convenience (e.g.,
greater amount of published research. The persons already in treatment for a respec
sheer number of meta-analyses that have tive disorder) in which the differentiation
been conducted on the relationship of neu among pathoplastic, spectrum, and causal
roticism to other variables is itself a concrete relationships can at times be impossible to
testament as to the importance of this per disentangle. Any particular cross-sectional
sonality trait, or this domain of personal period of time may represent only an ar
ity functioning, within existing research. bitrary and unrepresentative slice along a
Cited within this chapter alone were neu continuously interacting and mutually reaf
roticism meta-analyses conducted by Born firming sequence of events. In sum, neuroti
stein and Cecero (2000), Cassin and von cism continues to provide a compelling un
Ranson (2005), Connor-Smith and Flachs- derstanding of the etiology of negative life
bart (2007), DeNeve and Cooper (1998), events, but this relationship is complex, and
Hoyle and colleagues (2000), Huo-Liang it is through the dismantling of pathoplastic,
(2006), Jorm (1989), Judge, Bono, and col spectrum, and etiological relationships that
leagues (2002), Judge, Heller, and colleagues continued progress will be made.
9. N eu ro ticism 143
Revised (EPQ-R ) and the Eysenck Personality Pro 214). Washington, DC: American Psychological As
filer (EPP). In M. J. Hilsenroth, D. I.. Segal, & M. sociation.
Hersen (Eds.), C o m p r e h e n s iv e h a n d b o o k o f p s y Ready, R. E., 8c Clark, L. A. (2005). Psychiatric patient
c h o lo g ic a l a s s e ss m e n t (Vol. 2 , 9 9 - 1 0 7 ) . New York: and informant reports of patient behavior. J o u r n a l
Wiley. o f P erson ality , 73, 1-21.
M iller,]. D., 8c Pilkonis, P. A. ( 2 0 06 ). Neuroticism and Russell, J. A. (2 003). Core affect and the psychologi
affective instability: The same or different? A m e r i cal construction of emotion. P s y c h o lo g ic a l R ev iew ,
can J o u r n a l o f P sy ch iatry , 163, 8 3 9 - 8 4 5 . 110, 1 4 5 - 1 7 2 .
Morey, L. C., 8c Zanarini, M . C. ( 2 0 0 0). Borderline Saroglou, V. (2002). Religion and the five factors of
personality: Traits and disorder. J o u r n a l o f A b n o r personality: A meta-analytic review. P erson ality
m a l P sy ch olog y , 109, 7 3 3 - 7 3 7 . a n d In d iv id u a l D iffer en c es , 3 2 , 1 5 - 2 5 .
Mullins-Sweatt, S. N., & Widiger, T. A. (20 06 ). The Saucier, G., 8c Goldberg, L. R. (2001). Lexical stud
five-factor model of personality disorder: A transla ies of indigenous personality factors: Premises,
tion across science and practice. In R. F. Krueger & products, and prospects. J o u r n a l o f P erson ality , 69,
J. L. Tackett (Eds.), P erson ality a n d p sy c h o p a th o l- 847-880.
o g y (pp. 3 9 - 7 0 ) . New York: Guilford Press. Saulsman, L. M ., 8c Page, A. C. (200 4). The five-factor
Mullins-Sweatt, S. N., & Widiger, T. A. (20 07 ). The model and personality disorder empirical literature:
Shedler-Westen Assessment Procedure from the A meta-analytic review. C lin ica l P sy ch o lo g y R e
perspective of general personality structure. J o u r view , 2 3 , 1 0 5 5 - 1 0 8 5 .
n a l o f A b n o r m a l P sy ch olog y , 116, 6 1 8 - 6 2 3 . Schinka, J. A. (2005). Measurement scale does moder
Munafo, M. R., Clark, T., 8c Flint, J . (2005a). Does ate the association between the serotonin transport
measurement instrument moderate the associa er gene and trait anxiety: Comments on Munafo et
tion between the serotonin transporter gene and al. M o le c u la r P sychiatry, IQ, 8 9 2 - 8 9 3 .
anxiety-related personality traits?: A meta-analysis. Schinka, J . A., Busch, R. M ., 8c Robichaux-Keene,
M o le c u la r P sy ch iatry , 10, 4 1 5 - 4 1 9. N. (20 0 4). A meta-analysis of the association be
M unafo, M. R., Clark, T., 8c Flint, J . (2005b). Promise tween the serotonin transporter gene polymorphism
and pitfalls in the meta-analvsis of genetic associa (5 -H T T L P R ) and trait anxietv. M o le c u la r P sy ch ia
tion studies: A response to Sen and Schinka. M o try, 9, 1 9 7 - 2 0 2 .
lec u la r P sy ch iatry , 10, 8 9 5 - 8 9 7 . Schmitz, A., Hennig, J ., Kuepper, Y., 8c Reuter, M.
Munafo, M. R., Clark, T. G., Moore, I.. R., Payne, ( 2 0 07 ). The association between neuroticism and
E., Walton, R., 8c Flint, J. (2003). Genetic polymor the serotonin transporter polymorphism depends
phisms and personality in healthy adults: A system on structural differences between personality mea
atic review and meta-analysis. M o le cu la r P sy ch ia sures. P erson ality a n d In d iv id u a l D iffer en c es , 4 2 ,
try, 8, 4 7 1 - 4 8 4 . 789-799.
Munafo, M. R., Clark, T. G., Roberts, K. H., 8c Jo h n Sen, S., Burmeister, M., 8c Ghosh, D. (20 04 ). M eta
stone, E. C. (20 06 ). Neuroticism mediates the as analysis of the association between a serotonin
sociation of the serotonin transporter gene with transporter polymorphism (5-H T T L P R ) and
lifetime major depression. N e u r o p s y c h o b io lo g y , anxiety-related personality traits. A m erica n J o u r
5 3, 1-8. n a l o f M e d ic a l G e n e tic s P art B: N eu r o p sy c h ia tric
Munafo, M. R., Zetteler, J . I., 8c Clark, T. G. (2007). G e n e tic s , I 2 7 B , 8 5 - 8 9 .
Personality and smoking status: A meta-analysis. Sen, S., Burmeister, M ., 8c Ghosh, I). (2005). 5 - H T
N ic o tin e a n d T o b a c c o R e se a rch , 9, 4 0 5 - 4 1 3 . T L P R and anxiety-related personality traits meta
O ’Connor, B. P. (200 5). A search for consensus on analysis revisited: Response to Mu nafo and col
the dimensional structure of personality disorders. leagues. M o le c u la r P sychiatry, 10, 8 9 3 - 8 9 5 .
J o u r n a l o f C lin ic a l P sy ch olog y , 61, 3 2 3 - 3 4 5 . Shedler, J ., 8c Westen, D. (200 4). Dimensions of per
O ’Connor, B. P., 8c Dyce, J . A. (1998). A test of mod sonality pathology: An alternative to the five-factor
els of personality disorder configuration. J o u r n a l o f model. A m erica n jo u r n a l o f P sy ch iatry , 161, 1 7 4 3 -
A b n o r m a l P sy ch olog y , 107, 3 - 1 6 . 1754.
Ozer, D. J., 8c Benet-Martfnez, V. (200 6). Personality Smith, T. W., 8c MacKenzie, J. (200 6). Personality and
and the prediction of consequential outcomes. A n risk of physical illness. A n n u al R e v ie w o f C lin ica l
n u al R e v ie w o f P sy ch olog y , 57, 4 0 1 - 4 2 1 . P sy ch olog y , 2 , 4 3 5 - 4 6 7 .
Parslow, R. A., Jo rm , A. F., 8c Christensen, H. (200 6). Steel, P., Schmidt, J ., 8c Schultz, J . (200 8). Refining
Associations of pre-trauma attributes and trauma the relationship between personality and subjective
exposure with screening positive for PTSD: Analysis well-being. P s y c h o lo g ic a l B u lletin , 134, 1 3 8 - 1 6 1 .
of a community-based study of 2 0 8 5 young adults. Suls, J., 8c Bunde, J. (2005). Anger, anxiety, and de
P sy c h o lo g ic a l M ed icin e, 3 6 , 3 8 7 - 3 9 5 . pression as risk factors for cardiovascular disease:
Penke, L., Denissen, J. J . A., 8c Miller, G. F. (2007). T he problems and implications of overlapping af
T he evolutionary genetics of personality. E u r o p ea n fective dispositions. P sy c h o lo g ic a l B u lletin , 131,
J o u r n a l o f P erson ality , 2 1, 5 4 9 - 5 8 7 . 260-300.
Piedmont, R. L. (2001). Cracking the plaster cast: Big Tellegen, A. (1982). B r ie f m a n u a l f o r th e M u ltid im en
Five personality change during intensive outpatient s io n a l P erson ality Q u estio n n a ir e. Unpublished
counseling. J o u r n a l o f R e se a rc h in P erson ality , 35, manuscript, University of Minnesota, Minneapolis.
500-520. Tellegen, A., 8c Waller, N. G. (1987). E x p lo rin g p e r
Pincus, A. I,. (2 002). Constellations of dependency so n a lity th rou g h test c o n str u c tio n : D e v e lo p m e n t o f
within the five-factor model of personality. In P. T. the M u ltid im e n sio n a l P erson ality Q u estio n n a ire.
Costa 8c T. A. Widiger (Eds.), P erson ality d iso rd er s Unpublished manuscript.
a n d th e fiv e - fa c t o r m o d e l o f p e r s o n a lity (pp. 2 0 3 - ten Have, M ., Oldehinkel, A., Vollebergh, W., 8c
146 III. E M O T IO N A L D IS P O S IT IO N S
Ormel, J. (2005). Does neuroticism explain varia olds, S. K. (200 5). Validation of the UPPS Impulsive
tions in care service use for mental health problems Behaviour scale: A four-factor model of impulsivity.
in the general population?: Results from The Net h E u r o p ea n J o u r n a l o f P erson ality , 19, 5 5 9 - 5 7 4 .
erlands Mental Health Survey and Incidence Study Widiger, T. A. (2001). Social anxiety, social phobia,
(N EM ESIS ). S o c ia l P sy ch iatry a n d P sy ch iatric E p i and avoidant personality disorder. In W. R. Corzier
d e m io lo g y , 4 0 , 4 2 5 - 4 3 1 . 8c L. Alden (Eds.), I n te r n a tio n a l h a n d b o o k o f s o
Trapmann, S., Hell, B., Hirn, J . O. W., 8c Schuler, H. c ia l a n x ie ty (pp. 3 3 5 - 3 5 6 ) . New York: Wiley.
(20 0 7). Meta-analysis of the relationship between Widiger, T. A. (2005). A temperament model of bo r
the Big Five and academic success at university. derline personality disorder. In M. Zanarini (Ed.),
J o u r n a l o f P sy ch olog y , 2 1 5 , 1 3 2 - 1 5 1 . B o rd e rlin e p e r s o n a lity d is o r d e r (pp. 6 3 - 8 1 ) . Wash
Trull, T. J ., 8c Widiger, T. A. (1997). S tru ctu red in te r ington, DC : American Psychiatric Press.
v iew f o r th e fiv e - fa c t o r m o d e l o f p er so n a lity . Odes Widiger, T. A., 8c Clark, L. A. ( 2 0 00 ). Toward D SM-V
sa, FL: Psychological Assessment Resources. and the classification of psychopathology. P sy c h o
Turkheimer, E. ( 2 0 0 0). Three laws of behavior genetics lo g ic a l B u lletin , 1 2 6 , 9 4 6 - 9 6 3 .
and what they mean. C u rren t D irec tio n s in P sy ch o Widiger, T. A., 8c Costa, P. T., Jr. (2002). Five-factor
lo g ic a l S c ien c e , 14, 4 1 0 - 4 1 1 . model personality disorder research. In P. T. Costa,
Van Os, J ., 8c Jones, P. B. (2001). Neuroticism as a risk Jr. 8c T. A. Widiger (Eds.), P erson ality d iso rd er s
factor for schizophrenia. P sy c h o lo g ic a l M ed icin e, a n d th e fiv e - fa c t o r m o d e l o f p e r s o n a lity (2nd ed.,
31, 1 1 2 9 - 1 1 3 4 . pp. 5 9 - 8 7 ) . Washington, DC: American Psychologi
Vitousek, K. M ., 8c Stumpf, R. E. (2005). Difficulties cal Association.
in the assessment of personality traits and disorders Widiger, T. A., 8c Samuel, D. B. (2005). Evidence-
in eating-disordered individuals. E atin g D iso rd ers, based assessment of personality disorders. P sy c h o
13, 3 7 - 6 0 . lo g ic a l A s sessm en t, 17, 2 7 8 - 2 8 7 .
Watson, D. (2 005). Rethinking the mood and an x i Widiger, T. A., 8c Simonsen, E. (2005). Alternative di
ety disorders: A quantitative hierarchical model mensional models of personality disorder: Finding a
for D S M -V. J o u r n a l o f A b n o r m a l P sy ch olog y , 114, common ground. J o u r n a l o f P erson ality D iso r d er s,
522-536. 19, 1 1 0 - 1 3 0 .
Watson, D., 8c Clark, L. A. (1994). M a n u a l f o r the Widiger, T. A., 8c Smith, G. T. (200 8). Personality and
P ositiv e a n d N eg a tiv e A ffe c t S c h ed u le— E x p a n d e d psychopathology. In O. P. Joh n, R. W. Robins, 8c L.
F o rm . Iowa City: University of Iowa. A. Pervin (Eds.), H a n d b o o k o f p e r s o n a lity : T h e o r y
Watson, D., 8c Clark, L. A. (1997). Extraversion and a n d resea rch (3rd ed., pp. 7 4 3 - 7 6 9 ). New York:
its positive emotional core. In R. Hogan, J. Joh nson, Guilford Press.
8c S. Briggs (Eds.), H a n d b o o k o f p e r s o n a lity p s y Widiger, T. A., 8c Trull, T. J. ( 2 00 7). Plate tectonics in
c h o lo g y (pp. 767-793). San Diego, CA: Academic the classification of personality disorder: Shifting to
Press. a dimensional model. A m erica n P sy ch o lo g ist, 6 2,
Watson, D., Clark, L. A., 8c Harkness, A. R. (1994). 71-83.
Structures of personality and their relevance to psy Widiger, T. A., Trull, T. J ., Clarkin, J . F., Sanderson,
chopathology. J o u r n a l o f A b n o r m a l P sy ch olog y , C., 8c Costa, P. T. (2002). A description of the DSM-
10 3 , 1 8 - 3 1 . IV personality disorders with the five-factor model
Watson, D., Clark, L. A., 8c Tellegen, A. (1988). of personality. In P. T. Costa 8c T. A. Widiger (Eds.),
Development and validation of brief measures of P erson ality d iso rd er s a n d th e fiv e - fa c t o r m o d e l o f
positive and negative affect: The PANAS scales. p er so n a lity (2nd ed., pp. 8 9 - 9 9 ) . Washington, DC:
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 5 4 , American Psychological Association.
1063-1070. World Health Organization. (1992). T h e IC D -1 0 c la s
Watson, D., Gamez, W., 8c Simms, L. J . (200 5). Basic s ific a tio n o f m e n ta l a n d b e h a v io u r a l d iso rd er s:
dimensions of temperament and their relation to C lin ic a l d es c rip tio n s a n d d ia g n o s tic g u id elin es . G e
anxiety and depression: A symptom-based per neva, Switzerland: Author.
spective. J o u r n a l o f R e se a rc h in P erson ality , 39, Yamagata, S., Suzuki, A., Ando, J . , One, Y., Kijima,
46-66. N., Yoshimura, K., et al. ( 2 0 0 6). Is the genetic
Watson, D., 8c Tellegen, A. (1985). Toward a consen structure of human personality universal?: A cross-
sual structure of mood. P s y c h o lo g ic a l B u lletin , 9 8 , cultural twin study from North America, Europe,
219-235. and Asia. J o u r n a l o f P erson ality a n d S o c ia l P sy
Westen, D., 8c Shedler, J . (20 0 7). Personality diagno c h o lo g y , 9 0 , 9 8 7 - 9 9 8 .
sis with the Shedler-Westen Assessment Procedure Zuckerman, M. (2002). Zuck erm an-K uhlm an Per
(SWAP): Integrating clinical and statistical mea sonality Questionnaire (ZKPQ): An alternative
surement and prediction. J o u r n a l o f A b n o r m a l P sy five-factorial model. In B. de Raad 8c M. Perugini
c h o lo g y , 116, 8 1 0 - 8 2 2 . (Eds.), B ig F ive a s se ss m e n t (pp. 3 7 7 - 3 9 6 ) . Seattle,
Whiteside, S. P., 8c Lynam, D. R. (2001). The five-factor WA: Hogrefe 8c Huber.
model and impulsivity: Using a structural model of Zuckerman, M. (2003). Biological bases of personal
personality to understand impulsivity. P erson ality ity. In T. Millon, M. J. Lerner, 8c I. B. Weiner (Eds.),
a n d In d iv id u a l D iffe r e n c e s , 3 0 , 6 6 9 - 6 8 9 . H a n d b o o k o f p sy c h o lo g y : Vol. 5. P erso n a lity a n d
Whiteside, S. P., Lynam, D. R., Miller, J . D., 8c Reyn s o c ia l p s y c h o lo g y (pp. 8 5 - 1 1 6 ). New York: Wiley.
CHAPTER 10
Happiness
E d D ie n e r
P e l in K e s e b i r
W il l i a m T o y
147
148 III. E M O T IO N A L D IS P O S IT IO N S
(global judgments of one’s life), satisfaction “so far I have gotten the important things I
with important life domains (e.g., marriage want in life.” Individuals express the degree
or work satisfaction), positive affect (preva to which they agree with these statements
lence of positive emotions and moods), and using a 7-point Likert scale. PANAS, unlike
low levels of negative affect (prevalence of the SW LS, is interested in capturing directly
unpleasant emotions and moods). In many the positive and negative affectivity compo
studies, these dimensions of subjective well nents of subjective well-being. Respondents
being are studied separately, and the differ are given a list of emotion words that sample
ent patterns of the predictors with various positive (e.g., interested, excited, proud)
forms of subjective well-being are exam as well as negative (e.g., distressed, guilty ,
ined. scared) affects and are asked to evaluate on
Prominent conceptualizations of happi a scale from 1 -5 the extent to which they
ness other than subjective well-being in experience these emotions in general. The
clude Ryff and Singer’s (1996) “psychologi directions of PANAS can be rephrased to get
cal well-being” and Ryan and Deci’s (2000) at how the respondents have felt during the
“self-determination theory.” These theories previous week, for example, or how they are
exemplify a less subjective and more pre feeling at the moment. The Subjective Hap
scriptive approach toward happiness in that piness Scale, on the other hand, is an instru
they stipulate the fulfillment of certain needs ment measuring individuals’ perceptions of
(such as autonomy, self-acceptance, or pur how happy they are. Individuals indicate on
pose in life) as a prerequisite for well-being. a Likert scale from 1 to 7 how happy they
Whereas these theories embody valuable consider themselves, and respond to items
contributions to the definition of the good such as “Some people are generally not very
life, researchers working in the subjective happy. Although they are not depressed,
well-being tradition focus their efforts on they never seem as happy as they might be.
understanding people’s own evaluations of To what extent does this characterization
their lives, believing in the meaningfulness describe you?”
and scientific credibility of these evalua Although indispensable to the appraisal
tions. It is important to emphasize at this of something as private and personal as sub
point that individuals’ appraisals of their jective well-being, self-reports of happiness
own well-being hardly reflect empty-headed suffer from the same weaknesses associated
cheerfulness or raw hedonism. To the con with other self-report measures; most nota
trary, major constituents of subjective well bly, an oversensitivity to mood and context
being, such as life satisfaction and positive effects (Schwarz & Strack, 1999). Neverthe
affect, seem to emanate first and foremost less, a great number of studies attest to the
from one’s goals and values. People are most adequate validity of self-report subjective
likely to experience high levels of subjective well-being measures by showing that they
well-being when they strive for and make converge with friend and spousal reports of
progress toward personal goals derived from the individual’s well-being (Lyubomirsky &
their hallowed values, rendering feelings of Lepper, 1999), with recall of satisfying as
meaning, purpose, and fulfillment promi opposed to unsatisfying times in one’s life
nent predictors of subjective well-being (Di- (Pavot, Diener, Colvin, & Sandvik, 1991),
ener & Larsen, 1993). with smiling behavior (Harker & Keltner,
Subjective well-being is typically assessed 2001), and with greater relative left frontal
through self-report measures, such as the activation in the brain (Tomarken, David
Satisfaction with Life Scale (SW LS; Pavot son, & Henriques, 1990). Similarly, tempo
&C Diener, 1993), the Positive and Negative ral stabilities for self-reports of well-being
Affect Schedule (PANAS; Watson, Clark, & have been found to be in the range of 0 .5 -
Tellegen, 1988), or the Subjective Happiness 0.7 over a period of several years (Diener &
Scale (Lyubomirsky & Lepper, 1999). The Suh, 1997). Although multimethod mea
SW LS, for instance, is a five-item instrument surements of subjective well-being should be
that measures global cognitive judgments of undertaken whenever feasible, accumulated
one’s life. The scale includes items such as “in evidence suggests that self-report measures
most ways my life is close to my ideal” and of well-being possess satisfactory validity
10. H appiness 149
and reliability to be employed in happiness legen et al., 1988) testify to the genetically
research. determined part of subjective well-being, as
do findings regarding the relative stability
of happiness over the years (Costa & M c
The A ntecedents o f Happiness Crae, 1988; Magnus & Diener, 1991). It is
widely believed that these genetically deter
In this part of the chapter, we attempt to mined and relatively immutable differences
offer answers to the ever-fascinating ques in responding to people and events set a
tion of how to be happy, drawing on the fixed point for individuals, around which
extant literature on happiness. A general their happiness level fluctuates. According
model specifying the major sources of varia to these set-point theories, major life events
tion in happiness can be useful at the outset. such as the birth of a child or the death of
Lyubomirsky, Sheldon, and Schkade (2005) a partner have only a temporary effect on
have proposed that a person’s chronic hap the person’s happiness level, after which it
piness level is determined by three major reverts to the default level determined by ge
factors: a genetically determined set point netic traits.
for happiness, circumstantial factors (e.g., Closely affiliated with set-point theories
gender, education, culture), and the activi is the “hedonic treadmill theory,” which
ties and practices that the person engages suggests that our emotional systems adjust
in. This model is remarkably similar to Se- to just about anything that happens in our
ligman’s happiness formula, according to lives, positive or negative, just as our noses
which one’s enduring level of happiness is quickly adapt to any kind of scent (Brickman
the sum of (1) one’s set range for happiness, & Campbell, 1971). Early studies showing
(2) life circumstances, and (3) factors under that lottery winners tend to be not much
one’s voluntary control (Seligman, 2002). A happier and that paraplegics tend to be not
survey of the literature suggests that whereas much unhappier than a control group fol
the genetically determined set point accounts lowing an initial adjustment period (Brick
for about 50% of variation in happiness, life man, Coates, & Janoff-Bulm an, 1978) have
circumstances account for only 10% , and been broadly cited to illustrate the powerful
intentional activities are responsible for the role of adaptation in happiness. Set-point
remaining 4 0 % (Lyubomirsky et al., 2005). theory, in conjunction with the hedonic
In our review of the causes and correlates of treadmill idea, implies that individual and
happiness, we start with the genetic determi societal attempts at increasing happiness are
nants of happiness and move on to circum ultimately doomed to failure. In their paper
stantial and demographic factors (e.g., age, documenting the high heritability coefficient
gender, intelligence, religion), finally turn of happiness, Lykken and Tellegen (1996)
ing to the antecedents of happiness that are have indeed noted that trying to be happier
relatively more amenable to individual con may be “as futile as trying to be taller and
trol (e.g., social relationships, goals, leisure). therefore is counterproductive” (p. 189).
Without a doubt, these three sources of hap However, findings from longitudinal and
piness are not fully independent from each cross-sectional studies, as well as from in
other, yet, in our view, they provide a fairly tervention research, fail to corroborate such
accurate and useful schema for understand pessimistic conclusions. These findings sug
ing the antecedents of happiness. gest that time may be ripe for a revision of
the hedonic adaptation theories of well-being
(Diener, Lucas, & Scollon, 2 0 0 6 ; Easterlin,
G enes and H appiness Set Point
20 0 6). People do not rapidly and completely
There is virtually no dispute among scholars adapt to everything life has in store for them,
that genetic inheritance plays a significant and this fact is powerfully demonstrated by
role in determining one’s chronic happiness differences in average national happiness
level. Studies demonstrating that identical levels. Factors such as wealth, human rights,
twins are considerably more similar to each and societal equality significantly predict
other in their happiness levels than frater well-being in a society, which means that
nal twins (Lykken & Tellegen, 1996; Tel people do not automatically adapt to any
150 III. E M O T IO N A L D IS P O S IT IO N S
objective life condition (Diener, Diener, & punishments. Specifically, extraverts are
Diener, 1995). Similarly, Fujita and Diener more responsive to positive mood induc
(2005) have found in a large German sample tions, whereas neurotics are more responsive
that over a 17-year period, almost 9% of to negative mood inductions (Derryberry &
the sample changed an average of 3 or more Reed, 1994; Larsen &C Ketelaar, 1991). In
points on a 10-point scale from the first 5 to addition to this direct effect of personality
the last 5 years of the study and that average traits on happiness, studies have also un
life satisfaction in the first 5 years correlated covered an indirect route in that extraverted
only .51 with average life satisfaction during people experience more frequent positive ob
the last 5 years. Other studies investigating jective life events and neurotic people experi
the longitudinal effects of unemployment ence more frequent negative objective events
(Lucas, Clark, Georgellis, & Diener, 20 0 4 ), (Headey & Wearing, 1989; Magnus, Diener,
marriage (Lucas, Clark, Georgellis, &C Di Fujita, &c Pavot, 1993).
ener, 2003), and even winning the lottery Other than extraversion and neuroticism,
(Gardner & Oswald, 2 0 0 7 ) on life satisfac personality traits such as dispositional op
tion levels confirm the view that set-point timism, trust, agreeableness, desire for con
and adaptation theories as they are typically trol, and hardiness have been found to be
conceived do not coincide with the current positively associated with happiness (DeNeve
empirical findings and need to be modified. & Cooper, 1998; Lucas, Diener, & Suh,
In short, there seems to be a substantial 1996; Scheier & Carver, 1993; Watson &
genetic component to subjective well-being, Clark, 1992). Another personality trait that
which contributes to the relative stability of is closely related to happiness is self-esteem.
subjective well-being over a person’s lifespan Research has consistently revealed moderate
and makes some people more prone to hap to high correlations between self-esteem and
piness and others to unhappiness. Even so, happiness (Lyubomirsky, Tkach, & DiMat-
only about half of the individual differences teo, 2006). It is worth noting, however, that
in happiness are accounted for by genetic these correlations are demonstrated to be
influences, meaning that people are hardly significantly stronger in individualist com
doomed to miserable lives as the victims of the pared with collectivist cultures (Diener &
genetic happiness lottery. Genes affect one’s Diener, 1995a). Furthermore, the direction
happiness through their expression in dispo of causality between the two constructs is
sitional patterns and personality characteris not entirely understood (Baumeister, Camp
tics, which is the subject we turn to next. bell, Krueger, & Vohs, 2003).
Intimately related to personality charac
teristics that are correlated with high levels
Personality
of happiness are what some have called “vir
Among different facets of personality, extra tues and character strengths.” Recently, a
version and neuroticism are the ones most number of psychologists undertook the mas
consistently and strongly related to happi sive project of coming up with an exhaustive
ness (Diener & Lucas, 1999; Rusting &C Lar list of virtues, and their efforts culminated in
sen, 1997). As expected, both of these traits a classification system made up of 24 char
are highly heritable, rooted in neurobiology, acter strengths, organized under six core
and exhibit little change over the lifespan virtues (Peterson & Seligman, 2004). These
(Lyubomirsky et al., 2005). A host of stud six core virtues are w isdom (e.g., love of
ies show that extraversion predicts positive learning, creativity), courage (e.g., bravery,
affect moderately to strongly (e.g., Lucas & persistence), humanity (e.g., kindness, so
Fujita, 2 0 0 0 ), whereas neuroticism is an ex cial intelligence), justice (e.g., fairness), tem
ceptionally strong predictor of negative af perance (e.g., forgiveness, self-regulation),
fect (e.g., Fujita, 1991). The exact processes and finally, transcendence (e.g., gratitude,
that underlie the extraversion-happiness religiousness/spirituality). Research has re
and neuroticism-unhappiness links have vealed that the character strengths of hope,
also been broadly explored. One such pro zest, gratitude, love, and curiosity are most
cess seems to be the differential sensitivity strongly and robustly linked to life satisfac
of extraverts and neurotics to rewards and tion. More cerebral virtues, such as love of
10. H appiness 151
learning, on the other hand, seem to be only vik (1991) have demonstrated that whereas
weakly associated with happiness (Park, Pe gender accounts for less than 1% of the vari
terson, & Seligman, 2004). ance in happiness, it accounts for over 13%
of the variance in the intensity of emotional
experiences. In other words, women and
A ge
men do not differ in their average happiness
O f the relation between age and happiness, levels, though women may be overrepresent
Tatarkiewicz (1976) confidently wrote, “it ed among both the extremely happy and the
is considered to be an elementary truth that extremely unhappy members of society (Di
happiness is the privilege of youth” (p. 165). ener et al., 1999).
Studies, however, make it plain that al
though young people are generally happy,
Intelligence and Education
happiness is hardly their exclusive privilege.
Longitudinal and cross-sectional data sug “By all the gods above,” wrote Dutch phi
gest that, of the three components of well losopher Erasmus, “is anyone happier than
being, positive affect slightly decreases in the sort of men who are usually called fools,
old age, yet so does negative affect (Charles, dolts, simpletons, nincompoops?” (2003,
Reynolds, & Gatz, 2 0 0 1 ; Mroczek & Spiro, p. 54). Studies, however, fail to validate this
2005). As to life satisfaction, Mroczek and observation and point to a positive (though
Spiro (2005) found that although there were weak) correlation between one’s level of ed
significant individual differences, life sat ucation and happiness after controlling for
isfaction increased from age 4 0 to 65, but other variables, explaining 1 -3 % of variance
then declined, particularly with impending in happiness (Witter, Okun, Stock, & H ar
death. Though more research on the subject ing, 1984). As to the effect of intelligence (as
is required, these results alone warn against measured by IQ tests) on happiness, it seems
a view of old age as a wellspring of unhappi to be very weak, if it exists at all. Emotional
ness and against oversimplified conclusions intelligence, on the other hand, has consis
about age trends in subjective well-being. tently been linked to happiness (Furnham &
Petrides, 2 0 0 3 ; Schutte, M alouff, Simunek,
McKenley, &c Hollander, 2002), most likely
G ender
because neurotic individuals tend to score
In his famous essay “On Women,” Scho low on measures of social and emotional in
penhauer (2004) argued that “the keenest telligence.
sorrows and joys” are not for a woman,
that “the current of her life should be more
Wealth
gentle, peaceful and trivial than man’s,
without being essentially happier or unhap- All in all, research suggests that money has
pier” (p. 51). Large-scale surveys dovetail a positive, yet diminishing, effect on happi
with Schopenhauer’s insight that women are ness. Although increased income contributes
not significantly happier or unhappier than significantly to happiness at low levels of de
men. When sex differences are observed in velopment across nations, the strong link be
studies, it is typically women who report tween wealth and life satisfaction appears to
higher happiness levels, yet these differences taper off at higher levels of income (Frey &
tend to disappear when other demographic Stutzer, 2002a). Reflecting this trend, when
variables are controlled for (Diener, Suh, Diener, Horowitz, and FLmmons (1985)
Lucas, & Smith, 1999). Schopenhauer’s ob asked wealthy people chosen from the Forbes
servation that women do not experience the list of the wealthiest Americans about their
greatest sufferings and the greatest joys, on happiness levels, they reported being only
the other hand, seems to be a poor reflec modestly happier than a comparable group,
tion of reality. Data indicate that, quite to and 37% of them turned out to be less happy
the contrary, women experience both nega than the average American. Whereas having
tive and positive emotions more frequently money is associated with a positive, albeit
and more intensely than men. In line with diminishing, effect on happiness, wanting
this observation, Fujita, Diener, and Sand- money too much has repeatedly been shown
152 III. E M O T IO N A L D IS P O S IT IO N S
to prove toxic to happiness. People who place among the poorest. National wealth is also
a lot of importance on money and on materi highly correlated with various social indica
al possessions, particularly to the expense of tors, such as democratic governance, human
family and social relationships, tend to feel rights, and longevity (Diener & Diener,
less satisfied with their lives and experience 1995b), which may partly account for its as
less positive affect and more negative affect sociation with subjective well-being. Societ
(Kasser & Kanner, 2004). ies also have differing norms regarding the
desirability of happiness and the appropriate
expression of positive and negative emotions
Religion
that contribute to cross-cultural subjective
A number of studies point to a positive yet well-being differences beyond the effect of
modest effect of religion on happiness. More economic development. For example, rela
specifically, participation in religious servic tive to other cultures, Confucian cultures
es, strength of religious affiliation, relation (such as China) regard the ideal level of life
ship with God, and prayer have all been as satisfaction as one of neutrality and display
sociated with greater happiness levels (e.g., higher acceptance of negative emotions and
Ferriss, 2 0 0 2 ; Poloma & Pendleton, 1990; lower acceptance of positive emotions. Ex
Witter, Stock, Okun, & Haring, 1985). tant norms regarding life satisfaction in a
Higher levels of religiosity have also been society are apparently mirrored in actual
linked to higher life satisfaction and lower levels of life satisfaction in that society, as
rates of suicide across nations (Diener &C Se- confirmed by the finding that the mean ideal
ligman, 2 0 0 4 ; Helliwell, 20 0 7 ). It is believed level for life satisfaction correlates .73 with
that the beneficial effects of religion on hap mean reported life satisfaction across na
piness stem largely from the sense of mean tions (Diener & Suh, 1999). The variables
ing and purpose that religious beliefs provide that most influence subjective well-being are
to the individual, as well as from the social also moderated by culture. For example, as
support networks associated with organized mentioned earlier, self-esteem is a stronger
religion (e.g., churches). Importantly, it is predictor of subjective well-being in indi
an intrinsic as opposed to extrinsic orienta vidualistic cultures than in collectivistic
tion toward religion that seems to be asso cultures. In line with this finding, people
ciated positively with subjective well-being in individualistic cultures tend to base their
(Ardelt, 2 0 0 3 ; Ardelt & Koenig, 20 0 7 ). It life satisfaction judgments on personal emo
is also worth noting that the positive link tional experiences, whereas people from col
between religion and happiness is stronger lectivistic cultures emphasize the appraisals
for women, African Americans, and older of others (Suh, Diener, Oishi, & Triandis,
adults and for Americans compared with 1998).
Europeans (Argyle, 1999). Religious people
in certain countries (e.g., Lithuania, Slova
Health
kia) have even reported lower levels of life
satisfaction, which highlights the need for Physical health inarguably affects well
further research in order to understand the being, as is evidenced by the considerably
exact nature of the relationship between lowered happiness levels of individuals who
happiness and religion. The link of spiritu suffer life-threatening illnesses or illnesses
ality— as a concept distinct from religious that interfere with their daily lives and cause
ness— to subjective well-being is a similarly pain. Given this fact, it is intriguing that re
unstudied topic. searchers have reported weak and sometimes
nonexistent correlations between happiness
and objective health as assessed by medical
Societal Conditions and Culture
personnel. Whereas associations between
International surveys of happiness reveal objective health and happiness are often
significant mean differences across societ weak, research documents that associations
ies (Diener & Suh, 20 0 0 ). These differences between happiness and subjective health— as
are substantially explained by the level of it is reported by the individual— are consis
economic development in a country: Some tently strong (Okun, Stock, Haring, & W it
of the unhappiest nations tend also to be ter, 1984). This curious phenomenon seems
10. H appiness 153
to be the consequence of (1) clinical error, born children significantly increase the hap
meaning that objective health measures are piness of their parents, whereas additional
sometimes not as objective as one would children reduce the happiness of their moth
hope, and (2) the notion that subjective re ers and leave the happiness of their fathers
ports of health reflect emotional adjustments unchanged. Another remarkable finding
on the part of the individual, thus inflating from their study was that having had chil
the correlation between self-reported health dren at one point in their lives did not have
and happiness. any effect on the happiness levels of men and
women at ages 5 0 -7 0 .
Social R elationships and F riend s
Goals and Sense o f M eaning
Having close friends and a network of social
support has a distinct positive effect on hap Research findings unambiguously illustrate
piness, to such a degree that some scholars that striving for and making progress toward
have suggested that this could be the single meaningful, enjoyable, moderately challeng
most important source of happiness (Reis ing goals is an important source of happi
& Gable, 2003). Corroborating this view, ness (Brunstein, 1993; Emmons, 1986; Lit
Diener and Seligman (2002) found in their tle, 1989). As Myers and Diener (1995) have
study of very happy people that every single suggested, happiness seems to grow “less
one of them had excellent social relation from the passive experience of desirable cir
ships. Other studies document that those who cumstances than from involvement in valued
enjoy close relationships are better at coping activities and progress toward one’s goals”
with major life stresses such as bereavement, (p. 17). Individuals who have goals that they
rape, unemployment, and illness (Myers, deem important tend to be more energetic,
1999), and perceived loneliness is robustly experience more positive affect, and feel that
linked to depression (Anderson & Arnoult, life is meaningful (Diener, Lucas, & Oishi,
1985). It is not to be forgotten, though, that 2002). Interestingly, positive affect in itself
happiness itself may lead to better relation has been found to predispose people to feel
ships. As we show later, happy people tend that life is meaningful (King, Hicks, Krull,
to be more outgoing, empathic, and trusting & Del Gaiso, 2006).
than unhappy people, presumably resulting
in enhanced quantity and quality of social
Leisure
relationships (Veenhoven, 1988).
George Bernard Shaw once observed that the
only way to avoid being miserable is not to
M arriage and Children
have enough leisure time to wonder whether
Empirical research regarding the relation you’re happy or not. A host of studies none
ship between happiness and marriage in theless document that leisure activities such
the last few decades has yielded the robust as music, exercise, and reading significant
finding that married individuals tend to be ly contribute to happiness (Argyle, 2002).
happier than unmarried or divorced ones Balatsky and Diener (1993) even reported
(e.g., Gove & Shin, 1989; W hite, 1992). We that, among Russian students, leisure satis
should again be cautioned that the arrow of faction was the single best predictor of hap
causality may point both ways: A number of piness. On a related note, people who work
studies have revealed that individuals who fewer hours have been demonstrated to have
are likely to get married and to stay married higher life satisfaction (Alesina, Glaeser, &
are happier long before the marriage com Sacerdote, 2006).
pared with individuals who remain single
(Lucas et al., 2003). Investigations about
the effects of having children on one’s hap The Consequences o f Happiness
piness have been rarer, yet available data do
not unequivocally support the conventional In the previous section, we attempted to
view that children are “the joy of life.” In paint a rough picture of what causes happi
a well-controlled study, Kohler, Behrman, ness. Our assumption throughout was that
and Skytthe (2004) documented that first happiness is something highly valuable, not
154 III. E M O T IO N A L D IS P O S IT IO N S
withstanding the aforementioned cultural problem, positive emotions and general well
differences in the perception of its desirabil being produce a readiness to explore the en
ity. Indeed, in a recent study conducted in vironment and approach new goals, thereby
48 nations, Diener and Oishi (2006) estab building enduring personal resources.
lished that respondents rated the importance The notion that happy moods render the
of happiness 8.03 on a 9-point scale, higher world an easier and safer place for people to
than the importance of any of the other 11 deal with is evidenced, for example, by Prof
attributes included in the survey, such as fitt’s (2006) finding that participants who
success, intelligence/knowledge, or material were put in a bad mood estimated the slope
wealth. Others have found that, in America, of a hill dropping down in front of them to be
happiness is rated as more relevant to the much steeper than they did when they were
judgment of a good life than wealth or moral put in a good mood. Individuals are hence
goodness, and happy people are deemed expected to generally perform better when
to be more likely to go to heaven (King & they are in good moods. In one test of the
Napa, 1998). “sadder-but-wiser vs. happier-and-smarter”
Happiness indisputably feels good, and hypotheses, Staw and Barsade (1993) as
people value it greatly, yet the question that sessed the positive affect levels of first-year
remains to be answered is whether happiness M BA students and found that positive affect
is as justifiable as it is desirable. The answer, significantly predicted decision-making ac
according to happiness research, seems to curacy, mastery of information, leadership,
be emphatically positive. A fascinating dis and ratings of managerial performance,
covery made recently by happiness scholars after controlling for the effects of Gradu
is that happiness is not only an epiphenom- ate Management Admission Test (GMAT)
enon but also plays a causal role in bringing scores, age, gender, and years of experience.
about a plethora of individually and socially These findings are further supported by data
beneficial outcomes. In the next section, we showing that individuals experimentally put
provide a review of how happiness cultivates in pleasant moods outperform others in var
better health and achievement outcomes, ious tasks, including efficient decision mak
better social relationships, and elevated ing (Forgas, 1989) or anagram solving (Erez
degrees of prosocial behavior. For a more & Isen, 2002), and they also persist longer at
comprehensive review, the reader is advised tasks that require perseverance (Kavanagh,
to refer to Lyubomirsky, King, and Diener 1987).
(2005). At the same time, however, some stud
ies show that those experiencing elevated
moods have an increased tendency to rely on
Benefits o f H appiness
heuristics. Heuristics are learned answers,
f o r A chievem ent O utcomes
or mental habits, that help people effortless
Whereas many romanticizers of unhap ly answer problems that are frequently faced
piness disparage happiness for dumbing in life. When they are used in an appropriate
people down and praise misery for its role context, they can efficiently yield accurate
in sharpening one’s mental faculties, the pic answers. However, high levels of positive af
ture emerging from available data is that it fect might lead to inappropriate use of heu
is rather happiness (and not unhappiness) ristics, possibly because good moods serve as
that leads to the development and better a cue that everything is going well and that
use of intellectual skills. Barbara Fredrick there is no need for the expenditure of extra
son’s “broaden-and-build theory” provides mental energy (Schwarz, Bless, Wanke, &C
a valuable framework to make sense of this W inkielman, 2003). In harmony with such
phenomenon. According to this theory (Fre an interpretation, studies show that people
drickson, 1998, 2001), positive emotions who are in a good mood perform as well as
allow individuals to broaden their thought- those who are not when they are reminded
action repertoires and build their intellectual, that the task is important or complicated
psychological, social, and physical resources (Lyubomirsky, King, &C Diener, 2005).
over time. Whereas negative emotions, such Happiness has also been linked to higher
as fear or anger, appropriately cause the in achievement in professional life. It has been
dividual to focus on the immediate threat or documented, for instance, that happier indi
10. H appiness 155
viduals are more likely to graduate from col pier people are more likely to be commu
lege, more likely to secure jobs, more likely nity volunteers and to invest more hours in
to have more prestigious jobs, more likely volunteer work (Thoits & Hewitt, 2001).
to receive favorable evaluations from their O f great significance is the fact that happi
supervisors, more likely to find their jobs ness has also been shown to increase ethical
more meaningful, less likely to lose their judgments: When Jam es and Chymis (2004)
jobs, quicker to be reemployed if they do, analyzed how justifiable respondents found
more likely to exhibit organizational citizen various ethical scenarios— such as cheating
ship behaviors, and, finally, more likely to on taxes if one has a chance or avoiding a
earn higher incomes (e.g., Borman, Penner, fare on public transport— those with higher
Allen, & Motowildo, 20 0 1 ; Cropanzano & happiness levels responded in more ethical
Wright, 1999; Diener, Nickerson, Lucas, & ways. This led the authors to conclude that
Sandvik, 2 0 0 2 ; M arks & Fleming, 1999; improving subjective well-being may play a
Roberts, Caspi, & M offitt, 2 0 0 3 ; Verkley & significant role in reducing improbity of all
Stolk, 1989). On the whole, available data kinds (e.g., corruption, criminality) nation
strongly suggest that happiness is not only ally and worldwide. Inglehart and Klinge-
a product of achievement but, at the same mann’s (2000) similar argument that gen
time, a producer of it. Research has yet to eral well-being is a harbinger of democratic
unveil all the mechanisms through which governance is corroborated by Tov and Di-
these effects are obtained. ener’s (2008) finding that on a national level,
happier countries tend to be higher on gen
eralized trust, volunteerism, and democratic
Benefits o f H appiness f o r Social
attitudes. These findings sharply contradict
Relationships and Prosocial Behavior
a view of happiness as self-indulgent hedo
Whereas some have argued that only a self- nism and attest to the intimate connection
centered person blind to the overwhelming between a moral life and a happy life de
suffering permeating the world could ever fended by many a philosopher throughout
be happy, research fails to justify the cyni the ages.
cism in these beliefs. Quite to the contrary,
what studies reveal is that happiness tends
Benefits o f H appiness f o r Health
to bring out the best in humans, rendering
them more social, cooperative, and even eth Accumulating evidence suggests that subjec
ical. People with chronically high or experi tive well-being affects physical health and
mentally increased positive affect have been longevity, endorsing the Biblical notion that
observed to judge persons they have recent “a merry heart does good like a medicine”
ly met in a more positive light, to become (Pressman & Cohen, 2005). Whereas it had
more interested in social interaction, and long been established that high levels of
also more prone to self-disclosure (Berry & negative emotion (e.g., stress, anger) are as
Hansen, 1996; Cunningham, 1988; Mayer, sociated with lowered immune functioning
Mamberg, & Volanth, 1988). Experimen and coronary heart disease, less was known
tally induced positive affect also increases until recently about the powerful protective
trust in others (Dunn & Schweitzer, 2005), influence that positive emotions exert. In a
which may partly help to explain the classic remarkable study revealing this influence,
finding that positive moods increase helping Danner, Snowdon and Friesen (2001) es
behavior (Isen & Levin, 1972). In a similar tablished that positive emotional content in
vein, those who report higher life satisfac handwritten autobiographies of Catholic sis
tion exhibit more generalized trust in oth ters, composed when they were at the mean
ers (Brehm & Rahn, 1997), which in turn age of 22 years, predicted their longevity six
predicts not only individual but also societal decades later. In this study, the nuns in the
well-being. highest quartile regarding the number of
The view that a virtuous cycle exists be positive emotion words (e.g., happy, good,
tween happiness and myriad socially desir fun) lived on average 9.4 years longer than
able outcomes is further substantiated by the nuns in the lowest quartile. In another
the finding that not only does volunteer study, participants were experimentally in
ing increase well-being but also that hap fected with a cold virus and then monitored
156 III. E M O T IO N A L D IS P O S IT IO N S
daily in quarantine. As anticipated, individ lack of negative emotions can be highly dan
uals who reported experiencing high levels gerous, as exemplified by psychopaths, could
of positive emotions (i.e., those that were an excessive amount of positive affect also
happy, pleased, relaxed) turned out to be result in suboptimal outcomes? We know, for
much less vulnerable to the common cold example, that people in a good mood tend to
(Cohen, Doyle, Turner, Alper, & Skoner, rely more on heuristics than people in a bad
2003) than those who reported experiencing or neutral mood, which also explains their
low levels of positive emotions. Marsland, more frequent use of stereotypes in person-
Cohen, Rabin, and Manuck (2006), inter perception tasks (Bodenhausen, Kramer, &
estingly, found that positive affect was a Siisser, 1994). Intrigued by the notion of op
stronger predictor of immune strength than timal happiness, Oishi, Diener, and Lucas
negative affect, and its predictive power per (2007) put to test the idea that once people
sisted when demographics and body mass are moderately happy, the most effective
were controlled. Studies showing that peo level of happiness may depend on the life do
ple put into a pleasant mood exhibit greater main under question. They found those who
pain tolerance compared with control par experience the highest levels of happiness to
ticipants also provide evidence for the fa be more successful in the domain of close
vorable impact of positive affect on health relationships and volunteering. People who
outcomes (Zelman, Howland, Nichols, & reported slightly lower levels of happiness,
Cleeland, 1991). on the other hand, were the most success
ful ones in terms of income, education, and
political participation. These findings imply
Conclusion that, whereas happy people in general fare
much better than unhappy people, the level
In this chapter we aimed to shed some light of most desirable happiness depends on an
on what influences happiness and what in individual’s value priorities.
turn is influenced by it, based on four de As scholars of happiness, we are tremen
cades of research. These decades of accumu dously delighted that various knowledge dis
lated research have revealed that happiness ciplines, from philosophy (Haybron, 2007)
is not only universally desired, but justifi to economics (Frey & Stutzer, 2002b) to
ably so. We have learned that happiness is a neuroscience (Klein, 20 0 6 ), have recently
worthwhile pursuit, because it functions as a started to exhibit a serious interest in this
resource that people unwittingly draw from once marginalized subject. We cannot wait
in their endeavors toward higher levels of to see the future of happiness studies shaped
success, kindness, and health. Therefore, at by this multidisciplinary effort, and we con
tempts at increasing happiness take on great tinue to hope for a tomorrow in which peo
importance, not only for individuals but also ple will be optimally happy.
for societies. Fortunately, the science of hap
piness has shown and continues to show the
empirically validated ways to increase hap R eferen ces
piness. We know that although some part of
Alesina, A., Glaeser, E. L., & Sacerdote, B. (2006).
our capacity for happiness is inherited and
Work and leisure in the U.S. and Europe: Why so
simply not amenable to change, we can still different? In M. Gertler & K. Rogoff (Eds.), N B E R
choose to do certain things that will make M a c r o e c o n o m ic s A n n u al 2 0 0 5 (pp. 1 - 6 4 ) . C a m
us lastingly happier, such as counting our bridge, M A : M I T Press.
blessings (Emmons & McCullough, 2003) Anderson, C. A., & Arnoult, L. H. (1985). Attribu-
tional style and everyday problems in living: De
or stopping to smell the roses (Bryant &
pression, loneliness, and shyness. S o c ia l C o g n itio n ,
Veroff, 2006). 3, 16-35.
Our review of the literature suggests that Ardelt, M . (200 3). Effects of religion and purpose in
happiness not only is a reward in itself but life on elders’ subjective well-being and attitudes to
ward death .J o u r n a l o f R elig io u s G e r o n to lo g y , 14,
also brings about various individually and
55-77.
socially desirable outcomes. Given that hap Ardelt, M ., & Koenig, C. S. (200 7). The importance
piness is functional, the optimal level of hap of religious orientation in dying well: Evidence from
piness becomes an essential matter for indi three case studies. J o u r n a l o f R elig io n , S p iritu ality
vidual and societal reasons. If an extreme a n d A gin g, 19, 61 -7 9 .
10. H appiness 157
Argyle, M. (1999). Causes and correlates of happiness. from the nun study. J o u r n a l o f P erson ality a n d S o
In D. Kahneman, E. Diener, 8c N. Schwarz (Eds.), cia l P sy ch olog y , 8 0 (5 ), 8 0 4 - 8 1 3 .
W ell-b ein g : T h e fo u n d a t io n s o f h e d o n ic p s y c h o l DeNeve, K. M ., Sc Cooper, H. (1998). The happy per
o g y (pp. 5 5 3 —373). New York: Russell Sage Foun sonality: A meta-analysis of 137 personality traits
dation. and subjective well-being. P sy c h o lo g ic a l B ulletin ,
Argyle, M . (2002). T h e p s y c h o lo g y o f h a p p in es s. L on 124, 1 9 7 - 2 2 9 .
don: Routledge. Derryberry, D., Sc Reed, M. A. (1994). Temperament
Balatsky, G., Sc Diener, E. (1993). Subjective well and attention: Orienting toward and away from
being among Russian students. S o c ia l In d ic a to r s positive and negative signals. J o u r n a l o f P erson ality
R esea rc h , 2 8 , 2 2 5 - 2 4 3 . a n d S o c ia l P sy ch olog y , 6 6 , 1 1 2 8 - 1 1 3 9 .
Baumeister, R. F., Campbell, J. D., Krueger, J. I., 8c Diener, FI. (1984). Subjective well-being. P sy c h o lo g ic a l
Vohs, K. D. (2003). Does high self-esteem cause bet B u lletin , 95, 5 4 2 - 5 7 5 .
ter performance, interpersonal success, happiness, Diener, E., 8c Diener, M. (1995a). Cross-cultural co r
or healthier lifestyles? P sy c h o lo g ic a l S cien ce in the relates of life satisfaction and self-esteem. J o u r n a l o f
P u b lic In terest, 4 , 1 - 4 4 . P erson ality a n d S o c ia l P sy ch olog y , 6 8 , 6 5 3 - 6 6 3 .
Berry, D. S., 8c Hansen, J. S. (1996). Positive affect, Diener, E., Sc Diener, C. (1995b). The wealth of na
negative affect, and social interaction. J o u r n a l o f tions revisited: Income and quality of life. S o c ia l
P erson ality a n d S o c ia l P sy ch olog y , 71, 7 9 6 - 8 0 9 . In d ic a to r s R e se a rc h , 3 6 , 2 7 5 - 2 2 8 .
Bodenhausen, G. V., Kramer, G. P., 8c Siisser, K. Diener, E., Diener, M ., Sc Diener, C. (1995). Factors
(1994). Happiness and stereotypic thinking in so predicting the subjective well-being of nations.
cial judgment. J o u r n a l o f P erson ality a n d S o cia l J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 69,
P sy ch olog y , 6 6 , 6 2 1 - 6 3 2 . 851-864.
Borm an, W. C., Penner, L. A., Allen, T. D., & Motowi- Diener, F.., Horowitz, J ., Sc Emmons, R. A. (1985).
ldo, S. J . (2001). Personality predictors of citizen Happiness of the very wealthy. S o c ia l In d ic a to r s
ship performance. In te r n a tio n a l J o u r n a l o f S e le c R e se a rch , 16, 2 6 3 - 2 7 4 .
tion a n d A ssessm en t, 9, 5 2 - 6 9 . Diener, FI., Sc Larsen, R. J. (1993). The experience of
Brehm, J ., 8c R ahn, W. (1997). Individual-level evi emotional well-being. In M. Lewis Sc J. M . Havi-
dence for the causes and consequences of social cap land (Eds.), H a n d b o o k o f e m o tio n s (pp. 4 0 4 - 4 1 5 ) .
ital. A m er ic a n J o u r n a l o f P o litic a l S cien ce, 41(3), New York: Guilford Press.
999-1024. Diener, E., Lucas, R., 8c Oishi, S. (2002). Subjective
Brickman, P., 8c Campbell, D. T. (1971). Hedonic rela well-being: The science of happiness and life satis
tivism and planning the good society. In M . H. Ap- faction. In C. R. Snyder 8c S. J. Lopez (Eds.), T h e
pley (Ed.), A d a p ta tio n lev e l th e o r y : A sy m p o siu m h a n d b o o k o f p o s itiv e p s y c h o lo g y (pp. 6 3 - 7 3 ) . New
(pp. 2 8 7 - 3 0 2 ) . New York: Academic Press. York: Oxford University Press.
Brickman, P., Coates, D., 8c Jano ff-Bulman, R. (1978). Diener, FI., 8c Lucas, R. FI. (1999). Personality and sub
Lottery winners and accident victims: Is happiness jective well-being. In D. Kahneman, E. Diener, 8c
relative? J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l N. Schwarz (Eds.), W ell-b ein g : T h e fo u n d a t io n s o f
og y , 3 6 , 9 1 7 - 9 2 7 . h e d o n ic p s y c h o lo g y (pp. 2 1 3 - 2 2 9 ) . New York: Rus
Brunstein, J. C. (1993). Personal goals and subjective sell Sage Foundation.
well-being: A longitudinal study. J o u r n a l o f P ers o n Diener, E., Lucas, R. E., 8c Scollon, C. N. (200 6 ). B e
ality a n d S o c ia l P sy ch olog y , 6 5 , 1 0 6 1 - 1 0 7 0 . yond the hedonic treadmill: Revisions to the adapta
Bryant, F. B., Si Veroff, J. (200 6). S av orin g : A n ew tion theory of well-being. A m er ica n P sy ch olog ist,
m o d e l o f p o s itiv e e x p e r ie n c e . Mahwah, NJ: Erl 61, 3 0 5 - 3 1 4 .
baum. Diener, E., Nickerson, C., Lucas, R. F.., 8c Sandvik, E.
Charles, S. T., Reynolds, C. A., 8c Gatz, M . (2001). (2002). Dispositional affect and job outcomes. So-
Age-related differences and change in positive and c ia l In d ic a to r s R e se a rc h , 59, 2 2 9 - 2 5 9 .
negative affect over 23 years. J o u r n a l o f P erson ality Diener, K., 8c Oishi, S. ( 2 0 06 ). T h e d esira b ility o f h a p
a n d S o c ia l P sy c h o lo g y , 8 0 , 1 3 6 - 1 5 1 . p in es s a c r o s s cu ltu res. Unpublished manuscript,
Cohen, S., Doyle, W. J . , Turner, R. B., Alper, C. M., University of Illinois, Urbana-Champaign.
Sc Skoner, D. P. (2 003). Emotional style and suscep Diener, F,., 8c Seligman, M. E. P. (2 002). Very happy
tibility to the common cold. P sy c h o s o m a tic M e d i people. P sy c h o lo g ic a l S cien ce, 13, 8 1 - 8 4 .
cin e, 6 5 , 6 5 2 - 6 5 7 . Diener, E., Sc Seligman, M. E. P. (200 4). Beyond
Costa, P. T., Sc McCrae, R. R. (1988). Personality in money: Toward an economy of well-being. P sy ch o
adulthood: A six-year longitudinal study of self- lo g ic a l S cien ce in th e P u blic In terest, 5, 1-31.
reports and spouse ratings on the N E O Personality Diener, E., Sc Suh, FI. (1999). National differences in
Inventory. J o u r n a l o f P erso n a lity a n d S o c ia l P sy subjective well-being. In D. Kahneman, FI. Diener,
c h o lo g y , 5 4 , 8 5 3 - 8 6 3 . 8c N. Schwarz (Eds.), W ell-b ein g : T h e fo u n d a tio n s
Cropanzano, R., 8c Wright, T. A. (1999). A 5-year o f h e d o n ic p s y c h o lo g y (pp. 4 3 4 - 4 5 0 ) . New York:
study of change in the relationship between well Russell Sage Foundation.
being and job performance. C o n su ltin g P sy ch olog y Diener, E., Sc Suh, E. M. (1997). Measuring quality
Jo u r n a l: P ra c tic e a n d R e se a rch , 51, 2 5 2 - 2 6 5 . of life: Economic, social, and subjective indicators.
Cunningham, M. R. (1988). Does happiness mean S o c ia l In d ic a to r s R e se a rch , 4 0 , 1 8 9 - 2 1 6 .
friendliness? Induced mood and heterosexual self Diener, E., 8c Suh, E. M. (Eds.). ( 2 0 00 ). C u ltu re a n d
disclosure. P erson ality a n d S o c ia l P sy ch o lo g y B u l su b je c tiv e w ell-b ein g . Cambridge, M A : M I T Press.
letin , 14, 2 8 3 - 2 9 7 . Diener, E., Suh, E. M., Lucas, R. E ., 8c Smith, H. L.
Danner, D., Snowdon D., Sc Friesen, W. (2001). Posi (1999). Subjective well-being: Three decades of
tive emotions in early life and longevity: Findings progress. P sy c h o lo g ic a l B u lletin , 1 2 5 , 2 7 6 - 3 0 2 .
158 III. E M O T IO N A L D IS P O S IT IO N S
Dunn, J. R., 8c Schweitzer, M. E. (2 005). Feeling (Eds.), T h e s c ie n c e o f su b je c tiv e w e ll-b ein g (pp. 1 7 -
and believing: T he influence of emotion on trust. 43). New York: Guilford Press.
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 8 8 , Headey, B., & Wearing, A. (1989). Personality, life
736-748. events, and subjective well-being: Toward a dynam
Easterlin, R. A. (200 6). Life cycle happiness and its ic equilibrium model. J o u r n a l o f P erso n a lity a n d
sources: Intersections of psychology, economics, S o c ia l P sy ch o lo g y , 57, 7 3 1 - 7 3 9 .
and demography. J o u r n a l o f E c o n o m ic P sy ch olog y , Helliwell, J. F. (200 7 ). Well-being and social capital:
27, 4 6 3 - 4 8 2 . Does suicide pose a puzzle? S o c ia l In d ic a to r s R e
Emmons, R. A. (1986). Personal strivings: An approach sea rch , 81, 4 5 5 - 4 9 6 .
to personality and subjective well-being. J o u r n a l o f Inglehart, R., 8c Klingemann, H.-D. ( 2 0 0 0). Genes,
P erso n a lity a n d S o c ia l P sy ch olog y , 47, 110 5-1117. culture, democracy, and happiness. In E. Diener 8c
Emmons, R. A., 8c McCullough, M. E. (2003). Coun t F^. M. Suh (Eds.), C u ltu re a n d su b je c tiv e w ell-b ein g
ing blessings versus burdens: An experimental in (pp. 1 6 5 -1 8 4 ). Cambridge, M A : M I T Press.
vestigation of gratitude and subjective well-being Isen, A. M., 8c Levin, P. F. (1972). Effect of feeling
in daily life. J o u r n a l o f P erson ality a n d S o c ia l P sy good on helping: Cookies and kindness. J o u r n a l o f
c h o lo g y , 8 4 , 3 7 7 - 3 8 9 . P erson ality a n d S o c ia l P sy ch o lo g y , 2 1, 3 8 4 - 3 8 8 .
Erasmus, D. (2003). T h e p r a is e o f folly . New Haven, James, H. S., 8c Chymis, A. (200 4). A re h a p p y p e o p le
CT: Yale University Press. e t h ic a l p e o p l e ?: E v id e n c e fr o m N o r th A m er ic a a n d
Erez, A., 8c Isen, A. M . (2002). The influence of positive E u ro jje. (Working Paper No. A EW P 2 0 0 4 - 8 ) . C o
affect on the components of expectancy motivation. lumbia: University of Missouri, Department of Ag
J o u r n a l o f A p p lie d P sy ch olog y , 87, 1 0 5 5 - 1 0 6 7 . ricultural Economics
Ferriss, A. L. (20 0 2). Religion and the quality of life. Kasser, T., 8c Kanner, A. D. (Eds.). (2 004). P sy ch olog y
J o u r n a l o f H a p p in es s S tu d ies, 3, 1 9 9 - 2 1 5 . a n d c o n s u m e r cu ltu re: T h e stru g g le f o r a g o o d life
Forgas, J. P. (1989). Mood effects on decision making in a m a ter ia lis tic w o r ld . Washington, DC: Ameri
strategies. A u stralian J o u r n a l o f P sy ch olog y , 41, can Psychological Association.
197-214. Kavanagh, D. J . (1987). Mood, persistence, and suc
Fredrickson, B. L. (1998). W hat good are positive em o cess. A u stralian J o u r n a l o f P sy ch olog y , 3 9, 3 0 7 -
tions? R e v ie w o f G e n e ra l P sy ch olog y , 2 , 3 0 0 - 3 1 9 . 318.
Fredrickson, B. L. (2001). T he role of positive em o King, L. A., Hicks, J. A., Krull, J., 8c Del Gaiso, A. K.
tions in positive psychology: The broaden-and-build (20 0 6). Positive affect and the experience of mean
theory of positive emotions. A m er ica n P s y c h o lo ing in life. J o u r n a l o f P erso n a lity a n d S o c ia l P sy
g ist, 5 6, 2 1 8 - 2 2 6 . ch o lo g y , 9 0 , 1 7 9 - 1 9 6 .
Frey, B. S., 8c Stutzer, A. (2002a). H a p p in es s a n d e c o King, L. A., 8c Napa, C. K. (1998). What makes a life
n o m ic s: H o w th e e c o n o m y a n d in stitu tion s a ffe c t good? jo u r n a l o f P erson ality a n d S o c ia l P sy c h o l
h u m a n w ell-b ein g . Princeton, NJ: Princeton Univer ogy, 75, 1 5 6 - 1 6 5 .
sity Press. Klein, S. (20 06 ). T h e s c ie n c e o f h a p p in es s: H o w o u r
Frey, B. S., 8c Stutzer, A. (2 002b). W hat can eco no b rain s m a k e us h a p p y a n d w h a t w e can d o to g et
mists learn from happiness research? J o u r n a l o f h a p p ier. New York: Marlowe.
E c o n o m ic L itera tu re , 4 0 , 4 0 2 - 4 3 5 . Kohler, H. P., Behrman, J . R., 8c Skytthe, A. (2005).
Fujita, F. (1991). An in v estig ation o f th e rela tio n b e Partner + children = happiness?: T he effect of fertil
tw een ex tr a v er sio n , n eu r o ticis m , p o s itiv e a ffe c t , ity and partnerships on subjective well-being. P o p u
a n d n eg ativ e a ffe c t . Unpublished master’s thesis, la tio n a n d D e v e lo p m e n t R e v iew , 31(3), 4 0 7 - 4 4 5 .
University of Illinois, Urbana-Champaign. Larsen, R. J 8c Ketelaar, T. (1991). Personality and
Fujita, F., 8c Diener, E. (2005) Life satisfaction set susceptibility to positive and negative emotional
point: Stability and change. J o u r n a l o f P erso n a lity states. J o u r n a l o f P erso n a lity a n d S o c ia l P sy c h o l
a n d S o c ia l P sy c h o lo g y , 88(1), 1 5 8 - 1 6 4 . og y , 61, 1 3 2 - 1 4 0 .
Fujita, P., Diener, E., 8c Sandvik, E. (1991). Gender Little, B. R. (1989). Personal projects analysis: Trivial
differences in negative affect and well-being: The pursuits, magnificent obsessions, and the search for
case for emotional intensity. J o u r n a l o f P erson ality coherence. In D. M. Buss 8c N. Cantor (Eds.), P er
a n d S o c ia l P sy ch olog y , 61, 4 2 7 - 4 3 4 . so n a lity p sy c h o lo g y : R e c en t tren d s a n d em erg in g
Furnh am, A., 8c Petrides, K. V. (2 003). Trait emotion d ire c tio n s (pp. 1 5 - 3 1 ). New York: Springer.
al intelligence and happiness. S o c ia l B e h a v io r a n d Lucas, R. E., Clark, A. E., Georgellis, Y., 8c Diener, E.
P erson ality , 3 1 , 8 1 5 - 8 2 4 . (2003). Reexamining adaptation and the set point
Gardner, J ., 8c Oswald, A. J. ( 2 00 7). Mo ney and men model of happiness: Reactions to changes in marital
tal well-being: A longitudinal study of medium status. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l
sized lottery wins. J o u r n a l o f H ea lth E c o n o m ic s , og y , 8 4 , 5 2 7 - 5 3 9 .
26, 49-60. Lucas, R. E., Clark, A. E., Georgellis, Y., 8c Diener, E.
Gove, W. R., 8c Shin, H. (1989). The psychologi (2 0 04 ). Unemployment alters the set point for life
cal well-being of divorced and widowed men and satisfaction. P sy c h o lo g ic a l S c ien ce , 15(1), 8 - 1 3 .
women. J o u r n a l o f F am ily Issu es, 10, 1 2 2 - 1 4 4 . Lucas, R. E., Diener, E., 8c Suh, E. (1996). Discrimi
Harker, L., 8c Keltner, D. (2001). Expressions of posi nant validity of well-being measures. J o u r n a l o f
tive emotion in women’s college yearbook pictures P erso n a lity a n d S o c ia l P sy ch o lo g y , 71, 6 1 6 - 6 2 8 .
and their relationship to personality and life out Lucas, R. E., 8c Fujita, F. ( 2 00 0). Factors influencing
comes across adulthood. J o u r n a l o f P erson ality a n d the relation between extraversion and pleasant af
S o c ia l P sy ch olog y , 8 0 , 1 1 2 - 1 2 4 . fect. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y ,
Haybron, D. M . ( 2 00 7). Philosophy and the science 79, 1 0 3 9 - 1 0 5 6 .
of subjective well-being. In M . Eid 8c R. J. Larsen Lykken, D., 8c Tellegen, A. (1996). Happiness is a
10. H appiness 159
stochastic phenomenon. P s y c h o lo g ic a l S cien ce, 7, (1991). Further validation of the Satisfaction with
1 8 6 -1 8 9. Life Scale: Evidence for the cross-method conver
Lyubomirsky, S., King, L., 8c Diener, E. (2005). The gence of well-being measures. J o u r n a l o f P ers o n a l
benefits of frequent positive affect: Does happi ity A ssessm en t, 57, 149-161.
ness lead to success? P sy c h o lo g ic a l B u lletin , 131, Peterson, C., 8c Seligman, M. E. P. (2 004). C h a ra c ter
803-855. stren g th s a n d v irtu es: A c la s sific a tio n a n d h a n d
Lyubomirsky, S., 8c Lepper, H. (1999). A measure of b o o k . Washington, DC: American Psychological
subjective happiness: Preliminary reliability and Association.
construct validation. S o c ia l In d ic a to r s R esea rch , Poloma, M. M ., 8c Pendleton, B. F. (1990). Religious
4 6 , 13 7-1 5 5. domains and general well-being. S o c ia l In d ic a to r s
Lyubomirsky, S., Sheldon, K. M ., 8c Schkade, D. R esea rch , 2 2 , 2 5 5 - 2 7 6 .
(20 0 5). Pursuing happiness: The architecture of Pressman, S. D., 8c Cohen, S. (2 005). Does positive af
sustainable change. R e v iew o f G e n e r a l P sy ch olog y , fect influence health? P sy c h o lo g ic a l B u lletin , 131,
9, 111-13 1. 925-971.
Lyubomirsky, S., T kach, C., 8c DiM atteo, M . R. Proffitt, D. R. (20 06 ). Distance perception. C u rren t
( 2 0 06 ). W hat are the differences between happiness D ire ctio n s in P sy c h o lo g ic a l R e se a rch , 15, 131 -1 35 .
and self-esteem? S o c ia l In d ic a to r s R e se a rc h , 78, Reis, H. T., 8c Gable, S. L. (2003). Toward a posi
363-404. tive psychology of relationships. In C. L. Keyes 8c
Magnus, K., Sc Diener, E. (1991, May). A lo n g itu d in a l J. Haidt (Eds.), F lou rish in g : T h e p o s itiv e p er so n
an aly sis o f p e r so n a lity , life ev en ts, a n d su b je c tiv e a n d th e g o o d life (pp. 1 2 9 -15 9). Washington, DC:
w ell-b ein g . Paper presented at the annual meeting American Psychological Association.
of the Midwestern Psychological Association, Chi Roberts, B. W., Caspi, A., 8c Moffitt, T. E. (2003).
cago. Work experiences and personality development in
Magnus, K., Diener, E „ Eujita, F., 8c Pavot, W. (1993). young adulthood. J o u r n a l o f P erso n a lity a n d S o c ia l
Extraversion and neuroticism as predictors of objec P sy ch olog y , 8 4, 5 8 2 - 5 9 3 .
tive life events: A longitudinal analysis. J o u r n a l o f Rusting, C. L., 8c Larsen, R. J. (1997). Extraversion,
P erson ality a n d S o c ia l P sy ch olog y , 6 5 , 1 0 4 6 - 1 0 5 3 . neuroticism, and susceptibility to positive and nega
Marks, G. N., 8c Fleming, N. (1999). Influences and tive affect: A test of two theoretical models. P erso n
consequences of well-being among Australian young a lity a n d In d iv id u a l D iffer en c es , 2 2 , 6 0 7 - 6 1 2 .
people: 1 9 8 0 - 1 9 9 5 . S o c ia l In d ic a to r s R e se a rc h , 4 6 , Ryan, R. M ., Sc Deci, E. L. ( 2 00 0). Self-determination
301-323. theory and the facilitation of intrinsic motivation,
Marsland , A. I.., Cohen, S., Rabin, B. S., 8c M an uck, social development and well-being. A m erica n P sy
S. B. (200 6). Trait positive affect and antibody re c h o lo g is t, 5 5 , 6 8 - 7 8 .
sponse to hepatitis B vaccination. B rain B eh a v io r Ryff, C. D., 8c Singer, B. (1996). Psychological well
a n d Im m u n ity , 2 0 , 2 6 1 - 2 6 9 . being: Meaning, measurement, and implications for
Mayer, J. D., Mamberg, M . H., Sc Volanth, A. J. psychotherapy research. P sy ch o th era p y a n d Psy-
(19 88). Cognitive domains of the mood system. c h o s o m a tic s , 6 5 , 1 4 - 2 3 .
jo u r n a l o f P erson ality , 5 6 , 4 5 3 - 4 8 6 . Scheier, M. F., 8c Carver, C. S. (1993). On the power
M cM a ho n, D. M. (2005). H a p p in es s: A h istory . New of positive thinking: The benefits of being optimis
York: Atlantic Monthly Press. tic. C u rren t D ire c tio n s in P sy c h o lo g ic a l S cien c e, 2 ,
Mroczek, D. K., 8c Spiro, A., III. (2005). Change in life 26-30.
satisfaction during adulthood: Findings from the Schopenhauer, A. (20 04 ). S tu d ies in p es sim is m : T h e
Veterans Affairs Normative Aging Study. J o u r n a l o f essa y s o f A rth u r S c h o p e n h a u e r (T. B. Saunders,
P erson ality a n d S o c ia l P sy ch o lo g y , 8 8 , 1 8 9 - 2 0 2 . Trans.). Whitefish, MT: Kessinger.
Myers, D. G. (1999). Close relationships and the Schutte, N. S., Malouff, J . M., Simunek, M., McKen-
quality of life. In D. Kahneman, E. Diener, 8c N. ley, J ., Sc Hollander, S. (2002). Characteristic em o
Schwartz (Fids.), W ell-b ein g : T h e fo u n d a t io n s o f h e tional intelligence and emotional well-being. C o g n i
d o n ic p s y c h o lo g y (pp. 3 7 4 - 3 8 0 ) . New York: Russell tion a n d E m o tio n , 16, 7 6 9 - 7 8 5 .
Sage Foundation. Schwarz, N., Bless, H., Wanke, M., 8c Winkielman, P.
Myers, D. G., 8c Diener, E. (1995). W ho is happy? Psy (2003). Accessibility revisited. In G. V. Bodenhaus-
c h o lo g ic a l S c ien c e , 6, 1 0 - 1 9 . en 8c A. J. Lambert (Fids.), F o u n d a tio n s o f s o c ia l
Oishi, S., Diener, E., 8c Lucas, R. E. (200 7). T he opti c o g n itio n : A fe s ts c h r ift in h o n o r o f R o b e r t S. Wyer,
mal level of well-being: Can we be too happy? P er Jr. (pp. 5 1 - 7 8 ). Mah wah, NJ: Erlbaum.
sp e c tiv e s on P sy c h o lo g ic a l S cien ce , 2 , 3 4 6 - 3 6 0 . Schwarz, N., 8c Strack, F. (1999). Reports of subjective
Okun, M . A., Stock, W. A., Haring, M. J . , 8c Witter, well-being: Judgmental processes and their method
R.A. (1984). Health and subjective well-being: A ological implications. In D. Kahneman, E. Diener,
meta-analysis. In te r n a tio n a l J o u r n a l o f A ging a n d 8c N. Schwarz (Eds.), W ell-b ein g : T h e fo u n d a tio n s
H u m an D e v e lo p m e n t, 19, 111-13 1. o f h e d o n ic p s y c h o lo g y (pp. 6 1 - 8 4 ) . New York: R us
Park, N., Peterson, C., 8c Seligman, M. E. P. (2004). sell Sage Foundation.
Strengths of character and well-being. J o u r n a l o f Seligman, M. E. P. (2002). A u th en tic h a p p in es s. New
S o c ia l a n d C lin ic a l P sy c h o lo g y , 2 3 , 6 0 3 - 6 1 9 . York: Free Press.
Pascal, B. (1995). P en sees (A. J. Krailsheimer, Trans.). Staw, B. M., 8c Barsade, S. G. (1993). Affect and man
London: Penguin Books. agerial performance: A test of the sadder-but-wiser
Pavot, W., 8c Diener, E. (1993). Review of the Satisfac vs. happier-and-smarter hypotheses. A d m in istrativ e
tion with Life Scale. P s y c h o lo g ic a l A ssessm en t, 5, S cien ce Q u arterly , 3 8 , 3 0 4 - 3 3 1 .
164-172. Suh, E., Diener, E., Oishi, S., 8c Triandis, FI. C. (1998).
Pavot, W., Diener, E., Colvin, C. R., 8c Sandvik, E. The shifting basis of life satisfaction judgments
160 III. E M O T IO N A L D IS P O S IT IO N S
Depression
Pa t r i c k H . F i n a n
H ow a rd T en n en
A l e x J. Z a u t r a
161
162 III. E M O T IO N A L D IS P O S IT IO N S
depression, most investigators use depres participants to accurately recall the details
sion severity scales to study some aspect of of an episode of depression that may have
subclinical depression, which is a “state in occurred years before the assessment. We
which depressive symptoms are present, but conclude our overview of depression mea
with too few symptoms or symptoms with surement with a brief discussion of what is
insufficient severity to warrant a diagnosis known about the recall of emotional experi
of major depressive episode” (Ingram & ences and implications for the assessment of
Siegle, 2002). Tennen, Eberhart, and Affleck depression.
(1999) reviewed a wide-ranging literature to
demonstrate that subclinical depression—■
Can People Accurately Recall Symptoms
and its assessment with depression severity
and Em otional E xp erien ces?
scales— predicts morbidity, mortality, and
health service utilization. Since Tennen and Even if depression severity scales captured
colleagues’ review, converging evidence re DSM symptoms with great precision and
veals that depressive symptoms anticipate if these scales demonstrated perfect item
the development of heart disease, and even overlap, they would still require research
minimal depressive symptoms appear to participants to recall their recent symptoms
increase mortality risk after a myocardial and emotional experiences. Tennen (2006)
infarction (see Stanton, Revenson, & Ten questioned whether people can, in fact, ac
nen, 200 7 ). The limitations of these scales curately recall their sadness, general satis
notwithstanding, there is thus good reason faction, guilt, sleep, and related symptoms
to continue to study subclinical depression over the previous week or two as required by
using depression severity scales. both self-report and interview-administered
depression measures. Extensive evidence
demonstrates that people do not actually
D epression H istory:
recall such experiences when asked to do
A n A rea o f Relative Neglect
so. Instead, they reconstruct their past ex
Depression is typically a recurrent disor perience by using implicit theories and vari
der, which means that many individuals ous cognitive heuristics (Kahneman, 1999;
with a current major depression have been Ross & Wilson, 2003). Robinson and Clore
depressed previously. Despite repeated calls (2002) demonstrated that people provide
for greater attention to the assessment of de very different answers depending on wheth
pression history, studies in the psychologi er mood is assessed in real time or retro
cal literature— especially in personality and spectively. As Tennen notes, because belief-
social psychology— have given depression based recollections of emotional experience
history scant attention. Yet many research rely on narrative coherence, recalled emo
participants with subclinical depression tional states— the currency of all struc
have had a depressive episode in the preced tured diagnostic depression interviews and
ing year (Shelbourne et al., 1994), and previ questionnaires— easily become dissociated
ously depressed individuals manifest coping from actual emotional experience (see R ob
vulnerabilities when compared with their inson & Clore, 2002). Overall, depression
never-depressed counterparts (Conner et researchers have been lulled into concluding
al., 20 0 6 ). Moreover, individuals with more that because people are able to respond to
than one previous depressive episode can their depression severity scales and inter
be distinguished from those with only one views without difficulty, their responses are
previous episode in several areas of consider accurate reflections of recent experience.
able interest to psychological investigators, The social and personality psychology lit
including greater sleep disturbance, lower erature indicates that such conclusions may
levels of perceived support, and more self be misguided.
generated stressful life events (see Zautra,
Parrish, et al., 20 0 7 ). It is therefore disquiet
ing that so few studies in the psychological Cognitive Vulnerability to Depression
literature have assessed depression history
and that fewer still have addressed the dif Perhaps the most dominant etiological con
ficulties associated with requiring research ceptualization of depression today lies in
164 III. E M O T IO N A L D IS P O S IT IO N S
leave room for speculation regarding why etiological connection of chronic stress to
some people display symptoms of depression depression, however, is unclear. Some evi
yet remain free of clinical illness, why some dence implicates chronic stressors, such as
people experience major depression yet are financial strains and loss of social support,
able to rebound and live largely depression- as mediators in the pathway from major life
free, and why others develop a depressive events to depression (Kessler, 1997). How
episode and are catapulted into a lifelong ever, inconsistencies in the operationaliza
cycle of chronic and recurrent depression. tion of chronic stress and methodological
Additionally, these theories are largely silent difficulties in the identification of the tem
on the extent to which varying levels of life poral location of chronic stress in the caus
stress affect the development of both first- al chain that produces depression have led
onset and recurrent depression. Understand major reviews to conclude that the precise
ing the various profiles of stable beliefs that role of chronic stress in depression remains
make up a cognitive vulnerability is a nec ambiguous (Hammen, 2 0 0 5 ; Kessler, 1997).
essary first step in the conceptualization of Chronic pain represents a primary target for
depression, but individual differences in the research on the relation of chronic stress to
ability to respond to stress requires a life- depression in that it provides a context in
stress perspective, as well. which stressful daily life events can exacer
bate both the perception of pain associated
with a pain condition and the affective state
Life Stress in the Developm ent reported by the individual. Furthermore,
and M aintenance o f Depression chronic pain itself can be considered a con
stant threat. Conditions such as fibromy
Making the claim that life stress influences algia (FM) and rheumatoid arthritis (RA)
the development of depression requires an are commonly associated with the onset
explanation of the meaning of the term life and maintenance of depression (Dickens,
stress. Life stress typically applies to single M cGowan, Clark-Carter, & Creed, 2002),
events and/or accumulated disruptions that as individuals with these disorders often
severely tax an individual’s coping resourc endure uncertainties in medical diagnoses,
es. The severity of a stressful event also erosion of social resources, and deterioration
contributes to variability in how individu of physical capabilities. Loss of control and
als respond to it (Brown & Harris, 1989). personal mastery are commonly reported in
A review by Mazure (1998) concluded that these patient populations as factors leading
stressful experiences classified as major and to dysfunction in affective regulation (Reich,
undesirable life events are reliably associated Johnson, Zautra, & Davis, 2006).
with onset of major depression. Events that Intensive repeated-measure designs have
involve the loss of a loved one and events examined pain, stress, and mood relations
that are perceived as uncontrollable by the in chronic pain populations without concern
individual are indeed more prevalent among for the inferential bias typically produced by
clinical populations (Mazure, 1998). Within aggregate analysis, in which extremely pain
depressed populations, individuals who have ful or extremely stressful days can skew the
experienced a high number of severe life- overall effect size. In a study of patients with
stress events exhibit more intense depres RA, Zautra and Smith (2001) found that
sive symptoms than depressed patients with elevated levels of weekly stress and weekly
relatively few severe negative life experiences negative life events, as well as greater stress
(Monroe & Hadjiyannakis, 2002). reactivity, resulted in elevated weekly de
Chronic stressors are conceptually differ pressive symptoms.
ent from major life stressors in the ways in These results tell us that chronic stressors
which they are thought to affect the indi that challenge the individual’s adaptive ca
vidual and contribute to the development of pacity to respond to physical, psychological,
depressive symptoms. Chronic stress takes and social threats on a daily basis evidence
many forms, including caring for a sick or a link to depressive symptoms. We do not
disabled loved one, having a chronic medical know, however, what levels and duration of
illness oneself, living in poverty, or under chronic stress are required to initiate a bout
going marital strife (Hammen, 2005). The of major depression. Furthermore, there is
166 III. E M O T IO N A L D IS P O S IT IO N S
ing which young women in effect create the episodes are likely to arise from previous
environments with which they will interact episodes. The numbers support this assump
for years to come. For many young women tion, with a significant proportion of people
these environments are depression-inducing, who have experienced a depressive episode
and contribute to a social context in which reporting at least one more in the future
depression and stress influence each other in (American Psychiatric Association, 2 0 0 0 ;
a dismally recurring pattern” (p. 49). Rao, Hammen, & Daley, 1999). Thus a
The parent-child relationship has been therapist’s work is rarely complete when a
consistently associated with children’s sub client is in remission from depression. The
sequent vulnerability to depression. Brown cognitive vulnerability model provides only
and his colleagues (Bifulco, Brown, & H ar a partial explanation of why people contin
ris, 1987; Brown, Harris, & Copeland, ue to experience episodes of depression: A
1977) have demonstrated the significance of high-risk cognitive style is stable over time,
parental loss in the development of depres and major adverse life events, such as loss
sion. Specifically, they have shown that (1) and trauma, are bound to happen more than
parental loss is a stronger predictor of future once in an individual’s lifetime. According to
depression when the deceased parent is the the cognitive vulnerability model, then, each
mother; (2) girls are disproportionately vul interaction of a major adverse life event with
nerable to future depression after enduring a stable latent cognitive vulnerability should
the loss of a mother; and (3) the quality of be just as likely to produce a depressive epi
parental care following the loss of one par sode as the initial depression-producing in
ent may mediate the subsequent development teraction. Yet we know that the likelihood
of depression postloss. of experiencing a new depressive episode
Goodman and Gotlib (1999) concluded increases as the number of past episodes ex
that breakdowns in communication between perienced increases.
parent and child, as well as the transmission The differential activation hypothesis
of negative cognitions through maladaptive (Teasdale, 1983) provides an alternative
behaviors in depressed mothers, both serve conceptualization of vulnerability to recur
as significant predictors of future depression rent depression. In this view, dysfunctional
among adolescents. The intimate social dy thinking akin to the cognitive distortions
namic between a depressed mother and child, described by Beck (1987) interacts with dys
in which negative cognitions shared by each phoric mood to produce depression. Individ
party can create and maintain communica uals who have already experienced depres
tion problems, further elevate a child’s risk sion are hypothesized to be at greater risk
for depression (Hammen, 2 0 0 3 ; Hammen, for new onsets of depression than those who
Burge, & Adrian, 1991). In the absence of have never had a depressive episode, because
parental support, through either the parent’s they are more prone to heightened dysfunc
death or his or her emotional unavailability, tional thinking and, consequently, dysphoric
the child’s coping resources are threatened, mood (Teasdale, 1983). Recent studies sup
just as they are among adults who lack ad port the differential activation hypothesis
equate social support (Coyne & DeLongis, with prospective designs demonstrating in
1986). The antecedents of depression are creased risk of major depression among peo
thus closely tied to the dynamics of social re ple with a history of depression identified
lationships. Indeed, a critical task for studies as high risk according to Teasdale’s criteria
of depression vulnerabilities and precipitants (Lewinsohn, Allen, Seeley, & Gotlib, 1999).
is the distinction between interpersonal and The role of stressful events in provoking
noninterpersonal forms of life stress. depression may change as a consequence of
repeated episodes. Post (1992) proposed the
kindling hypothesis, which contends that
Depression H istory: Past Events life stressors play a less important role in
as Kindling the onset of recurrent depression than they
do in the onset of first-episode depression.
If depression and a depressogenic style con As stressors become less influential in the
tribute to the generation of new stressors, onset of recurrent depression, biological fac
and if stressors contribute to the onset of tors may become more influential. Mazure
depression, it follows that new depressive (1998) found mixed support for the kindling
11. D epression 169
hypothesis and cautiously concluded that the 1985). Conceptualizing pain as both a phys
phenomenon appears to be real for people ical and psychosocial stressor, it becomes
with unipolar depression but not for those clear that chronic pain, like life stress, is not
with bipolar depression. Monroe and Hark- an epiphenomenon secondary to depression
ness (2005) provide less equivocal evidence but an interactive participant in the develop
that the association of life stress and depres ment and maintenance of depression. Rom a
sion diminishes as a function of episode re no and Turner (1985) provided a sweeping
currence. In their review of eight empirical review of the literature on the relationship of
studies testing the kindling hypothesis for chronic pain and depression and concluded
unipolar depression, they concluded that the that the available data at the time did not
presence of depression history contributes to support a single, unifying etiological path
a changed, but not necessarily a weakened, way for pain and depression. Subsequent re
relationship between life stressors and pro search has validated their interpretation by
spective onset of recurrent depression, in revealing a mutually influential relationship
dependent of any person or environmental between pain and depression (Williamson
factors that may exist. & Schulz, 1992). Not until recently, how
According to Monroe and Harkness ever, have researchers begun to question the
(2005), the interpretations made to date relation of depression history to the experi
from empirical tests of the kindling hypoth ence of chronic pain. Due to the sensitivity
esis have largely taken the view of “stress of certain pain-related syndromes, such as
autonomy,” stipulating that the effects ob rheumatoid arthritis (RA) and fibromyalgia
served occurred as a result of recurrent de syndrome (FMS), to psychosocial influences,
pression growing increasingly autonomous there is a strong clinical imperative that the
from life stressors. In the place of stress, it lingering residual mood effects, changes in
is presumed that a “quasi-independent pro social support dynamics, and adjustments
cess” (p. 427), perhaps biological and endog in coping strategies associated with prior
enous, assumes the role of depression provo depressive episodes be explored— even, per
cateur. Monroe and Harkness offer instead haps especially, among individuals who are
a “stress sensitization” explanation. In this not currently reporting depressive symp
framework, chronically depressed people, toms.
rather than becoming less reactive to life A recent series of studies have attempted to
stress, become increasingly sensitive to the evaluate how a history of depression affects
effects of stress, such that even minor stres the experience of pain, and vice versa. Fifield,
sors (not typically measured in the study of Tennen, Reisine, and M cQuillan (1998) pro
life stress) are capable of provoking a depres vided the first account of how a previous de
sive episode. For the stress-sensitized indi pression could residually affect current pain
vidual, the impact of both major and minor ratings in patients with RA under current
life events on the experience of the depres mood-priming conditions (i.e., dysphoria).
sive episode increases. However, due to the Using D SM -III-R (American Psychiatric As
increase in the likelihood of minor life events sociation, 1987) criteria, they found that pa
to precipitate depression, major life events tients who were not currently depressed but
will be associated with recurrent episodes of had experienced at least one past episode of
depression with a reduced frequency. Thus, depression and had reported elevated cur
the stress-sensitized individual will continue rent dysphoria reported more severe pain
to detect stressors at a similar rate, but with than both patients with RA who had a his
each new depression, minor stressors have tory of depression but were not experiencing
increasingly greater impact. elevated negative mood and never-depressed
individuals who were currently dysphoric.
It seems that the residual effects of previous
Im plications o f Depression H istory depressive episodes act as a latent vulner
for Chronic Pain ability (Lewinsohn, Hoberman, & Rosen
baum, 1989) that can be primed under cer
Depression history is of particular interest tain circumstances, such as dysphoric mood,
in the study of chronic pain. Depression and and exacerbate pain levels in the absence of
chronic pain have been well documented as current depression. This finding speaks to
concomitant disorders (Romano & Turner, the importance of identifying a variety of
170 III. E M O T IO N A L D IS P O S IT IO N S
physical and psychosocial stressors that can compared with individuals who have expe
be used to extend previous research on the rienced only a single episode. To that end,
multifarious manifestations of depression. Zautra, Parrish, and colleagues (2007)
Hammen and colleagues’ (1991) work on found that recurrently depressed patients
stress generation may offer clues regarding with RA reported higher levels of pain than
what in the experience of a prior depression never-depressed patients and those who had
contributes to a greater susceptibility to cur experienced only a single episode. Perhaps
rent pain. more important, increases in perceived stress
In an effort to extend the work of Fifield and pain and decreases in positive affect in
and colleagues (1998), Tennen, Affleck, and response to an experimentally induced inter
Zautra (2006) utilized an electronic diary personal stressor were greater among the re
method and daily process analysis to reveal currently depressed than among the never- or
both between- and within-person relations once-depressed patients with RA. Providing
of depression history and pain. In a sample further support for the notion that recurrent
of women with FM S, participants who had depression is implicated in the emotion-
experienced at least one previous episode regulation abilities of people with chronic
of major depression but were not currently pain, Zautra, Davis, and colleagues (2007)
depressed reported increases in emotion- found that a mindfulness-based emotion-
focused coping and decreases in perceived regulation intervention resulted in decreased
coping efficacy on high-pain days. Addition daily negative affect and increased daily pos
ally, current depressive symptoms moderated itive affect and decreased physician-assessed
the relation between daily pain and pleasant tenderness and swelling among patients with
mood for the formerly depressed, such that RA with two or more prior episodes of de
higher levels of daily depressive symptoms pression compared with those with one or
resulted in a reduction in pleasant mood no prior depressive episodes.
on high-pain days. The pleasant mood of Taken together, these studies provide sev
participants who had never been depressed eral key insights regarding depression’s in
was not affected by depressive symptoms on fluence on the experience of chronic pain.
high-pain days. Again we see evidence for a First, they show that depression history
priming effect, whereby maladaptive coping negatively affects people’s subsequent pain
strategies are stoked when depressive symp and perceived stress and does so in the ab
toms prime a latent vulnerability conferred sence of current depression diagnosis (Zau
by past depression status. In a similar daily tra, Parrish, et al., 2007). However, current
process study with patients with RA , Conner dysphoric mood may serve as a moderator in
and colleagues (2006) found that depression this relationship (Fifield et al., 1998). Second,
history was associated with greater reactiv they show that depression history may serve
ity to daily pain, as well as with the employ as a latent vulnerability to maladaptive cop
ment of more emotion-based pain coping ing strategies and poor affective regulation
strategies when problem-focused strategies when adaptive regulatory abilities are needed
might have been appropriate. Depression most: in the face of high levels of pain (Con
history, even episodes that occurred years ner et al., 2 0 0 6 ; Tennen et al., 20 0 6 ). Finally,
earlier, seems to narrow the field of coping they reveal the importance of distinguishing
resources that are crucial to successful navi between recurrent, multiple-episode depres
gation of high-pain days when they occur sion and single-episode depression when
throughout the course of chronic illness. evaluating the emotion-regulatory capacities
The studies just described operational and stress reactivity of people with a chronic
ized depression history as an all-or-nothing pain condition (Zautra, Davis, et al., 2007;
phenomenon in which group assignment did Zautra, Parrish, et al., 2007). Critics of the
not differ depending on whether a person depression history perspective might argue
had one prior depressive episode or several. that the vulnerability to pain, stress, and af
There is evidence, however, that individuals fective regulation conferred by a past history
with recurrent, multiple-episode depression of the illness could instead be explained by
exhibit unique cognitive and behavioral defi factors that actually predate the onset of the
cits (Basso & Bornstein, 1999) and increased first depressive episode. The studies just de
stress reactivity (i.e., “kindling”; Post, 1992) scribed, however, are well suited to answer
11. D epression 171
such a critique, as they avoid confounding ence of positive resources that can be used
depression history with a preexisting vulner to combat major and daily stressors. R e
ability by controlling for such “nuisance” garding trajectories, one prominent concep
variables as neuroticism. tualization of the etiological path that resil
Chronicity may reflect differences in the ience takes is that of recovery, whereby an
resources available to the person to ward off individual “bounces back” or “rebounds”
future bouts of depression. It may also reflect from a decrease in functioning following
qualitative differences in the type of depres an aversive event (Tugade, Fredrickson, &
sion and/or in the nature of genetic contribu Feldman-Barrett, 2004). Bonanno (2004)
tions to the onset of and recovery from an defines resilience as “the ability of adults
episode. Individual differences in resilience, in otherwise normal circumstances who
to which we now turn, may also contribute are exposed to an isolated and potentially
to depression recurrence vulnerability. highly disruptive event . . . to maintain rela
tively stable, healthy levels of psychological
and physical functioning . . . as well as the
Finding Resilience on the R oad capacity for generative experiences and posi
from Risk to Depression tive emotions” (pp. 2 0 -2 1 ). For Bonanno,
resilience is distinguished from a trajectory
Thus far in this chapter, we have highlighted of recovery in that resilient individuals never
social, cognitive, and psychophysiological experience clinically significant emotional
research on the vulnerability to and develop disruptions throughout the process of coping
ment of depression. In focusing on vulner with an adverse event. With these definitions
ability, we have omitted a discussion of resil in mind, we examine evidence related to the
ience. Indeed, until recently, most research trajectory of resilience following a major ad
in this area has focused on vulnerability and verse event and the personal resources that
pathology. However, the field has now begun have been linked to resilience in relation to
to pay attention to the other side of the coin, depression.
and so, too, do we in concluding this chap Longitudinal and prospective findings
ter. Now, the question is being asked, “W hat now offer evidence for the prevalence of
makes an individual resilient?” There are distinct grieving trajectories, including a re
several intriguing new findings in the litera silient trajectory, following loss. In a longi
ture on resilience to depression but at this tudinal study following bereaved spouses 18
point no clear conclusions. months after the loss, a resilient pattern of
As we have noted, the literature has iden bereavement, in which the bereaved spouse
tified the occurrence of a major adverse life experienced only minor perturbations in
event as a primary culprit in the onset of de depressive symptoms over time, was the
pression. Yet nearly every living person will predominant bereavement pattern observed
at some point encounter what, by common (Bonanno et al., 2002). This resilience to de
standards, is a major adverse event. Often, pression pattern is in stark contrast to the in
this is the death of a close loved one. Less evitability of depression postulated in stage
often, it is a traumatic event related to war, theories of loss and grief (e.g., Maciejewski,
assault, or sexual abuse. Despite this, most Zhang, Block, & Prigerson, 2007). The re
people are able to navigate their way through silient response was observed more frequent
life without experiencing the disabling ef ly than either a chronic grief response or a
fects of chronic depression (Masten, 2001). pattern of depression followed by recovery.
Resilience is often discussed in terms of Bonanno, Moskowitz, Papa, and Folkman
either the outcomes associated with it or the (2005) replicated this dominant resilient
trajectories it assumes. Regarding outcomes, pattern of responding to loss and extended
some theorists have defined resilience as the the explanatory power of a resilient trajec
absence of psychopathological symptoms tory in showing that even HIV-positive care
and the situations in which those symptoms givers who carried with them the burden
are fostered (Conrad & Hammen, 1993; of an incurable illness were highly resilient
Haeffel & Grigorenko, 20 0 7 ). An alterna in the months following the loss of a part
tive conceptualization incorporates both the ner to AIDS. Longitudinal studies such as
absence of psychopathology and the pres these provide convincing evidence for the
172 III. E M O T IO N A L D IS P O S IT IO N S
prevalence of resilient coping. In our work plications of depression to learn how people
examining resilience in everyday life, we adapt in the face of stress. Such an under
have found evidence that daily positive in taking can only serve to benefit and inform
terpersonal experiences fortify patients with the parallel exploration of dysfunction in the
chronic pain by reducing their negative af course of depression.
fective reactions to episodic pain and every
day interpersonal stress (Zautra, Johnson, &
Davis, 2005). Ong, Bergeman, Bisconti, and References
Wallace (2006) have investigated the role of
Abramson, L. Y., Alloy, L. B., Hankin, B. L., Haeffel,
resilience resources in the daily lives of older
G. J . , M a c C o o n , D. G., & Gibb, B. E. (2002). C o g
adults. These resilient responses in daily life nitive vulnerability-stress models of depression in a
and the resilient responses identified by Bo- self-regulatory and sociobiological context. In I. H.
nanno and colleagues warrant further exam Gotlib 8c C. L. Hammen (Eds.), H a n d b o o k o f d e
ination. Although Bonanno and colleagues p ressio n (pp. 2 6 8 - 2 9 4 ) . New York: Guilford Press.
Abramson, L. Y., iMetalsky, G. I., & Alloy, L. B. (1989).
were able to replicate their obtained preva
Hopelessness depression: A theory-based subtype of
lence rates for resilient bereavement, a recent depression. P s y c h o lo g ic a l R e v ie w , 9 6 , 3 5 8 - 3 7 2 .
study using a hierarchical clustering analytic American Psychiatric Association. (1987). D ia g n o stic
technique to identify bereavement patterns, a n d s ta tistic a l m a n u a l o f m e n ta l d iso rd er s (3rd ed.,
rather than a priori theoretical identification, text rev.). Washington, DC : Author.
American Psychiatric Association. (200 0). D ia g n o stic
discovered that among bereaved spouses, a a n d sta tistica l m a n u a l o f m e n ta l d is o r d e r s (4th ed.,
trajectory of depression followed by recov text rev.). Washington, DC : Author.
ery occurred with greater frequency than a Basso, M. R., 8c Bornstein, R. A. (1999). Relative
trajectory of resilience (O tt, Lueger, Kelber, memory deficits in recurrent versus first-episode
depression on a word-list learning task. N e u r o p s y
& Prigerson, 2007).
ch o lo g y , 13, 5 5 7 - 5 6 3 .
W hat promotes the resilience that pro Beck, A. T. (1967). D e p r es sio n : C a u ses a n d treatm en t.
tects some people from risk of depression? Philadelphia: University of Pennsylvania Press.
Laughter, positive emotional disclosure, Beck, A. T. (1976). C o g n itiv e th er a p y a n d the e m o
problem-focused coping, optimism, and tio n a l d iso rd er s. New York: International Universi
ties Press.
positive affect have all been linked to bet
Beck, A. T. (1987). Cognitive models of depression.
ter physical and psychological health among J o u r n a l o f C o g n itiv e P sy c h o th er a p y : A n In t e r n a
people at risk for negative health outcomes tio n a l Q u arterly , I, 5 - 3 7 .
(Tugade et al., 2004). For example, caregiv Beck, A. T., Kovacs, M ., & Weissman, A. (1975).
ers of recently deceased AIDS patients who Hopelessness and suicidal behavior: An overview.
J o u r n a l o f th e A m er ica n M e d ic a l A s so c ia tio n , 2 3 4 ,
express positive emotion more than negative 1 14 6-1 14 9 .
emotion for up to a year after the loss re Beck, A. T., Steer, R. A., 8c Brown, G. K. (1996).
port fewer depressive symptoms than care M an u al f o r th e B e c k D ep r es sio n In v en to ry : II. San
givers with a propensity for negative emo Antonio, T X : Psychological Corporation.
tional expression (Stein, Folkman, Trabasso, Bifulco, A. T., Brown, G. W., 8c Harris, T. O. (1987).
Childhood loss of parent, lack of adequate parental
& Richards, 1997). In addition to positive care and adult depression: A replication. J o u r n a l o f
emotional expression, sense making and A ffe c tiv e D iso r d er s, 1 2, 1 1 5 - 1 2 8 .
benefit finding have been identified as two Bonanno, G. (200 4). Loss, trauma, and human resil
key resources thought to confer resilience in ience. A m erica n P sy ch o lo g ist, 59, 2 0 - 2 8 .
Bonanno, G. A., Moskowitz, J . T., Papa, A., 8c Folk
response to loss or trauma. In a longitudinal
man, S. (2005). Resilience to loss in bereaved
study of bereaved spouses, making sense of a spouses, bereaved parents, and bereaved gay men.
loss conferred less distress at 1 year after the J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 8 8 ,
loss, and construing some personal benefit 827-843.
from the loss resulted in less distress, includ Bonanno, G. A., Wortman, C. B., Lehman, D. R.,
Tweed, R. G., Haring, M ., Sonnega, J ., et al. (2002).
ing fewer depressive symptoms, 18 months Resilience to loss and chronic grief: A prospective
after the loss (Davis, Nolen-FIoeksema, & study from preloss to 18 months’ postloss. J o u r n a l
Larson, 1998). o f P erson ality a n d S o c ia l P sy ch o lo g y , 8 3, 1 1 5 0 —
The field, it seems, is moving toward a 1164.
fuller understanding of depression through Brown, G. W., 8c Harris, T. O. (1989). L ife ev en ts a n d
illn ess. New York: Guilford Press.
the exploration of psychological markers of Brown, G. W., Harris, T. O., 8c Copeland, J. (1977).
resilience. In this effort, we hope to grow be Depression and loss. B ritish J o u r n a l o f P sychiatry,
yond our knowledge of the pathological im 130, 1-18.
11. D epression 173
Compton, W. M ., & Cottier, L. B. (20 04 ). T he Diag Gotlib, I. H. (1984). Depression and general psychopa
nostic Interview Schedule (DIS). In M. J . Hilsenroth thology in university students. J o u r n a l o f A b n o r m a l
&C D. I.. Segal (Eds.), T h e c o m p r e h e n s iv e h a n d b o o k P sy ch olog y , 93, 1 9 - 3 0 .
o f p s y c h o lo g ic a l a s se ss m e n t: Vol. 2 . P erso n a lity a s Haeffel, G. J ., & Grigorenko, FI. L. (20 0 7). Cognitive
sess m en t (pp. 1 5 3 -1 6 2 ). Hoboken, NJ: Wiley. vulnerability to depression: Exploring risk and re
Conner, T. S., Tennen, H., Zautra, A. J . , Affleck, G., silience. C h ild a n d A d o le s c e n t P sy ch iatric C lin ics
Armeli, S., & Fifield, J . (200 6 ). Coping with rheu o f N o r th A m er ic a , 16, 4 3 5 - 4 4 8 .
matoid arthritis pain in daily life: Within-person Hamilton, M. (1960). A rating scale for depression.
analyses reveal hidden vulnerability for the formerly J o u r n a l o f N eu r o lo g y , N eu rosu rg ery , a n d P sy ch ia
depressed. Pain, 1 2 6 , 1 9 8 - 2 0 9 . try, 2 3 , 5 6 - 6 2 .
Conrad, M ., 8c Hammen, C. (1993). Protective and Hammen, C. (1991). Generation of stress in the course
resource factors in high- and low-risk children: A of unipolar depression. J o u r n a l o f A b n o r m a l P sy
comparison of unipolar, bipolar, medically ill, and c h o lo g y , 100, 5 5 5 - 5 6 1 .
normal mothers. D ev e lo p m e n t a n d P s y c h o p a th o l Hammen, C. (2003). Interpersonal stress and depres
o g y , 5, 593. sion in women. J o u r n a l o f A ffec tiv e D iso r d er s, 74,
Coyne, J. C. (1994). Self-reported distress: Analog 49-57.
or ersatz depression? P sy c h o lo g ic a l B u lletin , 116, Hammen, C. (2005). Stress and depression. A n n u al
29-45. R e v ie w o f C lin ica l P sy ch olog y , I, 2 9 3 - 3 1 9 .
Coyne, J. C., & DeLongis, A. (1986). Going beyond Hammen, C., Burge, D., & Adrian, C. (1991). T im
social support: T he role of social relationships in ing of mother and child depression in a longitudinal
adaptation. J o u r n a l o f C o n su ltin g a n d C lin ic a l Psy study of children at risk. J o u r n a l o f C o n su ltin g a n d
c h o lo g y , 5 4 , 4 5 4 - 4 6 0 . C lin ica l P sy ch olog y , 5 9, 3 4 1 - 3 4 5 .
Coyne, J. C., &c Gotlib, I. H. (1983). The role of cogni Henkel, V., Bussfeld, P., Moller, H. J ., &c Hegerl, U.
tion in depression: A critical appraisal. P s y c h o lo g i (2002). Cognitive-behavioural theories of helpless
c a l B u lletin , 9 4 , 4 7 2 - 5 0 5 . ness/hopelessness: Valid models of depression? E u
Coyne, J. C., & Wiffen, V. E. (1995). Issues in per ro p e a n A rch iv es o f P sy ch iatry a n d C lin ica l N e u r o
sonality as diathesis for depression: The case s c ie n c e, 2 5 2 , 2 4 0 - 2 4 9 .
of sociotropy-dependency and autonomy-self- Ingram, R. FI., & Siegle, G. J. (2002). Contemporary
criticism. P sy c h o lo g ic a l B ulletin , 118, 3 5 8 - 3 7 8 . methodological issues in the study of depression:
Daley, S. E., Hammen, C., Burge, D., Davila, J., Paley, Not your father’s Oldsmobile. In I. H. Gotlib Sc
B., Lindberg, N., et al. (1997). Predictors of the gen C. L. Hammen (Eds.), H a n d b o o k o f d ep r es sio n
eration of episodic stress: A longitudinal study of (pp. 8 6 - 1 1 3 ). New York: Guilford Press.
late adolescent women. J o u r n a l o f A b n o r m a l Psy Joiner, T. E. (2002). Depression in its interpersonal
c h o lo g y , 106(2), 2 5 1 - 2 5 9 . context. In I. H. Gotlib & C. L. Hammen (Fids.),
Davis, C. G., Nolen-Hoeksema, S., & Larson, J. H a n d b o o k o f d e p r e s s io n (pp. 2 9 5 - 3 1 3 ) . New York:
(1998). Mak ing sense of loss and benefiting from Guilford Press.
the experience: Two construals of meaning. J o u r n a l Joiner, T. FI., Walker, R. L., Pettit, J. W., Perez, M.,
o f P erso n a lity a n d S o c ia l P sy ch olog y , 75, 5 6 1 - 5 7 4 . & Cukrowicz, K. C. (2 005). Evidence-based assess
Derogatis, L. R. (19 77). S C L - 9 0 -R : A d m in istra tio n , ment of depression in adults. P s y c h o lo g ic a l A s se ss
sc o rin g , p ro c ed u re s. M a n u a l I f o r th e rev ised v e r m en t, 17, 2 6 7 - 2 7 7 .
sio n . Baltimore: John Hopkins University School of Joiner, T. E., Wingate, L. R., Gencoz, T., & Gencoz,
Medicine. F. (2 005). Stress generation in depression: Three
Dickens, C., M cG o w an, L.., Clark-Carter, D., &c Creed, studies on its resilience, possible mechanism, and
F. (2002). Depression in rheumatoid arthritis: A sys symptom specificity. J o u r n a l o f S o c ia l a n d C lin ical
tematic review of the literature with meta-analysis. P sy ch olog y , 2 4 , 2 3 6 - 2 5 3 .
P sy c h o s o m a tic M ed icin e, 6 4 , 5 2 - 6 0 . Kahneman, D. (1999). Objective happiness. In D. K a h
Endicott, J., & Spitzer, R. L. (1978). A diagnostic in neman, E. Diener, Sc N. Schwarz (Fids.), W ell-b ein g :
terview: The Schedule for Affective Disorders and T h e fo u n d a tio n s o f h e d o n ic p s y c h o lo g y (pp. 8 5 -
Schizophrenia. A rch iv es o f G e n e r a l P sy ch iatry , 3 5, 105). New York: Russell Sage Foundation.
237-244. Kessler, R. C. (19 97). The effects of stressful life events
Fifield, J ., Tennen, H., Reisine, S., & McQuillan, J. on depression. A n n u al R e v ie w o f P sy ch olog y , 4 8 ,
(1998). Depression and the long-term risk of pain, 191-214.
fatigue, and disability in patients with rheumatoid Kovacs, M. (1992). C h ild r e n ’s D ep r es sio n In v en to ry
arthritis. A rth ritis a n d R h eu m a tis m , 4 1, 1851 — m a n u a l. North Tonawanda, NY: Multi-Health Sys
1857. tems.
First, M. B., Spitzer, R. L., Gibbon, M., Sc Williams, Lewinsohn, P. M ., Allen, N. B., Seeley, J . R., Sc G o t
J. B. W. (1997). S tru ctu red C lin ica l In te rv iew fo r lib, I. H. (1999). First onset versus recurrence of
D S M -IV A x is I D iso r d er s— P atien t ed itio n . New depression: Differential processes of psychosocial
York: New York State Psychiatric Institute. risk. J o u r n a l o f A b n o r m a l P sy ch olog y , 1 0 8 , 4 8 3 -
Flett, G. L., Vredenburg, K., & Krames, L. (1997). The 489.
continuity of depression in clinical and nonclinical Lewinsohn, P. M., Hoberman, H. M ., &c Rosenbaum,
samples. P sy c h o lo g ic a l B u lletin , 121, 3 9 5 - 4 1 6 . M . (1989). Probability of relapse after recovery from
Goo dm an, S., Sc Gotlib, I. (1999). Risk for psycho an episode of depression. J o u r n a l o f A b n o r m a l P sy
pathology in the children of depressed mothers: A c h o lo g y , 97, 2 5 1 - 2 6 4 .
developmental model for understanding mecha Lewinsohn, P. M ., Hops, H., Roberts, R. E., Seeley,
nisms of transmission. P sy c h o lo g ic a l R e v iew , 106, J . R., & Andrews, J. A. (1993). Adolescent psycho
458-490. pathology: 1. Prevalence and incidence of depres
174 III. E M O T IO N A L D IS P O S IT IO N S
Teasdale, J. D. (1983). Negative thinking in depression: Windle, M. (1992). A longitudinal study of stress buff
Cause, effect, or reciprocal relationship? A d v an ces ering for adolescent problems. D e v e lo p m e n ta l P sy
in B e h a v io u r R e se a rc h a n d T h era p y , 5, 3 - 2 5 . c h o lo g y , 2 8 , 5 2 2 - 5 3 0 .
Tennen, H. (20 06 ). Accuracy of recalled experience: Yesavage, J . A., Brink, T. I.., Rose, T. L., Lum, O.,
Depression measurement’s enduring illusion. M e a Huang, V., Adey, M ., et al. (1983). Development
s u r em e n t, 4, 180 -18 7 . and validation of a geriatric depression screening
Tennen, H., Affleck, G., 8c Zautra, A. J . (200 6 ). De scale: A preliminary report. J o u r n a l o f P sy ch iatric
pression history and coping with chronic pain: R e se a rc h , 17, 3 7 - 4 9 .
A daily process analysis. H ea lth P sy ch olog y , 2 5 , Zautra, A. J ., Davis, M . C., Nicassio, P., Tennen, H.,
370-379. Finan, P. H., Parrish, B. P., et al. (200 7). C o m p a riso n
Tennen, H., Eberhart, T., 8c Affleck, G. ( 1999). Depres o f c o g n itiv e -b e h a v io r a l a n d m in d fu ln ess m e d ita tio n
sion research methodologies at the social-clinical in te rv e n tio n s o n a d a p t a t io n to r h e u m a to id arth ritis
interface: Still hazy after all these years. J o u r n a l o f f o r p a tien ts w ith a n d w ith o u t h is to r y o f recu rren t
S o c ia l a n d C lin ic a l P sy ch olog y , 18, 121-159. d ep r es sio n . Manuscript submitted for publication.
Tennen, H., Hall, J. A., 8c Affleck, G. (1995). De Zautra, A. J . , Johnson, L. M., 8c Davis, M. C. (2005).
pression research methodologies in the J o u r n a l o f Positive affect as a source of resilience for women
P erso n a lity a n d S o c ia l P sy c h o lo g y : A review and in chronic pain. J o u r n a l o f C o n su ltin g a n d C lin ical
critique. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l P sy ch olog y , 7 3, 2 1 2 - 2 2 0 .
og y , 6 8 , 8 7 0 - 8 8 4 . Zautra, A. J ., Parrish, B. P., Van Puymbroeck, C. M .,
Thase, M . E. (2 007, June). D ep r es sio n ratin g sc a le s : Tennen, H., Davis, M. C., Reich, J. W., et al. (20 0 7).
H isto r y a n d cu rren t statu s. Paper presented at Depression history, stress, and pain in rheumatoid
the Consensus Conference on Advancing Signal arthritis patients. J o u r n a l o f B e h a v io r a l M ed icin e,
Strength in Proof of Concept Studies in Major De 30(3), 187-197.
pression, Bethesda, MD. Zautra, A. J., Schultz, A. S., 8c Reich, J. W. (20 0 0).
Tugade, M . M., Fredrickson, B. L., 8c Feldman- The role of everyday events in depressive symptoms
Barrett, E. (20 0 4). Psychological resilience and for older adults. In G. M. Williamson, D. R. Shaf
positive emotion granularity: Flxamining the benefit fer, 8c P. A. Parmalee (Fids.), P h y sical illn ess a n d
of positive emotions on coping and health. Jo u r n a l d e p r e s s io n in o ld e r a d u lts: A h a n d b o o k o f th eo ry ,
o f P erson ality , 7 2 , 1 1 6 1 -1 1 9 0 . resea rch , a n d p r a c tic e (pp. 6 5 - 9 1 ) . New York: Ple
Watson, D., O ’Hara, M. W., Simms, L. J., Kotov, num Press.
R., Chmielewski, M ., 8c McDade-M ontez, E. A. Zautra, A. J ., 8c Smith, B. W. (2001). Depression and
(2 00 7 ). Development and validation of the Inven reactivity to stress in older women with rheumatoid
tory of Depression and Anxiety Symptoms (IDAS). arthritis and osteoarthritis. P sy c h o s o m a tic M ed i
P s y c h o lo g ic a l A ssessm en t, 19, 2 5 3 - 2 6 8 . cin e, 6 3, 6 8 7 - 6 9 6 .
Williamson, G. M., 8c Schulz, R. (1992). Pain, activ Zimmerman, M., 8c Coryell, W. (1987). The Inventory
ity restriction, and symptoms of depression among to Diagnose Depression (IDD): A self-report scale to
community-residing elderly adults. J o u r n a l o f G e r diagnose major depressive disorder. J o u r n a l o f C o n
o n to lo g y , 47, 3 6 7 - 3 7 2 . su ltin g a n d C lin ica l P sy ch olog y , 55, 5 5 - 5 9 .
CHAPTER 12
R o w l a n d S. M i l l e r
176
12. S ocial A nxio u sn ess, Shyness, and E m b arrassab ility 177
to suppress goofy grins of chagrin that are it is obviously not gregarious and convivial,
noticeably different from smiles of genuine but whether it derives from shy trepidation,
amusement (Asendorpf, 1990). Speech er a mild manner, dullness, or unfriendly lack
rors and exaggerated movements may also of interest may be hard to judge. In any case,
occur (Edelmann & Hampson, 1979). Only because it involves anxious affect that is
a few of these various cues need to be present paired with inhibited behavior, shyness is
for a person’s embarrassment to be obvious best considered a syndrom e , rather than an
and unmistakable, so, ordinarily, observers emotion or mood per se (Leary, 1986a).
are aware of embarrassment in their midst
(Marcus & Miller, 1999). And because it is
detectable and transparent, embarrassment Social-Evaluative Siblings
has different effects on our interactions than
social anxiety and shyness do, as I discuss. Social anxiety, shyness, and embarrassment
In contrast, social anxiety is fretful dis are clearly separable, distinct states, but they
quiet that stems from the prospect of evalu share a common foundation. All of them
ations from others in the absence of any spring from a person’s attentive concern for
predicament. It occurs when we believe our interpersonal evaluation. Each emerges from
selves to be subject to real, implied, or imag situations in which one is subject to real or
ined social evaluation, and it takes the form imagined inspection by others, and each is
of nervous concern for what others may unlikely when one’s actions are genuinely
be thinking, even when nothing has gone private and will not become known to any
wrong (Leary, 2001a). Unlike embarrass one else. Consider social anxiety. A clas
ment, social anxiety often occurs over long sic formulation posits that social anxiety
periods of time, gradually waxing and wan arises when we wish to portray ourselves to
ing. It depends on contemplation of social others in a particular, desired fashion, but
settings that portrays them as daunting and we doubt that we can do so successfully
intimidating, so it is usually gradual, pro (Schlenker & Leary, 1982); the combina
longed, and mindful (rather than automatic; tion of desire and doubt is thought to trig
Miller, 2001a). Physically, social anxiety ger the aversive arousal of social anxiety.
resembles other fears; it involves activation In this model, without some motivation to
of the “fight or flight” responses of the sym construct a preferred image for others— that
pathetic nervous system—-causing higher is, if we really do not care what a particular
heart rates, faster, shallower breathing, and audience thinks of us— social anxiety will
increased blood pressure (Borkovec, Stone, not occur.1 Leary (2001b) has since revised
O ’Brien, & Kaloupek, 1974)— but it is not this formulation, suggesting that the funda
accompanied by a characteristic pattern of mental motivation at work is not merely to
coherent behavior that signals its presence succeed at impression management but, in
to other people. Outwardly, then, it may stead, to maintain our level of inclusion and
take various forms, but inwardly, it is expe acceptance by others. Self-presentational
rienced as aversive arousal that involves ten challenges fuel social anxiety, Leary assert
sion, apprehension, and unease. ed, when they engender the threatening pos
Shyness occurs when social anxiety is sibility that others will come to devalue their
paired with reticent, cautious, and guarded associations with us (because they think
social behavior (Leary, 2001a). Shy behav of us less favorably). Nevertheless, the key
ior may range from mild inhibition, involv role of social evaluation in social anxiety, it
ing bashful timidity or wary watchfulness, should be recognized, remains the same in
to stronger distancing behavior that can in the revised model: Absent the wish to be ac
clude total withdrawal from social settings. cepted by others, social anxiety is unlikely
That is a broad range, and no one pattern of to occur.
behavior reliably distinguishes shyness from Indeed, circumstances that make the
cooler, calmer states (such as those associated threat of rejection less imposing reduce the
with introversion) that lead one to be quiet social anxiety of those who encounter them.
and reserved in the absence of any anxiety Leary (1986b) provided a clever demonstra
(Henderson & Zimbardo, 2001). Shy behav tion of this when he placed young adults
ior may thus seem ambiguous to observers; with chronic social anxiety in a noisy room
178 III. E M O T IO N A L D IS P O S IT IO N S
with a stranger. The noise, which was sup Moreover, unlike other self-conscious
posed to mimic the ambience of a lively tav emotions such as shame and guilt (Tangney,
ern, was supplied by an audiotape that was Miller, Flicker, & Barlow, 1996), embarrass
always played at a constant volume— but it ment rarely occurs when we are alone (M ill
was explained in two different ways. Some er, 1992). The missteps and misbehavior that
participants were told that the noise was not give rise to embarrassment cause us chagrin
loud enough to interfere with their conversa either because others are present or because
tion, whereas others were told that the noise they are about to be. Inadvertently enter
was likely to be a hindrance. Participants in ing a restroom for the other sex can cause a
this latter condition were given an explicit flash of embarrassment even when the room
excuse for their interactions to go poorly, is unoccupied (and no one else knows of the
and it was evidently a comforting rationale: mistake) if one envisions a witness arriving.
Whereas those in the “soft” noise condition Similarly, most of us sometimes content
became anxious and behaved shyly, those in edly engage in private actions that suddenly
the “loud” environment stayed calm and did gain embarrassing potential when anyone
not appear to be shy at all. Clearly, social else shows up. Like social anxiety and shy
anxiety in these situations depended on the ness, embarrassment can occur when one is
amount of evaluative threat perceived to be entirely alone— but all three states occur in
present; it was unrelated to the objective set such circumstances only when we imagine
ting. the likely future reactions of others to our
Embarrassment is also tied to our marvel present conduct.
ous human ability to comprehend— and to Thus they are discrete states, but social
care— what others are thinking of us. For anxiety, shyness, and embarrassment are
instance, people with autism who do not all close cousins, or perhaps step-siblings,
possess a normal “theory of mind” (a typi with the same mother but different fathers
cal awareness of the likely content of oth (Miller, 2001b). Embarrassment is an unbid
ers’ thoughts) do not recognize and under den, emotional response to actual predica
stand embarrassment as readily as do people ments, whereas social anxiety and shyness
who do not have autism (Heerey, Keltner, & emerge from more deliberate contemplation
Capps, 2003). Embarrassment is difficult to of one’s situation that influences interaction
produce when one cannot take others’ points even when nothing has (yet) gone wrong.
of view and see oneself as they do. Further Each is also influenced by its own modera
more, the ability to infer others’ thoughts is tors, so that, as I discuss, shyness is more af
tied to the medial regions of the prefrontal fected by insecure doubts about one’s social
cortex; when those areas are damaged, un skills than embarrassment is (Miller, 1995).
derstanding of others’ thoughts and feelings Still, they all emerge from the same parent,
is impaired (Beer, Heerey, Keltner, Scabini, the same core ingredient: Each depends on
& Knight, 2003). Remarkably, the same re an awareness of, sensitivity to, and concern
gion becomes active when people encounter for the evaluations of us by others, and any
violations of social norms (Berthoz, Armo- dispositional characteristic that makes these
ny, Blair, & Dolan, 20 0 2 ), and children who components more potent will make these
suffer damage to the region never fully learn states more frequent and more intense.
ordinary social graces and rules of appropri
ate conduct (Anderson, Bechara, Damasio,
Tranel, & Damasio, 1999). M ost important, A B rie f H istory o f Assessment
adults who develop orbitofrontal lesions may
become incapable of experiencing embar Modern studies of social anxiety, shyness,
rassment at all (Beer et al., 2003); they can and embarrassment have become quite so
commit the most glaring improprieties with phisticated, involving cutting-edge tech
unruffled aplomb, being heedless of their nology. Neuroimaging methodologies such
social peril. The converging evidence seems as electroencephalography and functional
conclusive: Embarrassment quite clearly de magnetic resonance imaging (Schmidt &
pends on the ability to imagine others’ per Tasker, 2 0 0 0 ), various cardiovascular, neu
ceptions of us. When this ability is lacking, roendocrine, and electrodermal measures
embarrassment does not occur. (Marshall & Stevenson-Hinde, 2001), and
12. S ocial A nxio u sn ess, Shyness, and E m b arrassab ility 179
behavior genetics (Saudino, 2001) are all that describe avoidant and inhibited behav
now routinely used to assess individual dif ior that often— but not always— occurs in
ferences associated with the three states. In anxious episodes. Arguably, accurate assess
deed, studies of social anxiety, shyness, and ment of social anxiety should not be con
embarrassment have become robust, vibrant founded with measurement of shy behav
areas of investigation— but it has not always ior, a point averred by Leary (1983b), who
been so. Systematic inquiry began with the developed separate measures of interaction
creation of self-report scales that allowed anxiousness (e.g., “Parties often make me
convenient assessment of our chronic sus feel anxious and uncomfortable”) and au
ceptibilities to experience social anxiety, dience anxiousness (which pertains to ner
shyness, and embarrassment— termed social vous dread of audiences, e.g., “I usually get
anxiousness, trait shyness, and em barrass nervous when I speak in front of a group”)
ability , respectively. that exclude any references to specific social
Two measures jump-started studies of so behaviors.
cial anxiousness. In psychology, Watson and The gold standard for assessing social
Friend (1969) created the Social Avoidance anxiety that is accompanied by reticent, in
and Distress Scale, and in communication hibited behavior is the Shyness Scale created
studies, McCroskey (1970) developed a Per by Cheek and Buss in 1981 (and later ex
sonal Report of Communication Apprehen panded to a slightly longer form; see Hopko,
sion. Watson and Friend’s measure contained Stowell, Jones, Armento, & Cheek, 2005).
two subscales that addressed the tendency It contains both affective (e.g., “I feel tense
to feel anxious in social interactions (e.g., when I’m with people I don’t know well”)
“I often find social occasions upsetting”) and behavioral (e.g., “I have trouble looking
and to withdraw from them altogether (e.g., someone right in the eye”) items that appear
“I often think up excuses in order to avoid to comprise three different factors (Hopko
social engagements”). M cCroskey’s scale et al., 2005): (1) nervous distress in social
focused more broadly on anxiety associ situations, (2) wary awkwardness around
ated with public speaking in larger groups, strangers, and (3) difficulty with forthright
as well as in dyadic conversations. The two behavior. Systematic study of shyness began
measures thus had different emphases, but when Zimbardo (1977) created the Stanford
both were influential, stimulating new pro Shyness Survey, and other useful scales, such
grams of research in their respective disci as the Social Reticence Scale (Jones, Briggs,
plines. & Smith, 1986), have since been created, but
Since then, a variety of other inventories the revised Cheek and Buss Shyness Scale is
have been developed to assess social anxiety the measure most widely used in current re
in particular domains. These include mea search.
sures of physique anxiety (i.e., unease about Interest in embarrassment was rare, and
others’ evaluations of one’s body; H art, studies of embarrassability did not exist
Leary, & Rejeski, 1989), anxiety associated until Modigliani created an Embarrassabil
with public physical activity and sports par ity Scale in 1968.2 The measure asks respon
ticipation (Norton, Hope, &L Weeks, 2004), dents to imagine themselves in a variety of
and performance anxiety, or stage fright awkward situations and to rate the amount
(e.g., Osborne, Kenny, & Holsomback, of embarrassment they would feel in each
2005). Other measures have been designed case. The situations range widely, from in
specifically for children (e.g., Storch et al., nocent conspicuousness (e.g., “Suppose a
2 0 0 6 ), and still others have targeted the group of friends were singing ‘Happy Birth
stronger fears and more profound disruption day’ to you”) to inelegant interactions (e.g.,
of social life associated with social anxiety “Suppose you were talking to a stranger
disorder (e.g., Johnson, Inderbitzen-Nolan, who stuttered badly due to a speech impedi
& Anderson, 2 0 0 6 ) or “social phobia” ment”) and on to situations in which others
(Turner, Beidel, Dancu, & Stanley, 1989). are at risk and empathic embarrassment is
However, like Watson and Friend’s (1969) possible (Miller, 1987; e.g., “Suppose you
Social Avoidance and Distress Scale, many were watching an amateur show and one of
of these measures conflate items that tap the performers was trying to do a comedy
anxious phenomenology with other items act but was unable to make anyone laugh”).
180 III. E M O T IO N A L D IS P O S IT IO N S
Also included, of course, are several per shy, or embarrassable. Their higher chronic
sonal pratfalls and faux pas. Notably, these concern over social evaluation appears to re
diverse scenarios encompass all of the di sult in patterns of cognition, motivation and
verse types of situations that cause us ac emotion, and behavior that are recognizably
tual embarrassments (Miller, 1992), so the different from the thoughts, feelings, and ac
scale has substantial content validity. Kelly tions of those with weaker social-evaluative
and Jones (1997) have since developed a worries. This section of the chapter surveys
Susceptibility to Embarrassment Scale that these patterns. It begins with combined con
performs quite well, and Sabini, Siepmann, sideration of social anxiousness and shyness,
Stein, and Meyerowitz (2000) created a which share the same cognitive underpin
scale that assesses sensitivity to three differ nings.
ent classes of embarrassing situations; still,
the Embarrassability Scale remains better
Social A nxiousness and Shyness
known. Measures that assess one’s chronic
tendencies to blush in social situations have Social Cognition
also been developed by Leary and Meadows
(1991) and Bogels and Reith (1999). One of the most distinguishing characteris
Finally, studies of social anxiousness, trait tics of those who are high in social anxious
shyness, and embarrassability have been ness and trait shyness is the nature of their
aided and abetted by several other self-report thoughts in social situations (Clark & Wells,
measures that are meaningfully correlated 1995). They are prone to worry about their
with all three traits. Foremost among these interactions with others, and this nervous
are the Fear of Negative Evaluation Scale apprehension is manifested in three facets
created by Watson and Friend (1969; and of their mental lives: biased attention, inter
revised into a shorter, handier measure by pretation, and rumination.
Leary, 1983a), and the Self-Consciousness
Scale (SCS) developed by Fenigstein, Scheier, Attention. People with high social anx
and Buss (1975; and revised for use with the iousness appear to be edgily alert to stimuli
lay public by Scheier & Carver, 1985). The that signal the approach of social threats
SCS contains items that assess one’s chronic (Ledley & Heimberg, 2006). Studies using
awareness of and attention to one’s pub an emotional Stroop test— which asks par
lic image (e.g., “I’m concerned about what ticipants to name the color in which a pro
other people think of me”), and people who vocative word is printed while ignoring the
are high in such self-consciousness are more meaning of the word— routinely find that it
embarrassable and more prone to social is relatively difficult for people high in so
anxiety (Miller, 1995). Fear of negative eval cial anxiousness to divert their attention
uation refers to chronic dread of disapproval from words denoting social threats (e.g.,
from others (e.g., “I am afraid that others m ocked, rejected, disgraced) (e.g., Amir,
will not approve of me”), and it, too, is posi Freshman, &C Foa, 2002); they take longer
tively related to social anxiousness, shyness, to identify the colors of such words than of
and embarrassability (Miller, 1995). Any neutral words (e.g., house) or positive words
evaluation from others may be threatening (e.g., adm ired, accepted). In contrast, people
to some people, and new scales that will re low in social anxiousness are not differen
fine our assessment and understanding of tially attentive to semantic cues of social
the social-evaluative siblings continue to be peril (Maidenberg, Chen, Craske, Bohn, &
produced (e.g., Weeks, Heimberg, & Rode- Bystritsky, 1996).
baugh, 2008). However, much has obviously Words are relatively pallid stimuli, and
already been learned. several studies have now confronted partic
ipants with richer threat cues— angry facial
expressions— using a method that arguably
The N ature o f the Traits provides a purer measure of attention, the
dot-probe paradigm. In this computerized
To varying degrees, people who are prone to procedure, participants are asked to des
social anxiety, shyness, and embarrassment ignate as quickly as possible the position
live social lives that are different from those of a dot that flashes onscreen after the si
managed by others who are less anxious, multaneous presentation of two facial im
12. S ocial A nxiou sn ess, Shyness, and E m b arrassab ility 181
ages; shorter latencies in locating the dot When they do assay the responses they are
are thought to indicate greater attention to receiving from others, those with high social
the face that appeared in the same place as anxiousness more accurately detect disap
the dot. Investigations using this technique proving or bored reactions than accepting
have demonstrated that social anxiousness and approving ones. Two studies have placed
seems to be associated with a preconscious participants of high or low social anxious
vigilance for evaluative faces that is absent ness in a threatening setting, asking them to
in those who are less often anxious. Com give brief speeches to evaluative audiences
pared with those who are less prone to so that provided mixed reactions to the talk. In
cial anxiety, socially anxious people are Veljaca and Rapee (1998), participants were
more attentive to faces depicting positive face-to-face with the experimenter and two
(i.e., happy) or negative (i.e., anger, disgust, confederates who provided distinctly differ
fear, or sadness) emotions than to neutral ent nonverbal feedback; one of them leaned
faces when they face a daunting public forward, smiled, and nodded, while the
performance (a speech about a controver other looked sleepy, yawned, and looked at a
sial topic; M ansell, Clark, Ehlers, & Chen, watch. In Perowne and Mansell (2002), par
1999). They are also more attentive to angry ticipants gave their talks looking at a video
faces than to happy or neutral expressions, screen that showed six people who were ob
but only when the faces are presented too viously interested, bored, or neutral while
quickly— a duration of one-half second— to they ostensibly watched the speech from the
allow deliberate consideration of the imag next room. In both studies, socially anxious
es; when longer inspection is allowed— for speakers were clearly attuned to the negative
Wa seconds— the greater attention paid to reactions they encountered; compared with
angry faces disappears (Mogg, Philippot, & their less anxious counterparts, they noticed
Bradley, 2004). bored responses more often and enthusias
Socially anxious people are thus attuned tic responses less often (Veljaca & Rapee,
to emotive faces, and their attention is in 1998), and they knew who had disliked, but
stantly drawn to faces that denote hostility not who had liked, their talks (Perowne &
or rejection. Thereafter, given a chance to M ansell, 2002).
consciously consider their options, people These results are consistent with those
with high social anxiousness are relatively from visual search procedures that have
^attentive to evaluative stimuli. When they demonstrated that social anxiousness is
can take their time to react to their sur positively associated with the speed with
roundings, people high in social anxiousness which people can pick an angry face out of
avoid emotional facial expressions, paying a neutral crowd (e.g., Gilboa-Schechtman,
both smiling and angry faces less heed than Foa, & Amir, 1999). In general, then, when
they give to neutral stimuli; they also attend people with high social anxiousness scan
less to them than less anxious people do their social environments, they evidently
(Heuer, Rinck, & Becker, 20 0 7 ). This tem do so attuned to disregard. They are more
poral pattern is consistent with a vigilance- alert than others are for signs of antagonism
avoidan ce model that suggests that social or rejection, but they defensively work to
anxiousness leads people to be constantly withdraw from such signals when they are
watchful for signs of social evaluation— and, encountered. This is a pattern of attentive
preconsciously, particularly alert for anger ness that, rather than gradually promoting
and antagonism— that are then ducked and calmer, cooler reactions to social evaluation,
dodged once they are noticed (Mogg, Brad is likely instead to perpetuate groundless,
ley, de Bono, & Painter, 1997). In ordinary misplaced trepidation in social settings (Bo-
social interactions, both reactions may be gels & M ansell, 2004).
counterproductive. High levels of relentless, Compared with those who are less anx
automatic vigilance are probably needless ious, those with high social anxiousness are
and wearing, leaving socially anxious peo also more attentive to internal physiological
ple jittery and depleted. Then, their wish to cues regarding their arousal in public places.
evade the feedback they do encounter may Indeed, they are even more attentive to in
rob them of the opportunity to learn over formation about their internal states (such
time that rejection from others is actually as heart rate feedback) than to threatening
quite rare. external cues (such as angry faces; Pineles
182 III. E M O T IO N A L D IS P O S IT IO N S
& M ineka, 20 0 5 ), especially when they are Interpretation. Higher levels of social
placed in evaluative social situations (such anxiousness also lead people to imagine the
as giving a speech on a controversial topic; worst in innocuous stimuli that do not seem
Mansell, Clark, & Ehlers, 2003). Thus peo worrisome to those who are less routinely
ple with high social anxiousness are par anxious (Hirsch & Clark, 2004). They at
ticularly likely to be distracted by, and then tach pejorative interpretations to ambiguous
preoccupied with, the sensations of their actions from others, and they perceive disap
physical activation in social situations. The proval where it does not objectively exist—
presence of others is stimulating, and some and the higher their levels of social anxious
arousal is normal; people with high social ness, the more pronounced these tendencies
anxiousness do not necessarily become become (Huppert, Foa, Furr, Filip, & M at
more aroused than anyone else in social sit thews, 2003).
uations— but they think they do (Edelmann In one study, for instance, people with
& Baker, 2002). Then the self-focused at high or low social anxiousness read a sce
tention that accompanies social anxiousness nario describing a blind date (Constans,
makes one’s unease seem even more intense, Penn, Ihen, & Hope, 1999). At several junc
making anxiety even worse (Zou, Hudson, tures in the story, ambiguous statements
& Rapee, 2007). were made (e.g., “When meeting her date,
Moreover, people with high social anxious Lisa said ‘You’re certainly not what I ex
ness think differently about upcoming social pected’ ”), and the participants were asked
interactions before they even begin. Their to consider the plausibility of various inter
anticipatory processing is characterized by pretations of each ambiguity (e.g., “When
recurrent and intrusive thoughts regarding Lisa said to Steve, ‘You’re certainly not what
their physical nervousness, past failures, and I expected,’ she was impressed”). When the
fantasies of escape from the current threat ambiguity involved an interpersonal evalu
that are more frequent and compelling than ation, people with high social anxiousness
those experienced by those with less anx judged positive possibilities to be less likely
iousness (Vassilopoulos, 2004). This train of than less anxious people did; they were not
thought is disadvantageous, both interfering more pessimistic, however, about imperson
with their preparation for the event and inten al judgments (such as a first impression of a
sifying their anxiety (Vassilopoulos, 2005). restaurant).
In particular, when people who are not so In a subsequent investigation (Voncken,
cially anxious are instructed to adopt this Bogels, & de Vries, 2003), the participants
outlook before an upcoming speech— “think were themselves the targets of brief scenari
of a particular social situation that you felt os that involved a variety of outcomes rang
did not go well; try to anticipate the worst ing from favorable (e.g., “someone makes a
thing that could happen while you are giving compliment about your looks”) through am
the speech”— they get more nervous, both be biguous (e.g., “somebody you know looks
fore and during the performance (Hinrichsen in your direction”) to profoundly negative
& Clark, 2003). Furthermore, when people (e.g., “a friend tells you that a colleague dis
with high social anxiousness are distracted likes you”). People with high levels of social
by another task and their usual fearful an anxiousness made more negative interpre
ticipation is interrupted, an imminent threat tations of all of these events than did those
causes them less concern (Hinrichsen & with lower anxiousness, but they did not dif
Clark, 2 0 0 3 ; Vassilopoulos, 2005). fer from their less anxious counterparts in
Altogether, then, the greater attention judgments of nonsocial events.
chronically paid by those who are socially Evidently, social anxiousness does not lead
anxious to cues of rejection, internal arous people to be generally gloomy, but it does
al, and worst-possible outcomes appears to lead them to have pessimistic perceptions
leave them continually ill at ease in social of social situations that are most apparent
settings. Indeed, through their eyes, the so in uncertain interactions in which others’
cial environments they encounter probably evaluations are indistinct (Amir, Beard, &
seem more routinely menacing than those Bower, 2005). When they complete am
same situations seem to others with lower biguous sentences involving social evalua
social anxiousness. tions (e.g., “As you give a speech, you see a
12. S ocial A n xio u sn ess, Shyness, and E m b arrassab ility 183
person in the crowd smiling, which means anxiousness approach social life more cau
that your speech i s ________ ”), people with tiously than do those who are less anxious.
high social anxiousness provide more fret Arguably, they do not seek approval from
ful and derogatory responses than less anx others so much as they defensively strive
ious people do (Huppert, Pasupuleti, Foa, to avoid disapproval (Shepperd & Arkin,
& Matthews, 20 0 7 ). Social anxiousness 1990). This leads them to interact with oth
makes social interactions seem more costly ers in a manner that is self-protective rather
and dangerous than they would appear to than acquisitive, and their interpersonal be
be were one less anxious (Schofield, Coles, havior is intentionally innocuous: They sit
& Gibb, 2007). on the sides or in the back of classrooms,
People with high social anxiousness also they express neutral opinions, and they con
judge themselves more harshly. They un form readily (see Shepperd & Arkin, 1990).
derestimate their physical attractiveness, There is only a modest negative correlation
tend to blame themselves for disappointing (r = -.3 0 ) between shyness and sociability,
outcomes, and tend to doubt the accuracy so people with high social anxiousness do
of the praise they receive (Cheek & Briggs, not necessarily wish to be left alone; in
1990). Because they are usually nervous and deed, some shy people are quite sociable
uneasy, socially anxious people do make rel (Cheek & Buss, 1981). They generally do
atively poor impressions on their audiences not much like to interact with anyone they
when they give public presentations; never do not already know, however, because
theless, their self-evaluations of their poise they are averse to social risk (Brown, Silvia,
and preparation are more damning than the Myin-Germeys, & Kwapil, 20 0 7 ). Their
appraisals they actually receive from others. motivation to minimize the chances of dis
Social anxiousness is unrelated to judgments approval from others seems to leave them
of others’ performances, but it does seem wary, watchful, and relatively unwilling to
to make people unduly hard on themselves take advantage of social opportunities that
(Ashbaugh, Antony, M cCabe, Schmidt, & are attractive and pleasurable to those with
Swinson, 2005). lower anxiousness.
Indeed, people with high social anxious
Rumination. Finally, socially anxious ness are less curious about new ideas and
people ponder their problems more persis experiences than less anxious people are
tently than others do. In particular, they (Kashdan, 2007). Curiosity is a “pleasant,
brood after a public presentation, replaying appetitive state,” and “when people feel curi
the event in their minds and fussing over ous, they thrive on novel and challenging in
their (perceived) imperfections (Edwards, teractions with the world, with exploratory
Rapee, & Franklin, 2003). They may also behavior inevitably leading to an expansion
be beset by the fear that their nervous disar of knowledge, skills, and resources” (Kash
ray caused discomfort to others (Rector, Ko- dan, 2007, p. 350). Being relatively lacking
covski, & Ryder, 2006). This pejorative style in curiosity, socially anxious people would
of postevent processing is even more pro rather stay close to home.
nounced after one-on-one interactions than However, this preference may be costly.
after solo performances, and among people When respondents describe their feelings
with high social anxiousness, it is specific several times a day in experience-sampling
to social-evaluative situations as opposed studies, social anxiousness is associated with
to impersonal threats (Fehm, Schneider, & higher negative affect (Brown et al., 2007),
Hoyer, 2007). Thus the peril posed by social less positive affect, and the enjoyment of
settings does not end when an interaction is fewer positive events (Kashdan & Steger,
done; self-criticism and self-censure perse 2 0 0 6) day by day. As we have seen, bouts of
vere, often for days thereafter (Dannahy & social anxiety cause autonomic arousal, ner
Stopa, 2007). vousness, and disquiet; these unhappy sen
sations are disagreeable enough, but chronic
susceptibility to such states is also linked to
Motivation and Emotion
fewer positive emotions and more unpleas
Burdened with excessive self-focus and a ant feelings in one’s daily routine. Social
pessimistic outlook, people with high social anxiousness is no fun.
184 III. E M O T IO N A L D IS P O S IT IO N S
and they think that they have less dexterity and this may cumulatively be quite disad
in social situations than others do (Miller, vantageous. Down the road, shy men marry
1995). The behavioral inhibition that de 3 years later, on average, than men who are
fines shyness is tied both to nervous dread not shy, and they take longer to establish
of what others are thinking and to doubts their careers (Caspi, Bern, & Elder, 1988).
about one’s social competence that magnify Shy people of both sexes also suffer more
one’s unease. health problems, such as sleep disturbances
and nausea (Langston & Cantor, 1989); the
Interactive and R elational Effects. None greater stress they encounter in social situa
of this escapes the notice of those with whom tions may take a toll over time.
shy people interact. The reticence of shy peo This all sounds rather grim, but shyness
ple can make them seem detached and un is not all bad. The intimate partners of shy
friendly, and they make poorer impressions people describe them as modest, sensitive,
on conversation partners than do those who and tactful (Gough & Thorne, 1986), so
are less shy (Heerey & Kring, 2 0 0 7 ). This shy diffidence may play well in close quar
is regrettably ironic: Concern about others’ ters once a partner’s love is won. Moreover,
judgments leads shy people to behave in a it is entirely sensible to be cautious and
timid, cautious, and clumsy manner that reserved when one is in unfamiliar terri
engenders the disregard they hoped to avoid tory, governed by unknown norms. On the
(see Curtis & Miller, 1986); this confirms whole, however, the patterns of cognition,
their fears, conceivably leading to stronger motivation and emotion, and behavior that
shyness and further withdrawal. emerge from social anxiousness and shyness
These unhappy outcomes may add up operate as anticipatory preparation for so
over time. In school, the hesitance and dif cial rejection that rarely occurs— unless the
fidence of shy adolescents may be mistaken pessimistic outlook and inhibited behavior
for a lack of intelligence (Evans, 2001) and of anxious and shy people make their fears
lead to victimization by their peers (Blote & come true. In this sense, then, social anx
Westenberg, 200 7 ). Compared with those iousness and shyness are maladaptive. They
who are less shy, teenagers with higher are costly, and their benefits are few. Are the
levels of trait shyness have fewer friends, social-evaluative siblings undesirable, then?
with whom they share weaker emotional Not entirely.
ties and who provide them less support, so
their friendships are less satisfying and of
Em barrassability
lower quality (Rubin, Wojslawowicz, Rose-
Krasnor, Booth-LaForce, & Burgess, 2006). With their common roots in social-evaluative
They develop attachment styles that tend concern, social anxiousness, shyness, and
to be characterized by anxiety over aban embarrassability have much in common.
donment, and they are relatively unlikely to As you might expect, socially anxious and/
enjoy secure attachments (Darcy, Davila, & or shy people respond to social predica
Beck, 2005). ments with more intense embarrassment
When they go to college, shy people make than do those who are less anxious or shy
friends more slowly than others do, and they (Hofmann, Moscovitch, & Kim, 2006).
are less likely to start a new romance (Asen- They are clearly more embarrassable (M ill
dorpf, 20 0 0 ). They have fewer sex partners er, 1995). However, the traits are far from
(Leary & Dobbins, 1983), and the sex they being synonymous; the correlation between
have is of poorer quality (Bradshaw, 2006). embarrassability and social anxiousness is
Shyness is probably associated with these 0 .4 8 — sizable, but not enormous-—and shy
outcomes not only because it makes one’s ness is even more distinct (being correlated
interactions less rewarding but also because with embarrassability, r = .37; Miller, 1995).
it reduces one’s opportunities to make new Stemming from alert sensitivity to impro
friends and to find new loves (Leary & priety that is coupled with a dread of disap
Buckley, 2 0 0 0 ). Being less socially dexter proval, embarrassability leads to pessimistic
ous than others, shy people initiate fewer vigilance, much like the other traits. Never
conversations, interact with fewer people, theless, within its normal range, it has fewer
and go on fewer dates (Asendorpf, 20 0 0 ), adverse effects on daily life.
186 III. E M O T IO N A L D IS P O S IT IO N S
to the circumstances (Miller, 2 0 0 7 ). In one Too often, however, people are too con
study bearing on perils of this sort, a young cerned about what others are thinking of them
woman requesting help from a college class (Leary, 2004), and this is frequently the case
dropped a stack of forms and responded with with social anxiousness, shyness, and embar
evident chagrin in one of two ways (Levin & rassability. Social awareness is undoubtedly
Arluke, 1982): she either yelped, “Oh, my almost always valuable, but the fretful con
God, I can’t continue,” and ran from the cerns of social anxiety are only rarely useful.
room, or she stayed put and seemed obvi Beyond modest levels, social anxiousness
ously embarrassed as she made her request. has people edgily fearing threats that never
She received more assistance from her peers arrive. Or worse, as in the case of shyness,
when her embarrassment was apparent but it leads people to behave in ineffective ways
not excessive. Indeed, embarrassment that is that produce the outcomes they fear.
calibrated to its circumstances is usually met Within their normal ranges of operation,
with sympathy and support by the others embarrassability seems the most desirable of
present; oddly extreme reactions, however, these three traits, because a person complete
are not (Miller, 1996, 2007). ly without it would be impaired, being impla
The most important influence of embar cably unruffled in outrageous situations and
rassability on behavior, however, may be the seeming to observers to lack a conscience.
manner in which it may lead people to avoid In contrast, were the fretful anticipation of
situations that promise to be embarrassing, unlikely rejection that characterizes social
even if such situations will be good for them anxiousness and the usually misplaced ti
(Miller, 20 0 7 ). In particular, people too midity and inhibition of shyness entirely
often put off or avoid medical treatment for absent from our lives, most of us would be
such conditions as sexually transmitted in better off. Still, the social-evaluative siblings
fections (Hook & Sharma, 20 0 5 ), colorec exist for good reason; there are occasions
tal cancer (Hou, 2005), and urinary incon that are socially risky, and the caution that
tinence (Horrocks, Somerset, Stoddart, & the siblings promote is sometimes beneficial.
Peters, 2004) because of misplaced concern We might not wish to be without some mea
about others’ opinions of them. Despite hav sure of social anxiousness, shyness, and em
ing good intentions, they may also fail to barrassability, even if they could be avoided.
buy and use condoms (Moore, Dahl, Gorn, They are inevitable influences, then, on our
& Weinberg, 20 0 6 ). In these cases, social transactions with others. The trick lies in
sensitivity that at more appropriate levels keeping them in check so that they operate
would promote decorous, genteel behavior
within beneficial limits.
is actually disadvantageous.
N o te s
Conclusion
1. T h e accu racy of this poin t was driven ho me to me
If they are so often distressing and detrimen when, as a favor to a friend, I gave a brief ta lk on
tal, why are social anxiousness, shyness, and “ W h a t Professors D o ” to an audience o f 3 - and
embarrassability so prevalent? Most theo 4-ye a r-o ld s at a day care facility. T h e ta lk did
rists agree that ours is a social species with not go well. I was met with the m ost profound
uninterest I have ever en countered , and my audi
evolved mechanisms that help us maintain
en ce’s boredom was palpable. T h i s would have
congenial relations with our fellows (Miller, been distressing had my friend or any other adult
2004). The social-evaluative siblings pre been present, but I was alone with the p re scho ol
sumably arose because they were beneficial ers, all of them strangers. 1 was relieved to find
in some way— and, in moderation, keeping that 1 seemed to be imm une to their indifference.
track of others’ opinions of us is undoubt It was enough that my friend was grateful for
edly of service to our survival. In particular, my help; 1 would never see these kids again, and
it simply didn’t m atter w hether they liked me or
mechanisms that alert us when rejection is
not. Being accepted by them was of no im p o r
imminent provide invaluable feedback that ta nce to me. Given t h at, I could be frustrated by
interrupts undesirable behavior and pro my failure to en tertain them , but I was incapable
motes remedial action (Leary & Baumeister, o f being a n x io u s a bo ut the pro sp ect o f their re
2 0 0 0 ). je ctio n. It was rather a freeing exp erience.
188 III. E M O T IO N A L D IS P O S IT IO N S
2. The Em barrassability Scale was originally writ questionnaires to assess social fears: T he Dutch So
ten using only masculine pronouns. A version cial Phobia and Anxiety Inventory and the Blushing,
appropriate for use for both sexes, adapted by Trembling and Sweating Questionnaire. J o u r n a l o f
me with M odigliani’s permission, appears in P sy c h o p a th o lo g y a n d B e h a v io r a l A ssessm en t, 21,
Miller (1996). 51-66.
Borkovec, T. D., Stone, N., O ’Brien, G., 8c Kaloupek,
D. (1974). Identification and measurement of anxi
ety in an analogue social situation. J o u r n a l o f C o n
References su ltin g a n d C lin ica l P sy ch olog y , 4 4 , 1 4 3 - 1 5 3 .
Bradshaw, S. D. ( 2 00 6). Shyness and difficult rela
Alden, L. E., 8c Taylor, C. T. (20 0 4). Interpersonal tionships: Formation is just the beginning. In D. C.
processes in social phobia. C lin ica l P sy ch o lo g y R e Kirkpatrick, S. Duck, 8c M. K. Foley (Eds.), R e la t
v iew , 2 4 , 8 5 7 - 8 8 2 . ing d iffic u lty : T h e p r o c e s s e s o f co n stru ctin g a n d
Amir, N., Beard, C., 8c Bower, E. (2 005). Interpreta m a in ta in in g d iffic u lt in te ra ctio n (pp. 1 5 - 4 1 ) . Mah-
tion bias and social anxiety. C o g n itiv e T h era p y a n d wah, NJ: Erlbaum.
R esea rc h , 2 9 , 4 3 3 - 4 4 3 . Brown, L. H., Silvia, P. J ., Myin-Germeys, I., 8c Kwa-
Amir, N., Freshman, M ., 8c Foa, E. (200 2). Enhanced pil, T. R. ( 2 00 7). When the need to belong goes
Stroop interference for threat in social phobia. J o u r wrong: The expression of social anhedonia and so
n a l o f A n x iety D iso r d er s, 16, 1-9. cial anxiety in daily life. P sy c h o lo g ic a l S cien ce, 18,
Anderson, S. W., Bechara, A., Damasio, H., Tranel, 778-782.
D., 8c Damasio, A. R. (1999). Impairment of so Caspi, A., Bern, D. J., 8c Elder, D. J. (1988). Conti
cial and moral behavior related to early damage in nuities and consequences of interactional styles
human prefrontal cortex. N a tu re N eu r o sc ie n c e, 2, across the life course. J o u r n a l o f P erson ality , 57,
1 0 3 2 -1 03 7. 375-406.
Apsler, R. (1975). Effects of embarrassment on behav Cheek, J. M ., 8c Briggs, S. R. (1990). Shyness as a per
ior toward others. J o u r n a l o f P erson ality a n d S o cia l sonality trait. In W. R. Crozier (Ed.), S hyn ess a n d
P sy c h o lo g y , 3 2 , 1 4 5 - 1 5 3 . em b a r r a s s m e n t: P ersp ectiv es fr o m s o c ia l p s y c h o l
Asendorpf, J. (1990). T he expression of shyness and o g y (pp. 3 1 5 - 3 3 7 ) . Cambridge, UK: Cambridge
embarrassment. In W. R. Crozier (Ed.), S hyn ess University Press.
a n d em b a r r a s s m e n t: P ersp ectiv es fr o m s o c ia l p sy Cheek, J . M., 8c Buss, A. H. (1981). Shyness and so
c h o lo g y (pp. 87 -1 1 8 ). Cambridge, UK: Cambridge ciability. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o l
University Press. og y , 4 1 , 3 3 0 - 3 3 9 .
Asendorpf, J . B. ( 2 0 00 ). Shyness and adaptation to the Chen, V., 8c Drummond, P. (20 08 ). Fear of negative
social world of university. In W. R. Crozier (Ed.), evaluation augments negative affect and somatic
S h y n ess: D ev e lo p m e n t, c o n s o lid a tio n a n d ch an g e symptoms in social-evaluative situations. C o g n itio n
(pp. 1 0 3 - 1 2 0 ) . New York: Routledge. a n d E m o tio n , 2 2 , 2 1 - 4 3 .
Ashbaugh, A. R., Antony, M. M ., M cC abe, R. E., Clark, D. M ., 8c Wells, A. (1995). The cognitive model
Schmidt, L. A., 8c Swinson, R. P. (2 005). Self- of social phobia. In R. G. Heimberg, M. R. Lie-
evaluative biases in social anxiety. C o g n itiv e T h e r a bowitz, D. A. Hope, 8c F. R. Schneier (Eds.), S o
p y a n d R e se a rc h , 2 9 , 3 8 7 - 3 9 8 . c ia l p h o b ia : D ia g n o sis, a s se ss m e n t, a n d tre a tm e n t
Baker, S. R., 8c Edelmann, R. J . (2002). Is social pho (pp. 6 9 - 9 3 ) . New York: Guilford Press.
bia related to lack of social skills?: Duration of Constans, J. I., Penn, D. L., Ihen, G. H., & Hope, D.
skill-related behaviours and ratings of behavioural A. (1999). Interpretive biases for ambiguous stimuli
adequacy. B ritish J o u r n a l o f C lin ica l P sy ch olog y , in social anxiety. B eh a v io u r R e se a rc h a n d T h era p y ,
4 1, 2 4 3 - 2 5 7 . 37, 6 4 3 - 6 5 1 .
Beer, J. S., Heerey, E. A., Keltner, D., Scabini, D., 8c Curtis, R. C., 8c Miller, L. (1986). Believing another
Knight, R. T. (2003). The regulatory function of likes or dislikes you: Behaviors making the beliefs
self-conscious emotion: Insights from patients with come true. J o u r n a l o f P erso n a lity a n d S o c ia l P sy
orbitofrontal damage. J o u r n a l o f P ers o n a lity a n d c h o lo g y , 51, 2 8 4 - 2 9 0 .
S o c ia l P sy ch olog y , 8 5, 5 9 4 - 6 0 4 . Dannahy, L., 8c Stopa, L. (20 07 ). Post-event process
Beidel, D. C., 8c Turner, S. M . (200 7). Shy ch ild ren , ing in social anxiety. B e h a v io u r R e se a rc h a n d T h e r
p h o b ic a d u lts: N a tu re a n d tre a tm e n t o f s o c ia l a n x i ap y , 4 5 , 1 2 0 7 - 1 2 1 9 .
ety d is o r d e r (2nd ed.). Washington, DC: American Darcy, K., Davila, J ., 8c Beck, J . G. (2 005). Is social
Psychological Association. anxiety associated with both interpersonal avoid
Berthoz, S., Armony, J . L., Blair, R. J. R., 8c Dolan, R. ance and interpersonal dependence? C o g n itiv e
J. (2002). An f M R I study of intentional and unin T h era p y a n d R e se a rch , 2 9, 1 7 1 -1 86 .
tentional (embarrassing) violations of social norms. Davila, J ., 8c Beck, J . G. (2 002). Is social anxiety as
B rain , 1 2 5 , 1 6 9 6 - 1 7 0 8 . sociated with impairment in close relationships?: A
Blote, A., 8c Westenberg, P. M . (200 7 ). Socially anxious preliminary investigation. B e h a v io r T h era p y , 33 ,
adolescents’ perceptions of treatment by classmates. 427-446.
B eh a v io u r R e se a rch a n d T h era p y , 4 5 , 1 8 9 - 1 9 8 . DePaulo, B. M ., Epstein, J. A., 8c LeMay, C. S. (1990).
Bogels, S. M ., 8c Mansell, W. (200 4). Attention pro Responses of the socially anxious to the prospect
cesses in the maintenance and treatment of social of interpersonal evaluation. J o u r n a l o f P erson ality ,
phobia: Hypervigilance, avoidance and self-focused 58, 6 2 3 -6 4 0 .
attention. C lin ic a l P sy ch o lo g y R ev iew , 2 4 , 8 2 7 - Drummond, P. D., Camacho, L., Formentin, N., Hef-
856. fernan, T. D., Williams, F., 8c Zekas, T. E. (2003).
Bogels, S. M ., 8c Reith, W. (1999). Validity of two The impact of verbal feedback about blushing on
12. S ocial A n xio u sn ess, Shyness, and E m b arrassab ility 189
social discomfort and facial blood flow during em a n d in te rv e n tio n s relatin g to th e s e l f a n d sh y n ess
barrassing tasks. B e h a v io u r R esea rch a n d T h era p y , (pp. 4 3 1 - 4 4 7 ) . Chichester, UK: Wiley.
41, 4 1 3 - 4 2 5 . Heuer, K., Rinck, M ., 8c Becker, E. S. (20 07 ). Avoid
Edelmann, R. J. (2001). Blushing. In W. R. Crozier ance of emotional facial expressions in social an x i
8c L. E. Alden (Eds.), In te r n a tio n a l h a n d b o o k o f ety: T he Approach-Avoidance Task. B e h a v io u r R e
s o c ia l a n x iety : C o n c e p ts , resea rch a n d in te rv e n sea rch a n d T h era p y , 4 5 , 2 9 9 0 - 3 0 0 1 .
tion s relatin g to th e s e l f a n d sh y n ess (pp. 3 0 1 - 3 2 3 ) . Hinrichsen, H., 8c Clark, D. M. (2003). Anticipato
Chichester, UK: Wiley. ry processing in social anxiety: Two pilot studies.
Edelmann, R. J ., 8c Baker, S. R. (2 002). Self-reported J o u r n a l o f B e h a v io r T h era p y a n d E x p e r im e n ta l
and actual physiological responses in social phobia. P sychiatry, 3 4 , 2 0 5 - 2 1 8 .
B ritish J o u r n a l o f C lin ic a l P sy ch olog y , 4 1, 1-14. Hirsch, C. R., 8c Clark, I). M. (2004). Information-
Edelmann, R. J., 8c Hampson, R. J. (1979). Changes in processing bias in social phobia. C lin ica l P sy ch ol
non-verbal behaviour during embarrassment. B rit og y R ev iew , 2 4 , 7 9 9 - 8 2 5 .
ish J o u r n a l o f S o c ia l a n d C lin ica l P sy ch olog y , 18, Hofmann, S. G., Moscovitch, D. A., 8c Kim, H. (2006).
385-390. Autonomic correlates of social anxiety and em bar
Edwards, S. L., Rapee, R. jVI., & Franklin, J . (2003). rassment in shy and non-shy individuals. In te r n a
Postevent rumination and recall bias for a social tio n a l J o u r n a l o f P sy c h o p h y sio lo g y , 61, 1 3 4 - 1 4 2 .
performance event in high and low socially anxious Hopko, D. R., Stowell, J., Jones, W. H., Armento,
individuals. C o g n itiv e T h era p y a n d R e se a rch , 27, M . E. A., 8c Cheek, J. M. (2005). Psychometric
603-617. properties of the Revised Cheek and Buss Shy
Evans, M. A. (2001). Shyness in the classroom and ness Scale. J o u r n a l o f P erson ality A ssessm en t, 84,
home. In W. R. Crozier 8c L. E. Alden (Eds.), In 185-192.
t e r n a tio n a l h a n d b o o k o f s o c ia l a n x iety : C o n c e p ts , Ho ok, E. W., Ill, 8c Sharma, A. K. (2005). Public toler
re se a r c h a n d in te rv e n tio n s relatin g to th e s e l f a n d ance, private pain: Stigma and sexually transmitted
sh y n ess (pp. 159-18 3 ). Chichester, UK: Wiley. infections in the American Deep South. C u ltu re,
Fehm, L., Schneider, G., 8c Hoyer, J. (20 07 ). Is post H ea lth , a n d S ex u ality , 7, 4 3 - 5 7 .
event processing specific for social anxiety? Jo u r n a l Horrocks, S., Somerset, M ., Stoddart, H ., 8c Peters, T.
o f B e h a v io r T h era p y a n d E x p e r im e n ta l P sychiatry, J. (200 4). W hat prevents older people from seeking
38, 11-22. treatment for urinary incontinence?: A qualitative
Fenigstein, A., Scheier, M . F., 8c Buss, A. FI. (1975). exploration of barriers to the use of community con
Public and private self-consciousness: Assessment tinence services. F am ily P ractice, 2 1, 6 8 9 - 6 9 6 .
and theory. J o u r n a l o f C o n su ltin g a n d C lin ica l P sy Hou, S. (2005). Factors associated with intentions for
c h o lo g y , 4 3 , 5 2 2 - 5 2 7 . colorectal cancer screenings in a Chinese sample.
Furmark, T. (2 002). Social phobia: Overview of co m P sy ch o lo g ica l R e p o rts , 9 6 , 1 5 9 - 1 6 2 .
munity surveys. A cta P sy ch iatrica S ca n d in a v ica , Huppert, J . D., Foa, FI. B., Furr, J. M ., Filip, J. C.,
105, 8 4 - 9 3 . 8c Matthews, A. (2003). Interpretation bias in so
Garcia, S., Stinson, L., Ickes, W., Bissonnette, V., 8c cial anxiety: A dimensional perspective. C o g n itiv e
Briggs, S. R. (1991). Shyness and physical attrac T h era p y a n d R esea rch , 27, 5 6 9 - 5 7 7 .
tiveness in mixed-sex dyads .J o u r n a l o f P erson ality Huppert, J. D., Pasupuleti, R. V., Foa, FI. B., 8c M at
a n d S o c ia l P sy ch olog y , 61, 3 5 - 4 9 . thews, A. (20 07 ). Interpretation biases in social
Gerlach, A. I... Wilhelm, F. H., 8c Roth, W. T. (2003). anxiety: Response generation, response selection,
Embarrassment and social phobia: T he role of para and self-appraisals. B e h a v io u r R e se a rch a n d T h e r a
sympathetic activation. J o u r n a l o f A n x iety D is o r py, 4 5 , 1 5 0 5 - 1 5 1 5 .
d ers, 17, 1 9 7 -2 1 0. Johnson, H. S., Inderbitzen-Nolan, H. M ., 8c Ander
Gilboa-Schechtman, E., Foa, E. B., 8c Amir, N. (1999). son, E. R. (200 6 ). T he Social Phobia Inventory:
Attentional biases for facial expressions in social Validity and reliability in an adolescent community
phobia: The face-in-the-crowd paradigm. C o g n i sample. P sy c h o lo g ic a l A ssessm en t, 18, 2 6 9 - 2 7 7 .
tion a n d E m o tio n , 13, 3 0 5 - 3 1 8 . Jones, W. H., Briggs, S. R., 8c Smith, T. G. (1986). Shy
Gough, H. G., 8c Thorne, A. (1986). Positive, negative, ness: Conceptualization and measurement. J o u r n a l
and balanced shyness: Self-definitions and the reac o f P erson ality a n d S o c ia l P sy ch olog y , 51, 6 2 9 - 6 3 9 .
tions of others. In W. FI. Jones, J. M. Cheek, 8c S. R. Kashdan, T. B. (20 07 ). Social anxiety spectrum and
Briggs (Fids.), S h y n ess: P ersp ectiv es on resea rch a n d diminished positive experiences: Theoretical syn
tre a tm e n t (pp. 2 0 5 - 2 2 6 ) . New York: Plenum Press. thesis and meta-analysis. C lin ical P sy ch o lo g y R e
Hart, E. A., Leary, M. R., 8c Rejeski, W . J . (1989). The v iew , 27, 3 4 8 - 3 6 5 .
measurement of social physique anxiety. J o u r n a l o f Kashdan, T. B., 8c Roberts, J. FI. (200 6). Affective out
S p o rt a n d E x e r c is e P sy ch olog y , 11, 9 4 - 1 0 4 . comes in superficial and intimate interactions: Roles
Heerey, E. A., Keltner, D., 8c Capps, I.. M. (2003). of social anxiety and curiosity. J o u r n a l o f R e se a rc h
Making sense of self-conscious emotion: Linking in P erson ality , 4 0 , 140 -1 67 .
theory of mind and emotion in children with au Kashdan, T. B., 8c Steger, M. F. ( 2 0 0 6). FIxpanding
tism. E m o tio n , 394-400. the topography of social anxiety: An experience-
Heerey, E. A., 8c Kring, A. M. (20 0 7). Interpersonal sampling assessment of positive emotions, positive
consequences of social anxiety. J o u r n a l o f A b n o r events, and emotion suppression. P s y c h o lo g ic a l S ci
m a l P sy ch olog y , 116, 1 2 5 - 1 3 4 . en c e , 17, 1 2 0 - 1 2 8 .
Henderson, L., 8c Zimbardo, P. (2001). Shyness as Kelly, K. M ., 8c Jones, W. H. (1997). Assessment of
a clinical condition: The Stanford model. In W. dispositional embarrassability. A n x iety , Stress a n d
R. Crozier 8c L. E. Alden (Fids.), I n te r n a tio n a l C o p in g : An I n te r n a tio n a l J o u r n a l, 10, 3 0 7 - 3 3 3 .
h a n d b o o k o f s o c ia l a n x iety : C o n c e p ts , resea rch Keltner, S . , 8c Buswell, B. N. (1997). Embarrassment:
190 III. E M O T IO N A L D IS P O S IT IO N S
Its distinct form and appeasement function. P sy ch o Mansell, W., Clark, D. M ., Ehlers, A., 8c Chen, Y.
lo g ic a l B u lletin , 1 2 2 , 2 5 0 - 2 7 0 . (1999). Social anxiety and attention away from
Langston, C. A., 8c Cantor, N. (1989). Social anxiety angry faces. C o g n itio n a n d E m o tio n , 13, 6 7 3 - 6 9 0 .
and social constraint: When making friends is hard. Marcus, D. K., 8c Miller, R. S. (1999). T he perception
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 5 6, of “live” embarrassment: A social relations analysis
649-661. of class presentations. C o g n itio n a n d E m o tio n , 13,
Leary, M. R. (1983a). Brief version of the Fear of Neg 105-117.
ative Evaluation Scale. P erso n a lity a n d S o c ia l P sy Marshall, P. )., 8c Stevenson-Hinde, J . (2001). Behav
c h o lo g y B u lletin , 9, 3 7 1 - 3 7 5 . ioral inhibition: Physiological correlates. In W. R.
Leary, M. R. (1983b). Social anxiousness: The con Crozier 8c L. E. Alden (Eds.), I n te r n a tio n a l h a n d
struct and its measurement. J o u r n a l o f P erson ality b o o k o f so c ia l a n x iety : C o n c e p ts , resea rch a n d in
A ssessm en t, 47, 6 6 - 7 5 . terv en tio n s relatin g to th e s e l f a n d sh y n ess (pp. 5 3 -
Leary, M . R. (1986a). Affective and behavioral co mpo 76). Chichester, UK: Wiley.
nents of shyness: Implications for theory, measure McCroskey, J. C. (1970). Measures of communication-
ment, and research. In W. H. Jones, J . M . Cheek, bound anxiety. S p ee ch M o n o g r a p h s, 37, 2 6 9 - 2 7 7 .
& S. R. Briggs (Eds.), S h y n ess: P ersp ectiv es on r e Miller, R. S. (1987). Empathic embarrassment: Situ
se a r c h a n d tre a tm e n t (pp. 2 7 - 3 8 ) . New York: Ple ational and personal determinants of reactions to
num Press. the embarrassment of another. J o u r n a l o f P ers o n a l
Leary, M . R. (1986b). The impact of interactional im ity a n d S o c ia l P sy ch olog y , 5 3 , 1 0 6 1 -10 69 .
pediments on social anxiety and self-presentation. Miller, R. S. (1992). T he nature and severity of self-
J o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch olog y , 2 2 , reported embarrassing circumstances. P erson ality
122-135. a n d S o c ia l P sy ch o lo g y B u lletin , 18, 1 9 0 - 1 9 8 .
Leary, M . R. (2001a). Shyness and the self: Atten- Miller, R. S. (1995). On the nature of embarrassability:
tional, motivational, and cognitive self-processes in Shyness, social evaluation, and social skill. J o u r n a l
social anxiety and inhibition. In W. R. Crozier 8c L. o f P erson ality , 6 3 , 3 1 5 - 3 3 9 .
E. Alden (Eds.), I n te r n a tio n a l h a n d b o o k o f s o c ia l Miller, R. S. (1996). E m b a r ra s sm e n t: P o ise a n d p er il
a n x iety : C o n c e p ts , re se a r ch a n d in terv en tio n s r e in ev e r y d a y life. New York: Guilford Press.
latin g to th e s e l f a n d sh y n ess (pp. 2 1 7 - 2 3 4 ) . Chich Miller, R. S. (2001a). Embarrassment and social pho
ester, UK: Wiley. bia: Distant cousins or close kin? In S. G. Hofmann
Leary, M . R. (2001b). Social anxiety as an early 8c P. M. DiBartolo (Eds.), F rom s o c ia l a n x ie ty to
warning system: A refinement and extension of the so c ia l p h o b ia : M u ltip le p er sp e c tiv e s (pp. 6 5 - 8 5 ) .
self-presentation theory of social anxiety. In S. G. Boston: Allyn 8c Bacon.
Hofmann 8c P. M. DiBartolo (Eds.), F ro m s o c ia l Miller, R. S. (2001b). Shyness and embarrassment co m
a n x ie ty to s o c ia l p h o b ia : M u ltip le p er sp e c tiv e s pared: Siblings in the service of social evaluation. In
(pp. 3 2 1 - 3 3 4 ) . Boston: Allyn & Bacon. W. R. Crozier 8c L. E. Alden (Eds.), In te r n a tio n a l
Leary, M . R. (20 04 ). T h e cu rse o f th e se lf: Self- h a n d b o o k o f s o c ia l a n x iety : C o n c e p ts , resea rch
a w a r en e ss , eg o tism , a n d th e q u a lity o f h u m a n life. a n d in terv en tio n s relatin g to th e s e l f a n d sh y n ess
New York: Oxford University Press. (pp. 2 8 1 - 3 0 0 ) . Chichester, UK: Wiley.
Leary, M . R ., 8c Baumeister, R. F. ( 2 0 0 0). The nature Miller, R. S. (20 04 ). Emotion as adaptive interpersonal
and function of self-esteem: Sociometer theory. In communication: The case of embarrassment. In L.
M . Zanna (Ed.), A d v a n ce s in e x p e r im e n ta l s o c ia l Z. Tiedens 8c C. W. Leach (Eds.), T h e s o c ia l life o f
p s y c h o lo g y (Vol. 3 2 , pp. 1-6 2). San Diego, CA: e m o tio n s (pp. 8 7 - 1 0 4 ). Cambridge, UK: Cambridge
Academic Press. University Press.
Leary, M . R ., 8c Buckley, K. E. ( 2 0 00 ). Shyness and the Miller, R. S. ( 2 00 7). Is embarrassment a blessing or a
pursuit of social acceptance. In W. R. Crozier (Ed.), curse? In J . L. Tracy, R. W. Robins, 8c J. P. Tangney
S h y n ess: D e v e lo p m e n t, c o n s o lid a tio n a n d ch an g e (Eds.), S elf-c o n sc io u s e m o tio n s (pp. 2 4 5 - 2 6 2 ) . New
(pp. 1 3 9 - 1 5 3 ). New York: Routledge. York: Oxford University Press.
Leary, M . R., 8c Dobbins, S. E. (1983). Social anxiety, Modigliani, A. (1968). Embarrassment and embarrass
sexual behavior, and contraceptive use. J o u r n a l o f ability. S o c io m etr y , 3 1, 3 1 3 - 3 2 6 .
P erson ality a n d S o c ia l P sy ch olog y , 4 5, 1 3 4 7 - 1 3 5 4 . Mogg, K., Bradley, B. P., de Bono, J ., 8c Painter, M.
Leary, M. R., 8c Meadows, S. (1991). Predictors, elici- (1997). Time course of attentional bias for threat
tors, and concomitants of social blushing. J o u r n a l o f information in non-clinical anxiety. B eh a v io u r R e
P erson ality a n d S o c ia l P sy ch olog y , 6 0 , 2 5 4 - 2 6 2 . se a r ch a n d T h era p y , 3 5, 2 9 7 - 3 0 3 .
Ledley, D. R., 8c Heimberg, R. G. (20 0 6). Cognitive Mogg, K., Philippot, P., 8c Bradley, B. P. (200 4). Se
vulnerability to social anxiety. J o u r n a l o f S o c ia l lective attention to angry faces in clinical social
a n d C lin ic a l P sy c h o lo g y 2 5 , 7 5 5 - 7 7 8 . phobia. J o u r n a l o f A b n o r m a l P sy ch olog y , 113,
Levin, J ., 8c Arluke, A. (1982). Embarrassment and 160-165.
helping behavior. P s y c h o lo g ic a l R e p o rts , 5 1, 9 9 9 - Moore, S. G., Dahl, D. W., Gorn, G. J ., 8c Weinberg,
1002 . C. B. (200 6 ). Coping with condom embarrassment.
Maidenberg, E., Chen, E., Craske, M ., Bohn, P., 8c P sy ch olog y , H ea lth a n d M ed icin e, 11, 7 0 —79.
Bystritsky, A. (1996). Specificity of attentional bias Norton , P. J . , Hope, D. A., 8c Weeks, J. W. (200 4 ). The
in panic disorder and social phobia. J o u r n a l o f A n x Physical Activity and Sport Anxiety Scale (PASAS):
iety D iso r d er s, 10, 5 2 9 - 5 4 1 . Scale development and psychometric analysis. A n x
Mansell, W., Clark, D. M ., 8c Ehlers, A. (2003). In iety, Stress a n d C o p in g : An In te r n a tio n a l J o u r n a l,
ternal versus external attention in social anxiety: 17, 3 6 3 - 3 8 2 .
An investigation using a novel paradigm. B eh a v io u r Osborne, M. S., Kenny, D. T., 8c Holsomback, R.
R e se a rc h a n d T h era p y , 4 1 , 5 5 5 - 5 7 2 . (2005). Assessment of music performance anxiety
12. S ocial A nxiou sn ess, Shyness, and E m b arrassab ility 191
J u n e P r ic e T a n g n e y
K e r s t in Y o u m a n
J e f f r e y St u e w ig
192
13. S ham e and G uilt 193
(3) the degree to which the person construes exposure is no more likely to evoke shame
the emotion-eliciting event as a failure of self than guilt.
or of behavior. The most widely used basis for distin
There is surprisingly little empirical evi guishing between shame and guilt— focus
dence that shame and guilt differ reliably on self versus behavior— was first proposed
in terms of the types of situations that elicit by Helen Block Lewis (1971) and more
them. Analyses of personal shame and guilt recently elaborated by Tracy and Rob
experiences provided by children and adults ins (2004) in their appraisal-based model
revealed few, if any, reliably shame-inducing of self-conscious emotions. According to
or guilt-inducing situations (Keltner 8c Lewis, shame involves a negative evaluation
Buswell, 1996; Tangney, 1992; Tangney, of the global self, whereas guilt involves a
M arschall, Rosenberg, Barlow, 8c Wagner, negative evaluation of a specific behavior.
1994; Tracy 8c Robins, 2006). Research Although the self-behavior distinction may,
ers agree that guilt is more narrowly linked at first glance, appear subtle, research sup
to moral transgressions, whereas shame is ports that this differential emphasis on self
evoked by a broader range of situations, in (“/ did that horrible thing”) versus behavior
cluding both “m oral” and “nonmoral” fail (“I did that horrible thing”) sets the stage
ures (Ferguson, Stegge, & Damhuis, 1991; for different emotional experiences and dif
Sabini 8c Silver, 1997; Smith, Webster, Par ferent patterns of motivations and subse
rott, 8c Eyre, 2002), but most types of events quent behavior.
(e.g., lying, cheating, stealing, failing to help Shame is typically the more painful, dis
another, disobeying parents, etc.) are some ruptive emotion because the self, not simply
times cited by people in connection with one’s behavior, is the object of judgment.
feelings of shame and sometimes in connec When people feel shame about the self, they
tion with guilt. feel “small,” worthless, powerless, and ex
A frequently cited distinction between posed. Even though an actual observing
shame and guilt highlights the public-versus- audience need not be present, they often
private nature of the emotion-eliciting situ imagine how one’s defective self would ap
ation (e.g., Benedict, 1946). From this point pear to others. Lewis (1971) described a split
of view, shame is the more “public” emo in self-functioning in which the self is both
tion, arising from exposure to disapproving agent and object of observation and disap
others, whereas guilt is a more “private” proval. Regarding motivations or “action
experience that arises from internal pangs tendencies,” shame is apt to prompt efforts
of conscience. However, empirical tests to hide and defend the diminished, defective
have not supported this distinction (Tang self and to escape the shame-inducing situ
ney et al., 1994; Tangney, Miller, Flicker, ation (Ketelaar 8c Au, 2 0 0 3 ; Lewis, 1971;
& Barlow, 1996). For example, a systematic Lindsay-Hartz, 1984; Tangney, Miller, et
analysis of shame and guilt events described al., 1996; W allbott 8c Scherer, 1995; Wick
by several hundred children and adults er, Payne, 8c Morgan, 1983).
(Tangney et al., 1994) indicated that both Guilt, on the other hand, typically wreaks
emotions were typically experienced in the less havoc. Although painful, guilt is less
presence of others. “Solitary” shame experi overwhelming because the object of condem
ences were no less common than “solitary” nation is a specific behavior, somewhat apart
guilt experiences. Moreover, the frequency from the self. Instead of feeling compelled to
with which others were aware of the respon defend the naked core of one’s identity, peo
dents’ behavior did not vary as a function of ple stricken with guilt are drawn to consider
shame or guilt. Similarly, although achieve their behavior and its consequences. People
ment and personal events are each more feeling guilt often ruminate over the mis
private than relational and familial events, deed, feeling the pain of remorse and regret.
the former were more likely to elicit shame Regarding action tendencies, whereas shame
rather than guilt in a study of personal emo often motivates hiding, guilt often motivates
tion narratives (Tracy 8c Robins, 2006). reparative action (e.g., confession, apology,
Several other studies (Smith et al., 2002) efforts to make amends for the wrongdoing)
provide ample evidence that actual public (de Hooge, 2 0 0 8 ; de Hooge, Zeelenberg, 8c
194 III. E M O T IO N A L D IS P O S IT IO N S
akin to low self-esteem, whereas proneness and is thus of little use in examining shame
to shame is the propensity to experience epi and guilt proneness in psychological and
sodic shame states in response to failures or social functioning. Moreover, because cor
transgressions. relates of shame proneness and guilt prone
ness sometimes differ in sign, measures that
confound shame and guilt may produce
Assessing Individual Differences null results, as the differential relationships
in Proneness to Sham e and Guilt cancel each other out, leading to erroneous
conclusions (e.g., that guilt is not important
How are shame and guilt proneness mea to the context under study). For this reason,
sured at the dispositional or trait level? researchers are advised to use caution when
Often researchers look to people’s self- considering measures that purport to assess
reports to assess dimensions of personality the propensity to experience guilt without
or affective style, but in the case of guilt and explicitly considering shame.
shame, self-reports can be problematic be Fewer measures assess shame proneness
cause most people have difficulty recogniz without reference to guilt proneness. The
ing the distinction between them. Research most widely used measure of this type is
indicates that feelings of shame and guilt fre the Internalized Shame Scale (ISS; Cook,
quently co-occur, that it is difficult for peo 1988). Ironically, the potential conceptual
ple to verbalize the difference between the confound here is not with guilt but rather
two, and that, in Western contexts at least, with its strong conceptual and operational
people are apt to avoid the term sham e alto similarity to low self-esteem. Many of the
gether, using guilt to refer to either or both items composing the ISS were drawn from
emotions. Thus simply asking a person, “In Rosenberg’s (1965) Self-Esteem Scale. Con
general, do you feel guilt rarely, sometimes, sequently, the ISS correlates very highly with
often, or very often?” may tell us something self-esteem (Cook, 1991), raising concerns
about his or her propensity to experience about its discriminant validity.
guilt, shame, or both. Fortunately, a number
of researchers have tackled this measurement
M easures A ssessing
challenge, although much more work needs
(and D istinguishing between)
to be done at the trait level and especially at
Sham e Proneness and G uilt Proneness
the state level.1 Here, we focus on measures
at the level of trait or emotion disposition— Measures designed to distinguish between
proneness to shame and proneness to guilt. shame proneness and guilt proneness vary
substantially in structure or format due to
different conceptual distinctions between
M easures Assessing O nly O n e Disposition
shame and guilt and to the unique challenges
Much of the pioneering work on moral emo posed by the assessment of these two emo
tions, and thus the early measures, focused tions in particular (e.g., people don’t always
exclusively on the propensity to experi use the emotion terms precisely; there is no
ence guilt without consideration of shame identifiable facial expression for guilt). In se
(Buss &c Durkee, 1957; Klass, 1987; Ku- lecting a measure, it is important to consider
gler & Jones, 1992; Mosher, 1966; Zahn- the measure’s suitability for the population
Waxler, Kochanska, Krupnick, & Mayfield, to be studied and the match between empiri
1988). These measures utilized a range of cally supported distinctions between shame
formats— selection of a single adjective, rat and guilt and the way in which they are op
ings of descriptive statements, forced-choice erationalized.
alternatives, ratings of emotional responses
to specific situations, and qualitative analy
Shame- versus Guilt-Inducing Situations
sis of narratives. Because these measures do
not take into account the difference between An approach first introduced by Perlman
shame and guilt, the assessment is apt to (1958) assesses emotional reactivity to
confound the propensity to experience guilt “shame-inducing” versus “guilt-inducing”
with the propensity to experience shame situations, under the assumption that differ
196 III. E M O T IO N A L D IS P O S IT IO N S
ent kinds of situations induce shame versus attachment were positively related to both
guilt. Measures by Crouppen (1976), Joh n emotional dispositions among lesbian, gay,
son and colleagues (1987), and Cheek and or bisexual adults. The correlation between
Hogan (1983) were designed under this as PFQ-2 shame and guilt was .73, affording
sumption. In light of research showing no little discriminant validity. A third, and
reliably shame-specific or guilt-specific elic perhaps most problematic, aspect of global
iting situations, discussed earlier, research checklists is that the process of filling them
ers should consider the rationale for using out is essentially a shame-like task— making
such an approach. global ratings about oneself (or one’s general
affective state) in the absence of any specific
situational context (Tangney, 1995). W here
Global Adjective Checklists
as this approach may be appropriate for the
This approach draws on a list of shame- and assessment of shame, which involves rather
guilt-related adjectives for which people are global negative assessments of the entire self,
asked to make overall ratings of how much it is a problem when attempting to assess the
they experience each affective term or how tendency to experience guilt about specific
well each term describes them. Examples of behaviors apart from the global self.
such measures include Hoblitzelle’s (1987)
Revised Sham e-G uilt Scale (RSGS) and
Scenario-Based Measures
Harder and colleagues’ (Harder, Cutler, &
Rockart, 1992; Harder &c Lewis, 1987) Per A third method for assessing shame prone
sonal Feelings Questionnaire (PFQ) and re ness and guilt proneness is the scenario-
vised PFQ-2. based approach exemplified by the Test of
These measures have the advantages of Self-Conscious Affect (TOSCA) measures
high face validity and ease of administra (Tangney, Wagner, & Gramzow, 1989) and
tion. There are some limitations, however, the Adolescent Shame Measure (ASM; Re-
that may outweigh the advantages. First, imer, 1995). In these measures, people rate
extended adjective checklists require ad how they would respond to a series of com
vanced verbal skills. The RSGS, for exam mon hypothetical situations (e.g., “You make
ple, includes vocabulary that is challenging a mistake at work and find out a coworker
for most college students. The PFQ mea is blamed for the error”). Importantly, the
sures utilize somewhat less sophisticated terms sham e and guilt are not used, thereby
vocabulary. A second limitation is that avoiding confusion common among layper
adjective checklists rely heavily on respon sons. Instead, responses reflect brief phe
dents’ ability to accurately distinguish be nomenological descriptions of shame and
tween “shame” and “guilt” in an abstract guilt reactions (as described in the theoreti
context, which is questionable. Even among cal, phenomenological, and empirical litera
well-educated adults, providing meaning ture). For the scenario described, the shame
ful definitions of shame and guilt is difficult response is “You would keep quiet and avoid
(Lindsay-Hartz, 1984; Tangney & Dearing, the coworker.” The guilt response is “You
2002). As a consequence, the correlation be would feel unhappy and eager to correct the
tween shame proneness and guilt proneness situation.” People rate their likelihood of
assessed via global adjective checklists is responding in each manner indicated. Thus
typically in the .70s, raising concerns about people may endorse both shame and guilt,
multicollinearity and discriminant validity. which can co-occur in a given situation.
Not surprisingly, research using measures Although scenario based, the distinction
such as the RSGS and PFQ-2 rarely identi between shame and guilt here is not in the
fies unique variance in proneness to shame content of the situation but rather in the phe
and guilt that is differentially related to nomenological reaction of the respondent.
other theoretically relevant constructs. For The primary strength of this approach is
instance, using the PFQ -2, Sherry (2007) that the structure of scenario-based measures
found that secure attachment was nega is conceptually consistent with our current
tively correlated with both shame and guilt understanding of guilt as a behavior-specific
proneness, whereas fearful and preoccupied negative appraisal within a given situational
13. S h am e and Guilt 197
context. Scenario-based measures provide eliciting situations that can be used. Efforts
a vehicle for assessing tendencies to experi are generally made to include scenarios
ence guilt about specific behaviors, distinct from diverse settings (e.g., home, work/
from shame about the self, by avoiding the school, peer, and significant others) and
global nature of adjective rating scales that to focus on diverse behaviors (e.g., miss
are more apt to tap into the characteristics ing an appointment, breaking something,
of shame. A second advantage of the sce hurting another person’s feelings, failing a
nario-based approach is that the situation- test). Nonetheless, such measures cover only
specific phenomenological descriptions a small subset of possible transgressions or
of shame and guilt do not require the re failures. In particular, preference is given
spondent to distinguish between the terms to situations and behaviors likely to be en
sham e and guilt. Third, the likelihood of a countered by most respondents at some
defensive response bias is lower than with point in their day-to-day lives— ones that
adjective checklist-type measures. As Lewis people can relate to easily and can readily
(1971) and others have noted, repression or imagine themselves. W hat is missing are less
denial of shame experiences are not uncom common, more idiosyncratic events (e.g.,
mon. Scenario-based measures may partly eating your roommate’s food, behaving in
circumvent people’s defensiveness because sensitively with a mentally ill family mem
they are not directly asked to acknowl ber) and more serious transgressions (e.g.,
edge tendencies to experience “shame” and hitting a child with a car, losing the family
“guilt” but rather to rate phenomenological fortune in an ill-advised business deal) or
descriptions of shame and guilt experiences events for which no reparation seems pos
with respect to specific situations that avoid sible (e.g., involuntary manslaughter) that
use of the emotionally charged words sham e are irrelevant to most respondents but may
and guilt. dominate a specific person’s emotional life.
Scenario-based measures are easily adapt These events may lead individuals to experi
ed for use with younger participants. There ence “maladaptive” levels of guilt (Luyten,
are TOSCAs for adolescents and for children Fontaine, & Corveleyn, 2002). Stated an
ages 8 -1 2 (see Tangney & Dearing, 2002), other way, measures such as the TO SCA are
and Stegge and Ferguson (1990) have de less apt to capture intense but more circum
veloped the Child Attribution and Reaction scribed shame and guilt experiences focused
Survey— Child Version (C-CARS) for chil in a specific domain (e.g., failures at dieting,
dren as young as 5 years. Common to these marital infidelity, mistreating a vulnerable
measures is a range of age-appropriate situ or stigmatized family member).2
ations (sampling from home, work/school, A third concern is whether scenario-based
peer, and other domains) that are likely to measures such as the TO SCA assess emo
elicit shame and/or guilt responses. tional response tendencies (shame and guilt)
O f course, scenario-based measures have as opposed to emotion-prompted behavior
limitations. In general, they yield some (hiding vs. amending). Some researchers
what lower internal consistency estimates have raised the possibility that in eschewing
of reliability than adjective checklists, with the use of the terms sham e and guilt in favor
Cronbach’s alphas ranging from .71 to .86 of phenomenological descriptions, scenario-
for checklists versus .61 to .83 for scenario- based proponents may have thrown out the
based measures (Tangney & Dearing, 2002). emotional baby with the linguistic bathwater,
Coefficient alphas, however, are apt to un as it were (Eisenberg, 2 0 0 0 ; Ferguson, Brug-
derestimate reliability due to the variability man, White, & Eyre, 20 0 7 ). A close analy
introduced by the use of different scenarios. sis, however, reveals that only 25% (4 of 16)
In contrast, test-retest estimates of reliabil of guilt responses on the TOSCA-3 describe
ity for scenario-based measures tend to be actual behavior (hiding for shame, amend
higher than internal consistencies, equiva ing for guilt). The rest refer to thoughts and
lent to those observed for global adjective feelings about what one should have done in
checklist measures. the past or what one should do in the fu
A second limitation is the necessary con ture. Only 2 of the 16 shame items describe
straint on the types of shame- and guilt- behavioral responses. More important, the
198 III. E M O T IO N A L D IS P O S IT IO N S
shame items, such as “feel incompetent,” list: Screening Version; Hare, Cox, & Hare,
“feel inadequate,” “feel immature,” “think: 1995), was unrelated to proneness to shame
‘I ’m terrible’” and the guilt items, such as and only weakly negatively correlated with
“think: ‘this is making me anxious. I need proneness to guilt (r = -.1 6 ), showing that
to either fix it or get someone else to’,” “feel psychopaths are not prone to either shame
unhappy and eager to correct the situation” or guilt. But low scores on the TO SCA do
are anything but affectively barren. The not imply a pathological absen ce of shame
TO SC A -C and TOSCA-A also hold up well and guilt. Stated another way, it is meaning
under this same scrutiny. The TO SCA-SD, ful for someone to score (1) higher than his
developed for inmates, however, is heavily or her peers on shame but not guilt, (2) high
weighted toward behavior, based on initial er than his or her peers on both shame and
assumptions about the need to use concrete guilt, and (3) higher than his or her peers on
responses with this population. Based on guilt but not shame. Low scores on both are
several years of research with jail inmates, not terribly informative.
we believe it is feasible to employ language
and concepts similar to those employed on
the other versions of the TO SCA . Thus, the W hat Is S h am e-Free Guilt?
TOSCA-SD is currently under revision.
In summary, global adjective checklists Theoretically, the adaptive features of guilt
and scenario-based measures each have pros should be most evident when unaccompanied
and cons. Both approaches yield reasonably by the painful feelings of shame (Tangney
valid indices of proneness to shame, but sce- & Dearing, 2002). Similarly, shame unac
nario-based measures seem uniquely able to companied by guilt may have unique nega
capture proneness to guilt about behaviors, tive consequences. To model this important
independent of shame about the self. unique variance, it is common to calculate
semipartial (part) correlations that reflect
“shame-free” guilt and “guilt-free” shame.
Sham e Proneness and Guilt Proneness For example, in one study the relationship
A re N ot at Opposite Ends of parental rejection to shame proneness (r =
o f a Single Continuum .15) and guilt proneness (r = -.0 9 ) changed
substantially once the semipartial correla
Just as people may experience shame, guilt, tion was used (rs = .27 and -.2 4 ) (Stuewig
or some combination of the two in response & McCloskey, 2005). Another way to think
to a single event, at the dispositional level about the relationship of shame and guilt to
some people are prone to shame, some to constructs is as suppressors (Paulhus, R ob
guilt, and some to both. The correlation be ins, Trzesniewski, & Tracy, 2004). As with
tween shame proneness and guilt proneness self-esteem and narcissism, for instance, dif
is positive— about .42 for the TO SCA -3, ferential relationships of shame proneness
higher among children using the TO SCA -C and guilt proneness become evident once
(about .6), and lower among inmates using each is residualized on the other. These dif
the TOSCA-SD (about .2). We believe these ferential patterns of results have been found
two moderately correlated measures repre in many samples (Dearing, Stuewig, &c
sent unipolar as opposed to bipolar dimen Tangney, 2 0 0 5 ; Paulhus et al., 2 0 0 4 ; Tang
sions (see Russell & Carroll, 1999). Specifi ney, 1991; Tangney, Wagner, Fletcher, &
cally, high scores on shame proneness and Gramzow, 1992) and are theoretically con
guilt proneness carry meaning, but low scores sistent with the notion that it is the capacity
are less informative, particularly for shame. to experience guilt about behaviors without
There is no polar opposite to shame prone the interference of shame about the self that
ness. The unipolar, as opposed to bipolar, leads to more adaptive intrapersonal and
nature of these scales was underscored in our interpersonal outcomes. For this reason, in
longitudinal study of jail inmates (Tangney, terpretation of the correlates of shame-free
Mashek, & Stuewig, 200 7 ). In a sample of guilt (and sometimes guilt-free shame) may
500 male and female inmates, psychopathy, be necessary to identify relationships that
a serious form of antisocial personality dis might otherwise be obscured by suppressor
order (assessed by the Psychopathy Check effects.3
13. Sham e and Guilt 199
guilt when one legitimately takes the respon tween shame proneness and externalization
sibility for one’s failures and transgressions. of blame, hostility, anger, and unconstruc-
A recent study comparing two clinical popu tive expression of anger (Ahmed & Braith-
lations suggested that guilt proneness might waite, 2 0 0 4 ; Andrews, Brewin, Rose, &
be related to psychopathology. Rusch and Kirk, 2 0 0 0 ; Bennett, Sullivan, & Lewis,
colleagues (2007) reported that guilt prone 2 0 0 5 ; Harper & Arias, 2 0 0 4 ; Flarper, Cer-
ness was higher in women with comorbid cone, & Arias, 2 0 0 5 ; Lutwak, Panish, Ferra
borderline personality disorder (BPD) and ri, & Razzino, 20 0 1 ; Robinson et al., 2007).
posttraumatic stress disorder as compared Shame-prone individuals may also express
with women with only a BPD diagnosis. verbal or physical aggression, although the
Furthermore, shame had an analogous pathways and circumstances leading to such
though nonsignificant positive relationship behavior are unclear (Stuewig & Tangney,
with comorbidity. Because shame and guilt 2007). Perhaps feelings of shame prompt a
were not partialled out, however, interpre strong tendency to become defensive, shift
tation of these results should be made with blame, and attack others (verbally or physi
caution. It is unclear whether shame-free cally) in order to escape the pain of shame.
guilt would be stronger among those women This proclivity to lash out may satisfy the
with comorbid diagnoses. short-term goal of regaining a sense of con
trol and moral superiority, but at what cost?
It is difficult to maintain healthy relations
Sham ing, B lam ing, and M aim ing
when friends, coworkers, and loved ones are
One robust empirical finding involves the frequently exposed to outbursts of anger. In
differential link of shame and guilt to contrast, guilt proneness is unrelated to an
blame and anger. In addition to assessing ger— that is, guilt-prone people are as prone
proneness to shame and guilt, the TOSCA to anger as anyone else. But when angered,
measures assess externalization of blame, guilt-prone individuals are inclined to man
initially included as filler items. Externaliza age their anger constructively (e.g., through
tion of Blame (blaming the situation or other nonhostile discussion or direct corrective
people for one’s failure or transgression) has action), and they are disinclined toward
emerged as a reliable, valid scale in its own aggression (Ahmed & Braithwaite, 2 0 0 4 ;
right. As expected, guilt-prone individuals Lutwak et al., 2 0 0 1 ; Paulhus et al., 2 0 0 4 ;
are inclined to take responsibility for their Tangney, Wagner, Hill-Barlow, M arschall,
blunders; externalization of blame has been & Gramzow, 1996).
consistently negatively correlated with prone
ness to guilt. But, whereas attribution theory
Risky, Illegal, a n d /o r Im m oral Behavior
would predict that shame-prone individu
als would be inclined to blame themselves Because shame and guilt are painful emo
for their failures, studies consistently show tions providing negative feedback for wrong
a positive link between shame proneness doing, it is often assumed that both motivate
and blaming others. How is it that shame- individuals to do the right thing. But re
prone people (in attributional terms, people search tells a different story. There is stron
who make internal, stable, and global attri ger empirical support for the moral func
butions for failures and transgression; see tion of guilt as opposed to shame (Stuewig
Tangney, 1990; Tracy & Robins, 2 0 0 6 ) are & Tangney, 2007). Among all age groups,
also inclined to blame others? People suffer guilt proneness is associated with low levels
ing from the pain and self-diminishment of of consensually immoral behavior, but there
shame may become defensive and angry and is little evidence for the presumed moral
attempt to deflect blame outward. Lewis inhibitory functions of shame. If anything,
(1971) described the “humiliated fury” un shame-prone individuals have difficulty fol
leashed by clients’ shame in clinical practice, lowing the straight and narrow. In one study
and Scheff’s (1987) qualitative research de of incarcerated adolescent offenders and
scribes a “shame-rage spiral” that can lead a comparison group from the community,
to blame, rage, and occasionally aggression. shame-free guilt proneness was negatively
In fact, research with individuals of all related to anger and antiauthority and dis
ages consistently demonstrates a link be trustful attitudes, whereas shame proneness
13. S ham e and Guilt 201
was positively related to anger and distrust unrelated to risky intravenous-drug use dur
ful attitudes across groups (Robinson et al., ing the year prior to incarceration, but guilt
2007). Contrary to expectations, however, was negatively related to number of sexual
shame proneness and guilt proneness only partners and to an index of risky sexual be
marginally differentiated between the two havior (Stuewig, Tangney, Mashek, Forkner,
groups. Using a sample of incarcerated indi & Dearing, in press).
viduals, Hosser, Windzio, and Greve (2008)
found that shame was related to higher re
cidivism rates, whereas guilt was related to Understanding Adaptive
less recidivism. In a study of college students, and M aladaptive Effects o f Shame
Tibbetts (2003) entered a number of shame and Guilt: Mediational Models
and guilt measures simultaneously into a re
gression analysis; the TO SCA Shame scale Much evidence shows that shame and guilt
was unrelated to illegal behaviors, whereas are differentially related to a number of
guilt was negatively related to illegal acts. psychological and behavioral constructs.
Similarly, in a longitudinal study, Stuewig Research has begun to delve deeper by ex
and McCloskey (2005) found a negative re amining the mediational pathways that un
lationship between guilt proneness and de derlie these relationships. A number of stud
linquency; shame proneness was unrelated ies show support for several hypothesized
to delinquency. processes that may explain how shame and
Shame proneness and guilt proneness are guilt influence social behavior. Notably,
also related to substance use and abuse. anger and externalization of blame appear
Compared with individuals in commu to mediate the relationship between shame
nity settings, adults in recovery programs and aggression. Specifically, men’s anger
had lower guilt-prone scores and higher has been found to mediate the relationship
shame-prone scores (Meehan et al., 1996; between shame proneness and perpetra
O ’Connor, Berry, Inaba, Weiss, & M orri tion of psychological abuse in dating rela
son, 1994). Among college students and jail tionships (Harper et al., 2005). Stuewig,
inmates, shame proneness was consistently Tangney, Heigel, and Harty (2006) found
positively related to alcohol and drug prob that across four diverse samples (early ado
lems. There was also evidence for a negative lescents, at-risk older adolescents, college
relationship between substance use prob students, and incarcerated adults), external
lems and guilt proneness (Dearing et al., ization of blame mediated the relationship
2005). In a longitudinal study, shame and between shame proneness and both verbal
guilt proneness in the fifth grade predicted and physical aggression. Guilt proneness
alcohol and drug use as reported at 18 years had the opposite effect in that proneness to
of age (Tangney, Stuewig, Kendall, Rein guilt was negatively related to aggression in
smith, & Dearing, 200 6 ). Children high in three of the four samples, partially mediated
shame tended to start drinking earlier than through other-oriented empathy and accept
those low in shame and were more likely to ing responsibility.
later use heroin, “uppers,” and hallucino Ashby, Rice, and M artin (2006) identi
gens. Those high in guilt started drinking at fied shame as a mediator of the effects of
a later age than those low in guilt and were maladaptive perfectionism on depression in
less likely to use heroin, with similar trends a sample of college students. Among men,
for marijuana and “uppers.” internalized shame fully mediated the rela
Very few studies have examined the rela tionship. Among women, maladaptive per
tionship of the moral emotions to other risky fectionism directly predicted depression, but
behaviors, such as needle use or risky sexual there was also partial mediation through
behavior, although one study of college stu shame and low self-esteem. This finding is
dents reported little relationship between re consistent with earlier empirical support of
ports of previous high-risk sexual behaviors the relationship between negative perfection
and current state shame or guilt (Murray, ism and both state and trait shame and the
Ciarrocchi, & Murray-Swank, 200 7 ). In an negative relationship between adaptive per
other study of recently incarcerated inmates, fectionism and state shame (Fedewa, Burns,
shame proneness and guilt proneness were & Gomez, 2005).
202 III. E M O T IO N A L D IS P O S IT IO N S
In a sample of several hundred undergrad ge, 1995; Kochanska, 1993; Kochanska &
uates, W illiamson, Sandage, and Lee (2007) Aksan, 2 0 0 6 ; Lewis, 1992; Miyake & Ya-
evaluated several mediational models to ex mazaki, 1995; Potter-Efron, 1989; Rosen
amine the implications of social connected berg, 1997; Zahn-W axler & Robinson,
ness for guilt proneness, shame proneness, 1995)— the research and measurement lags
differentiation of self, and hope. At the bi- far behind theory (Eisenberg, 2000).
variate level, proneness to shame-free guilt In the developmental literature, simi
was positively related to social connectedness larities between parents and offspring have
and hope, whereas proneness to (guilt-free) been found for a number of attributes and
shame was negatively associated with social behaviors (Serbin & Stack, 1998). There is
connectedness, hope, and differentiation of good reason to expect intergenerational con
self. Support was found for two of three al tinuity for shame proneness and guilt prone
ternative models. In one, social connected ness, as well. There may be a direct link be
ness positively predicted guilt and negatively tween parents’ affective styles and those of
predicted shame. In turn, guilt proneness their children through behavioral modeling.
positively predicted both hope and differ Children observe how their parents react
entiation of the self; shame negatively pre to negative events and may learn, via direct
dicted both hope and differentiation of the modeling, that a certain pattern of emotion
self. In an alternative model, dependent vari al, cognitive, and behavioral responses is ap
ables (hope and differentiation of self) were propriate in certain kinds of situations. To
switched with mediating variables (shame the degree that direct modeling occurs, one
and guilt); this second model, with shame would expect a direct link between parents’
and guilt as dependent variables, fit equally affective styles and those of their children.
well. Little research has examined whether
In sum, the bivariate correlates of prone shame proneness and guilt proneness show
ness to shame and guilt have been pretty continuities across generations. In one lon
well mapped out. Research that examines gitudinal study of fifth-grade children with
more complex models involving mediation follow-up in seventh grade, measures of
and moderation has just begun. We antici shame proneness and guilt proneness were
pate that future research will expand on this collected from children, parents, and grand
work considerably, clarifying the functional parents. Children were interviewed a third
nature of the relationship of shame prone time when they were 18. There was only very
ness and guilt proneness to a range of per modest evidence of transmission of shame
sonality factors, psychological symptoms, and guilt, with a weighted mean correlation
and patterns of interpersonal behavior. of .09 across generations (Stuewig, Kendall,
& Tangney, 2004). Although the direct rela
tionship between parent and child was mini
W here D o Sh am e-Prone mal, there may be important moderators of
and G uilt-Prone Styles C om e From ? intergenerational continuity in shame-prone
and guilt-prone styles. For example, age may
Given the implications of shame proneness play a role, such that the similarity between
and guilt proneness described thus far, it is parent and child may be strongest at a simi
clear that these individual differences mat lar developmental stage for each.
ter. How does one become shame or guilt Perhaps families play other roles in the de
prone? This remains largely a mystery. Few velopment of shame-prone and guilt-prone
prospective studies have examined the de styles. Tendencies toward shame proneness
velopment of shame and guilt proneness, es may be perpetuated through family dynam
pecially starting in early childhood (Mills, ics shaped by family members’ affective
2 0 0 5 ; Reimer, 1996). Whereas a large styles that in turn reinforce individual mem
number of possible mechanisms have been bers’ characteristic emotional responses. The
proposed— including genetic/temperament literatures on family systems and on code
factors (e.g., Dienstbier, 1984; Kochanska, pendence describe, for example, a sham e-
1993; Zahn-W axler & Robinson, 1995) and based family system that is characterized by
socialization factors, especially parenting maladaptive patterns of communication and
style (e.g., Barrett, 1995; Ferguson & Steg- extremes of family conflict or enmeshment
13. Sham e and G uilt 203
(Bradshaw, 1988; Fossum & M ason, 1986). & McCloskey, 2005). There are a number of
However, little empirical research has been possible reasons for these inconsistent find
conducted in this area. ings, including small sample sizes and dif
Another possibility is that parents’ child- ferences in operational definitions. An inter
rearing practices are most important to the esting hypothesis is that the specific findings
development of children’s moral affective may depend on the coping style and recovery
styles. In general, studies provide support process of the individual (Bonanno, Keltner,
for parental practices as a component in the & Noll, 2 0 0 2 ; Negrao, Bonanno, Noll, Put
socialization of moral emotions. In a study nam, & Trickett, 2005).
of 5- to 12-year-old children, Ferguson and Finally, temperament may play a role in
Stegge (1995) found that children’s guilt was the development of proneness to shame and
associated with parents’ reports of induction to guilt. The strongest support for a tem
and parental anger in negative situations, peramental perspective on children’s de
whereas children’s shame was associated velopment of conscience has been reported
with parental hostility, little recognition of by Kochanska and colleagues (Kochan-
positive outcomes, and a lack of discipline. ska, DeVet, Goldman, Murray, & Putnam,
Alessandri and Lewis (1993) reported that 1994; Kochanska, Gross, Lin, &C Nichols,
parents’ specific (but not global) negative 2002), who found that expression of behav
comments were associated with children’s ioral and affective discomfort subsequent
displays of shame, an unexpected result. to misbehaving was related to temperamen
Gilbert, Allan, and Goss (1996) found that tal qualities of fearfulness and reactivity.
recalled put-downs and shaming from child In one study, Kochanska and colleagues
hood were associated with shame proneness (2002) found that fearfulness at 2 2 , 33, and
in adulthood. Finally, evidence suggests that 45 months of age was related concurrently
children of depressed mothers may be at risk to guilt (measured as observed discomfort
for developing “maladaptive” patterns of after misbehaving) at each time. Further
guilt (Zahn-Waxler & Robinson, 1995). more, this measure of guilt (a composite
Child maltreatment in its different forms from all three previous waves) mediated
(physical abuse, sexual abuse, harsh parent the relationship between fearfulness and
ing, neglect) may leave children vulnerable a tendency to violate rules at 56 months.
to the development of a shame-prone dispo Toddlers who responded fearfully to risky
sition and less likely to acquire an adaptive activities were more likely to show discom
guilt-prone style. Research indicates a link fort after transgressing, which in turn led
between retrospective reports of abuse and to lower likelihood of violating rules. These
shame (Andrews, 1995; Andrews & Hunter, studies did not differentiate between shame
1997; Hoglund & Nicholas, 1995; Webb et and guilt, however.
al., 200 7 ). Alessandri and Lewis (1996) ob
served that mothers’ negative behaviors were
correlated with children’s shame reactions Gender and Culture
during laboratory tasks and that girls with a
history of maltreatment showed higher non A consistent empirical finding is that women
verbal shame than did girls with no history have higher levels than men of both shame
of abuse. Moreover, longitudinal research proneness and guilt proneness. This gender
shows that negative or harsh parenting is as difference has been observed, without excep
sociated with shame proneness (Bennett et tion, in studies involving over 3 ,0 0 0 individ
al., 2 0 0 5 ; M ills, 2 0 0 3 ; Stuewig & McClo- uals from early childhood through the elder
skey, 2005). years and from all walks of life (Tangney &
Taken together, evidence supports a link Dearing, 2002). Females’ higher scores on
between emotional or physical abuse and both shame and guilt proneness could be
proneness to shame. Surprisingly, evidence due to a number of factors: Females may, in
for the relationship between sexual abuse fact, experience shame and guilt more often
and shame is less clear-cut, writh some stud and more intensely, females may be more
ies finding positive results and others finding willing and/or able to report on emotional
null results (Alessandri & Lewis, 1996; An experiences, females may be more self-
drews, 1995; Andrews et al., 2 0 0 0 ; Stuewig reflective and hence more inclined to expe
204 III. E M O T IO N A L D IS P O S IT IO N S
rience self-conscious emotions, and females ers and to feel anger relative to their less
may be more attuned to issues of morality, shame-prone peers. Notably, in no case did
especially those involving interpersonal re shame seem to inhibit aggression-relevant
lationships (Gilligan, 1982). In short, mul cognitions, emotion, or behavior. In short,
tiple features of these self-reflective, moral although there were significant cultural dif
emotions may account for higher shame ferences in children’s propensity to experi
proneness and guilt proneness among girls ence self-conscious emotions, the correlates
and women. Tangney and Dearing (2002) of individual differences in shame and guilt
cautioned that females’ higher propensity were remarkably similar across these three
to “m oral” emotions does not necessarily cultures in at least one important domain—
imply that they are more moral, as the moral anger and aggression.
benefits of proneness to guilt may be negated
to some degree by the negative consequences
of proneness to shame. Psychobiological Correlates
Theory and research presented thus far
has been grounded in traditional Western A recent focus of moral-emotions research is
cultural norms that emphasize ideals of in the identification of psychobiological mark
dividualism and responsibility for one’s own ers of shame and guilt in response to labora
actions, but non-Western cultures embrace tory manipulations designed to threaten the
more collectivist ideals of interdependence social self (Dickerson, Kemeny, Aziz, Kim,
and group responsibility. Cross-cultural & Fahey, 2 0 0 4 ; Gruenewald, Kemeny, Aziz,
research highlights how culture may influ & Fahey, 2 0 0 4 ; see Dickerson, Gruenewald,
ence the intensity and frequency of moral & Kemeny, 2 0 0 4 , for a review). Participants
emotions, as well as their causes and conse who wrote about incidents involving heavy
quences (Lagattuta & Thompson, 2 0 0 7 ). For doses of self-blame, compared with those
example, Furukawa, Tangney, Higashihara, who wrote about more mundane daily ac
and Pak (2008) examined differences in tivities, evidenced increased levels of self-
proneness to shame, guilt, and pride among reported shame (and guilt) from pretest to
children residing in Japan, Korea, and the posttest. Importantly, increases in shame
United States. Significant group differences (but not guilt or general negative affect) co
were observed in children’s propensity to incided with increased proinflammatory cy
experience self-conscious emotions. Specifi tokine activity (Dickerson, Kemeny, et al.,
cally, Japanese children were more shame 2004).
prone than children in the United States and Other immunological research is equally
Korea. In this sense, Japan may represent a suggestive: Among H IV positive individuals,
“shame” culture (Benedict, 1946; Hogan persistent feelings of shame (but not other
& Sussner, 2001) in a way that is distinct negative emotions) were positively related
from another Asian culture, Korea. Korean to prospective T-cell decline, an indicator
children were more prone to guilt than Japa of compromised immune function (Weitz-
nese and American children (results incon man, Kemeny, & Fahey, 2004). Experiences
sistent with the notion of a Western “guilt of shame have also been linked to elevated
culture”). Regarding the correlates of shame cortisol in studies of adults (Gruenewald
proneness, it was hypothesized that shame et al., 2004) and children (Lewis & Ram
would be less problematic among Japanese say, 2002). Importantly, Dickerson, Gru
children relative to those raised in Korea enewald, and colleagues (2004) noted that
and the United States, owing to the fact that shame, cortisol, and proinflammatory cy
shame is more normative and would there tokine system activation increased specifi
fore be less painful in the self-critical Japa cally in response to social-evaluative threat
nese culture. There were, however, surpris (negative social evaluation and rejection) but
ingly few cross-cultural differences in the not in response to more general negative af
relationship of shame to aggression-related fect or distress. They hypothesized that indi
cognitions, emotions, and behavior. In the vidual differences in shame proneness may
face of failure or transgression, shame-prone be correlated with individual differences in
children in Japan, Korea, and the United immunosystem responsivity and that state
States were all more inclined to blame oth experiences of shame and related emotions
13. S h am e and Guilt 205
Benedict, R. (1946). T h e ch ry s a n th em u m a n d the Positive and negative perfectionism and the shame/
sw o rd . Boston: Houghton Mifflin. guilt distinction: Adaptive and maladaptive charac
Bennett, D. S., Sullivan, M . W., 8c Lewis, M . (2005). teristics. P erson ality a n d In d iv id u a l D iffer en c es ,
Young children’s adjustment as a function of mal 38, 1609-1619.
treatment, shame, and anger. C h ild M a ltre a tm e n t, Ferguson, T. J ., Brugman, D., White, J ., 8c Eyre, H.
30(4), 3 1 1 - 3 2 3 . L. (2007). Shame and guilt as morally warranted
Bonanno, G., Keltner, D., 8c Noll, J. (2002). When the experiences. In J. L. Tracy, R. W. Robins, 8c J. P.
face reveals what words do not: Facial expressions Tangney (Eds.), T h e se lf-c o n s c io u s e m o tio n s : T h e o
of emotions, smiling, and the willingness to disclose ry a n d re se a r ch (pp. 3 3 0 - 3 4 8 ) . New York: Guilford
childhood sexual abuse. J o u r n a l o f P erso n a lity a n d Press.
S o c ia l P sy ch olog y , 83(1), 9 4 - 1 1 0 . Ferguson, T. J ., & Stegge, H. (1995). Emotional states
Bradshaw, J . (1988). H ea lin g th e sh a m e th at b in d s y ou . and traits in children: The case of guilt and shame. In
Deerfield Beach, FL: Health Communications. J. P. Tangney 8c K. W. Fischer (Eds.), S elf-c o n sc io u s
Buss, A., 8c Durkee, A. (1957). An inventory for as e m o tio n s : T h e p s y c h o lo g y o f sh a m e, g u ilt, e m b a r
sessing different kinds of hostility. J o u r n a l o f C o n ra ssm en t, a n d p r id e (pp. 1 7 4 - 1 9 7 ). New York:
su ltin g P sy ch olog y , 22(4), 3 4 3 - 3 4 9 . Guilford Press.
Cheek, J. M ., 8c Hogan, R. (1983). Self-concepts, self Ferguson, T. J ., Stegge, H ., 8c Damhuis, I. (1991).
presentations, and moral judgments. In J. Suls 8c A. Children’s understanding of guilt and shame. C h ild
G. Greenwald (Eds.), P sy c h o lo g ic a l p e r s p e c tiv e s on D e v e lo p m e n t, 6 2 , 8 2 7 - 8 3 9 .
th e s e l f (Vol. 2 , pp. 2 4 9 - 2 7 3 ) . Hillsdale, NJ: Erl Fossum, M. A., Sc M ason, M . J. (1986). F acin g s h a m e:
baum. F am ilies in reco v ery . New York: Norton.
Cook, D. R. (19 8 8, August). T h e m ea s u r e m e n t o f Furukawa, E., Tangney, J. P., Higashihara, F., Sc Pak,
s h a m e : T h e In te rn a liz ed S h a m e S cale. Paper pre H. (2008). C ro ss-cu ltu ra l co n tin u ities a n d d is c o n ti
sented at the annual meeting of the American Psy n u ities in s h a m e, gu ilt, a n d p r id e : A stu d y o f c h il
chological Association, Atlanta, GA. dren in Ja p a n , K o r e a , a n d th e U n ited S tates. M a n u
Cook , D. R. (1991). Shame, attachment, and addic script under review.
tions: Implications for family therapists. C o n t e m Gilbert, P., Allan, S., 8c Goss, K. (1996). Parental rep
p o r a r y F am ily T h era p y , 13, 4 0 5 - 4 1 9 . resentations, shame, interpersonal problems, and
Crouppen, G. A. (1976). Field dependence- vulnerability to psychopathology. C lin ica l P sy c h o l
independence in depressive and “normal” males as og y a n d P sy ch o th era p y , 3, 2 3 - 3 4 .
an indicator of relative proneness to shame or guilt Gilligan, C. (1982). In a d iffe r e n t v o ic e : P sy c h o lo g i
and ego-functioning. D is s er ta tio n A b stra cts I n t e r c a l th e o r y a n d w o m e n ’s d e v e lo p m e n t. Cambridge,
n a tio n a l, 37, 4 6 6 9 B - 4 6 7 0 B . (UMI No. 7 7 - 6 2 9 2 ) M A : Harvard University Press.
de Hooge, I. E. (2 008). M o r a l e m o tio n s in d ec is io n Gruenewald, T. L., Kemeny, M . E., Aziz, N., 8c Fahey,
m a k in g : T o w a rd s a b e t t e r u n d er sta n d in g o f sh a m e J . L. (2004). Acute threat to the social self: Shame,
a n d gu ilt. Unpublished doctoral dissertation, Uni social self-esteem, and cortisol activity. P s y c h o s o
versity of Tilburg, The Netherlands. m a tic M e d icin e, 6 6 , 9 1 5 - 9 2 4 .
de Hooge, I. E., Zeelenberg, M ., 8c Breugelmans, S. M. Harder, D. W., Cutler, L., 8c R ock art, L. (1992). As
(200 7). Moral sentiments and cooperation: Differ sessment of shame and guilt and their relationship
ential influences of shame and guilt. C o g n itio n a n d to psychopathology. Jo u r n a l o f P erson ality A ssess
E m o tio n , 2 1 , 1 0 2 5 - 1 0 4 2 . m en t, 5 9, 5 8 4 - 6 0 4 .
Dearing, R. L., Stuewig, J ., 8c Tangney, J. P. (2005). Harder, D. W., 8c Lewis, S. J. (1987). The assessment
On the importance of distinguishing shame from of shame and guilt. In J . N. Butcher 8c C. D. Spiel-
guilt: Relations to problematic alcohol and drug berger (Eds.), A d v a n ce s in p e r s o n a lity a s se ss m e n t
use. A d d ic tiv e B e h a v io r s, 30 ( 7 ), 1 3 9 2 - 1 4 0 4 . (Vol. 6, pp. 8 9 - 1 1 4 ). Hillsdale, NJ: Erlbaum.
Dickerson, S. S., Gruenewald, T. L., 8c Kemeny, M . E. Hare, S. D., Cox, D. N., 8c Hare, R. D. (1995). T h e
(200 4). When the social self is threatened: Shame, H a re P sy ch o p a th y C h e c k lis t: S creen in g V ersion
physiology, and health. J o u r n a l o f P erson ality , 7 2, (PCL-.SV). Toronto, Ontario, Canada: Multi-
1191-1216. Health Systems.
Dickerson, S. S., Kemeny, M. E., Aziz, N., Kim, K. Harper, F. W. K., 8c Arias, I. (2004). T he role of shame
H., 8c Fahey, J. L. (200 4). Immunological effects of in predicting adult anger and depressive symptoms
induced shame and guilt. P sy c h o s o m a tic M ed icin e, among victims of child psychological maltreatment.
66, 124-131. J o u r n a l o f F am ily V io len ce, 19(6), 3 6 7 - 3 7 5 .
Dienstbier, R. A. (1984). The role of emotion in moral Harper, F. W. K., Cercone, J ., 8c Arias, I. (2005). The
socialization. In C. Izard, J . Kagan, 8c R. B. Za- role of shame, anger, and affect regulation in men’s
jonc (Eds.), E m o tio n s , c o g n itio n s , a n d b e h a v io r s perpetration of psychological abuse in dating rela
(pp. 4 8 4 - 5 1 3 ) . New York: Cambridge University tionships. J o u r n a l o f In t e r p e r s o n a l V io len ce, 2 0,
Press. 1648-1662.
Eisenberg, N. (200 0). Emotion, regulation, and moral Hoblitzelle, W. (1987). T h e m ea su re m en t o f sh a m e
development. A n n u al R e v ie w o f P sy ch olog y , S I, a n d g u ilt a n d th e r o le o f s h a m e in d e p r e s s io n . Un
665-697. published doctoral dissertation, Yale University.
Eisenberg, N., Valiente, C., 8c Champion, C. (2004). Hogan, J. D., 8c Sussner, B. D. (2001). Cross-cultural
Empathy-related responding: Mora l, social, and psychology in historical perspective. In L. L. Adler
socialization correlates. In A. G. Miller (Ed.), T h e 8c U. P. Gielen (Eds.), C ross-cu ltu ral to p ic s in p s y
s o c ia l p s y c h o lo g y o f g o o d a n d e v il (pp. 3 8 6 - 4 1 5 ) . c h o lo g y (pp. 1 5 - 2 8 ). Westport, CT: Praeger.
New York: Guilford Press. Hoglund, C. L., 8c Nicholas, K. B. (1995). Shame,
Fedewa, B. A., Burns, L. R., 8c Gomez, A. A. (2005). guilt, and anger in college students exposed to abu
13. Sham e and Guilt 207
Robins, & J. P. Tangney (Eds.), S elf-c o n s c io u s e m o Tangney, J. P. (1990). Assessing individual differences
tio n s: T h e o r y a n d resea rch (pp. 4 4 3 - 4 6 7 ) . New in proneness to shame and guilt: Development of the
York: Guilford Press. Self-Conscious Affect and Attribution Inventory.
Robinson, R., Roberts, W. L., Strayer, J ., 8c Koopman, J o u r n a l o f P erson ality a n d S o c ia l P sy ch o lo g y , 59,
R. ( 2 00 7). Empathy and emotional responsiveness 102 - 111 .
in delinquent and non-delinquent adolescents. S o Tangney, J. P. (1991). Moral affect: The good, the bad,
c ia l D e v e lo p m e n t , 16, 5 5 5 - 5 7 9 . and the ugly. J o u r n a l o f P erson ality a n d S o c ia l P sy
Rosenberg, K. L. (1997). T h e so c ia liz a tio n o f sh a m e ch o lo g y , 61, 5 9 8 - 6 0 7 .
a n d g u ilt. Unpublished doctoral dissertation, Tangney, J. P. (1992). Situational determinants of
George Mason University. shame and guilt in young adulthood. P erson ality
Rosenberg, M . (1965). S o c iety a n d the a d o le s c e n t self- a n d S o c ia l P sy ch o lo g y B u lletin , 18, 1 9 9 - 2 0 6 .
im ag e. Princeton, NJ: Princeton University Press. Tangney, J. P. (1995). Shame and guilt in interperson
Rusch, N., Corrigan, P. W., Bohus, M ., Kuhler, T., al relationships. In J. P. Tangney 8c K. W. Fischer
Jaco b, G. A., 8c I.ieb, K. (20 07 ). The impact of post- (Eds.), S elf-c o n sc io u s e m o tio n s : T h e p s y c h o lo g y o f
traumatic stress disorder on dysfunctional implicit s h a m e, g u ilt, e m b a r ra s sm e n t, a n d p r id e (pp. 1 14—
and explicit emotions among women with border 139). New York: Guilford Press.
line personality disorder. J o u r n a l o f N erv o u s a n d Tangney, J. P. (1996). Conceptual and methodological
M e n ta l D ise a se , 195, 5 3 7 - 5 3 9 . issues in the assessment of shame and guilt. B e h a v
Russell, J. A., 8c Carroll, J . M . (1999). On the bipo io u r R e se a rc h a n d T h era p y , 3 4 , 7 4 1 - 7 5 4 .
larity of positive and negative affect. P sy c h o lo g ic a l Tangney, J . P., 8c Dearing, R. (2002). S h a m e a n d gu ilt.
B u lletin , 1 2 5 , 3 - 3 0 . New York: Guilford Press.
Sabini, J ., 8c Silver, M. (1997). In defense of shame: Tangney, J . P., Marschall, D. E., Rosenberg, K., Bar-
Shame in the context of guilt and embarrassment. low, D. H., 8c Wagner, P. E. (1994). C h ild r en ’s a n d
J o u r n a l f o r th e T h e o r y o f S o c ia l B eh av iou r, 21, a d u lts a u t o b io g r a p h ic a l a c c o u n ts o f s h a m e, gu ilt
1 -15 . a n d p r id e e x p e r ie n c e s : An an aly sis o f situ a tio n a l
Scheff, T. J . (1987). The shame-rage spiral: A case d e te rm in a n ts a n d in te r p e r s o n a l c o n c e rn s. Unpub
study of an interminable quarrel. In H. B. Lewis lished manuscript.
(Ed.), T h e r o le o f s h a m e in s y m p to m fo r m a tio n Tangney, J . P., Mashek , D., 8c Stuewig, J. ( 2 00 7 ). Work
(pp. 1 0 9 - 1 4 9 ). Hillsdale, NJ: Erlbaum. ing at the social-cl inica l-com munity -crim in ology
Serbin, L. A., 8c Stack, D. M . (1998). Introduction interface: The G M U inmate study .J o u r n a l o f S o c ia l
to the special section: Studying inter-generational a n d C lin ica l P sy ch olog y , 2 6 , 1 -21 .
continuity and the transfer of risk. D e v e lo p m e n ta l Tangney, J. P., Miller, R. S., Flicker, L., 8c Barlow, D.
P sy ch olog y , 3 4 , 115 9-11 61 . H. (1996). Are shame, guilt and embarrassment dis
Sherry, A. ( 2 0 07 ). Internalized homophobia and adult tinct emotions?: An analysis of participant ratings.
attachment: Implications for clinical practice. P sy J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 70,
c h o th e r a p y : T h eo r y , R esea rch , P ractice, T raining, 1256-1269.
44, 219-225. Tangney, J. P., Stuewig, J ., Kendall, S., Reinsmith, C.,
Smith, R. H., Webster, J . M ., Parrot, W. G., 8c Eyre, H. 8c Dearing, R. ( 2 0 06 ). Im p lic a tio n s o f c h ild h o o d
L. (2002). The role of public exposure in moral and s h a m e a n d g u ilt f o r risky a n d illeg a l b e h a v io r s in
nonmoral shame and guilt. J o u r n a l o f P erso n a lity y o u n g a d u lt h o o d . Unpublished manuscript.
a n d S o c ia l P sy c h o lo g y , 8 3, 1 3 8 - 1 5 9 . Tangney, J. P., Stuewig, J., 8c Mashek, D. J . (2007a).
Stegge, H., 8c Ferguson, T. J . (1990). C h ild - C h ild A t Moral emotions and moral behavior. A n n u al R e
trib u tio n a n d R e a c tio n S u rv ey (C -C A R S ). Unpub v iew o f P sy ch olog y , 5 8 , 3 4 5 - 3 7 2 .
lished manuscript, Utah State University. Tangney, J . P., Stuewig, J ., 8c Mashek, D. J. (2007b) .
Stuewig, J., Kendall, S., Sc Tangney, J. ( 2 0 0 4 , April). W h a t ’s moral about the self-conscious emotions? In
In te rg e n e ra tio n a l tran sm ission o f m o r a l e m o tio n a l J . L. Tracy, R. W. Robins, 8c J. P. Tangney (Eds.),
sty le: H o w v a lid is th e m yth o f th e g u ilt-in d u cin g T h e se lf-c o n s c io u s e m o tio n s : T h e o r y a n d research
m o t h e r ? Poster presented at the Conference on (pp. 2 1 - 3 7 ) . New York: Guilford Press.
Human Development, Washington, DC. Tangney, J . P., Wagner, P., 8c Gramzow, R. (1989). T h e
Stuewig, J., 8c McCloskey, L. (2005). The impact of test o f se lf-c o n s c io u s a ffe c t . Unpublished manu
maltreatment on adolescent shame and guilt: Psy script, George Mason University.
chological routes to depression and delinquency. Tangney, J. P., Wagner, P. E., Fletcher, C., 8c G ram
C h ild M a ltrea tm en t, 10, 3 2 4 - 3 3 6 . zow, R. (1992). Shamed into anger?: The relation of
Stuewig, J ., 8c Tangney, J. P. (2007). Shame and guilt in shame and guilt to anger and self-reported aggres
antisocial and risky behaviors. In J. L. Tracy, R. W. sion. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y ,
Robins, 8c J. P. Tangney (Eds.), T h e se lf-c o n s c io u s 62, 669-675.
e m o tio n s : T h e o r y a n d resea rch (pp. 3 7 1 - 3 8 8 ) . New Tangney, J. P., Wagner, P. E., Hill-Barlow, D. H.,
York: Guilford Press. Marschall, D., 8c Gramzow, R. (1996). The relation
Stuewig, J . , Tangney, J . P., Heigel, C., 8c Harty, L. of shame and guilt to constructive vs. destructive
( 2 0 0 6 , August). T h e m o r a l e m o tio n s , e x t e r n a liz a responses to anger across the lifespan. J o u r n a l o f
tion o f b la m e , a n d a g g ression . Paper presented at P erson ality a n d S o c ia l P sy ch olog y , 70, 7 9 7 - 8 0 9 .
the meeting of the American Psychological Associa Tibbetts, S. G. (2003). Self-conscious emotions and
tion, New Orleans, LA. criminal offending. P sy c h o lo g ic a l R e p o rts , 93,
Stuewig, J ., Tangney, J. P., Mashek, D., Forkner, P., 101-126.
8c Dearing, R. L. (in press). The moral emotions, Tracy, J . L., 8c Robins, R. W. (20 04 ). Putting the self
alcohol dependence, and H IV risk behavior in an into self-conscious emotions: A theoretical model.
incarcerated sample. S u b sta n ce Use a n d M isuse. P sy c h o lo g ic a l In qu iry, 15(2), 1 0 3 - 1 2 5 .
13. S ham e and Guilt 209
Tracy, J . L.., & Robins, R. W. (200 6 ). Appraisal ante Wicker, F. W., Payne, G. C., & Morgan, R. D. (1983).
cedents of shame and guilt: Support for a theoretical Participant descriptions of guilt and shame. M o tiv a
model. P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , tion a n d E m o tio n , 7, 2 5 - 3 9 .
32 (10), 13 3 9 -1 3 5 1 . Williamson, I., Sandage, S. J ., &c Lee, R. M. (200 7 ).
Wallbott, H. G „ & Scherer, K. R. (1995). Cultural de How social connectedness affects guilt and shame:
terminants in experiencing shame and guilt. In J. Mediation by hope and differentiation of self. P er
P. Tangney & K. W. Fischer (Eds.), S elf-c o n sc io u s so n a lity a n d In d iv id u a l D iffe r e n c e s , 4 3 , 2 1 5 9 -
e m o tio n s : T h e p s y c h o lo g y o f s h a m e, g u ilt, e m b a r 2170.
ra ssm en t, a n d p r id e (pp. 4 6 5 - 4 8 7 ) . New York: Zahn-Waxler, C., Kochanska, G., Krupnick, J ., &
Guilford Press. Mayfield, A. (1988). C o d in g m a n u a l fo r children 's
Webb, M., Heisler, D., Call, S., Chickering, S. A., & in te rp r eta tio n s o f in te r p e r s o n a l d istress a n d c o n
Colburn, T. A. (20 07 ). Shame, guilt, symptoms of flict. Bethesda, M D : National Institutes of Mental
depression, and reported history of psychological Health, Laboratory of Developmental Psychology.
maltreatment. C h ild A b u se a n d N eg lect, 3 1, 1143— Zahn-Waxler, C., & Robinson, J. (1995). Empathy and
1153. guilt: F^arly origins of feelings of responsibility. In J.
Weitzman, O., Kemeny, M. E., & Fahey, J. L. (2 004). P. Tangney & K. W. Fischer (FMs.), S elf-c o n sc io u s
H IV -r e la te d sh a m e a n d g u ilt p r e d ic t C D 4 d eclin e. e m o tio n s : S h am e, gu ilt, e m b a r ra s sm e n t, a n d p r id e
Manuscript submitted for publication. (pp. 143-1 73 ). New York: Guilford Press.
CHAPTER 14
J ohn C. B a refo o t
St e p h e n H . B o y l e
210
14. H ostility and P ron en ess to A n g er 211
Various theoretical traditions have at of high dominance and low friendliness,
tempted to place the hostility components whereas withdrawal would be characterized
in their broader view of personality. Others as low dominance and low friendliness.
have emphasized particular components and The inclusion of the dominance dimension
distinctions. sets the Circumplex model somewhat apart
from other approaches to the hostility com
plex in that many of the behaviors that com
Trait Taxonom ies
pose it (e.g., desire to influence or gain status)
The factor-analytic approach to describing need not be motivated by feelings of ill will
the interrelationships of personality traits or anger. Therefore, this broadens the con
has an extensive history in personality psy cept, and the fact that studies of dominance
chology (e.g., Cattell, 1946; Eysenck & Ey (e.g., Houston, Babyak, Chesney, Black, &
senck, 1985). The five-factor model (FFM ; Ragland, 1997; Siegman, Kubzansky, et
McCrae & Costa, 1985) has been the most al., 2 0 0 0 ; Smith et al., 2004) have observed
prominent trait taxonomy in hostility re health effects that are similar to those seen in
search. The FFM divides personality into hostility studies supports this extension.
the domains of Extraversion, Neuroticism,
Openness, Agreeableness, and Conscien
Com ponent Model
tiousness, each of which has subscales, or
facets, in the system of Costa and McCrae The widely accepted division of experience
(1992). Hostility is most closely related to into cognition, affect, and behavior under
facets of the Neuroticism and Agreeable lies many of the conceptualizations of hostil
ness domains. The Neuroticism dimension, ity phenomena and guides the presentation
which emphasizes negative emotional expe in this chapter. The cognitive component of
riences, contains the Angry Hostility facet, hostility is generally thought to be primarily
and the Agreeableness domain emphasizes composed of negative beliefs about the na
altruistic and friendly tendencies versus pre ture of other people (cynicism) and suspicion
dispositions toward egocentrism and argu regarding their intent (mistrust). This should
mentativeness. lead to a vigilant and protective tone in one’s
interactions. Research on the affective com
ponent often focuses on angry feelings, but
C ircum plex M odel
there are other important related emotions,
Smith, Glazer, Ruiz, and Gallo (2004) have such as disgust and resentment, that are often
outlined the applicability of the Circumplex neglected. Likewise, the hallmark of the be
model (Wiggins & Trapnell, 1996) for the havioral component is verbal and physical
study and description of hostile personality aggression, but these can be suppressed or
characteristics as they relate to interpersonal expressed indirectly, a phenomenon that is
phenomena. This model contains two or sometimes underemphasized.
thogonal dimensions that describe a person’s The three-component model highlights
approach to interpersonal relationships. some important distinctions. If hostility is
One dimension is based on affective tone treated as a unitary construct it overlooks
and ranges from antagonism and coldness to the fact that the cognitive, affective, and be
friendliness and warmth. The other dimen havioral components are differentially asso
sion is based on power or control and ranges ciated with other variables, such as age and
from dominance to submissiveness. These socioeconomic position (Barefoot, Beckham,
dimensions also describe interpersonal moti Haney, Siegler, & Lipkus, 1993; Haukkala,
vations of communion (desire for affiliation 2 0 02), and may behave somewhat indepen
and intimacy) and agency (desire for achieve dently depending on the setting. Further
ment and status). Anger is certainly related more, studies of the associations between
to the friendliness dimension, and ambitious multiple hostility measures have found three
behavior defines one extreme of dominance factors that correspond to the dimensions of
orientations. Particular characteristics can the component model (Barefoot et al., 1993;
be located in the two-dimensional space M artin, Watson, & Wan, 2000). Much of
formed by these orthogonal dimensions. For the literature and much of this chapter treat
example, aggression can be seen as a blend the three components as part of the same
212 III. E M O T IO N A L D IS P O S IT IO N S
complex without stressing their potential physiology. For example, openly expressing
differential effects. That point should be one’s anger in an aggressive manner often
kept in mind when reading the literature and elicits reciprocal aggressive behavior and
conducting future research. acrimony, escalating conflicts, impeding
successful problem resolution, and eroding
important relationships. It is also associated
A n g er Regulation
with elevated indicators of cardiovascular re
Many of the social and personal consequences activity that have important implications for
of the experience of anger depend on the way health (Siegman, 1994). On the other hand,
that a person copes with it. Research in this researchers have also noted deleterious so
area has identified multiple anger-regulation cial and physiological effects of the failure to
strategies, including rumination, avoidance, express emotion. Accurately communicating
assertion, reflective coping, discussion, ag emotional states is important to interperson
gression, suppression, and various forms of al functioning, and suppression of emotion
repression (Deffenbacher, Oetting, Lynch, can impair communication, making it more
& Morris, 1996; Garssen, 20 0 7 ; Linden et difficult to form and maintain close relation
al., 2003). These strategies can be thought ships (Butler et al., 2 0 0 3 ; Gross, 2002). In
of as attempts to regulate anger at differ addition, some forms of covert coping with
ent points in the emotion-generative process anger are associated with delayed physiolog
through situation selection (e.g., avoidance), ical recovery after anger-producing episodes
modification of the situation (e.g., problem- (Brosschot & Thayer, 1998; Flogan & Lin
focused coping), deployment of attention den, 2004).
(e.g., rumination), change in cognitions (re As these findings suggest, research on
flective coping), and selection of behavioral anger-regulation styles has resulted in a pic
responses (e.g., expression and suppression) ture clouded by findings that appear to be
(John &c Gross, 2004). contradictory on the surface and by some
The focus of much of this work has been what conflicting theoretical approaches.
the congruity between negative emotional This state of affairs calls for theoretical ap
experiences of animosity and the overt be proaches that better recognize the complexi
havioral expression of those feelings. The ties of potential anger-regulation styles and
most widely used approach has contrasted their appropriateness for particular social
Anger-Out, involving overt behavior that settings.
displays the negative feelings in a verbally or
physically aggressive manner, and Anger-In,
involving efforts to conceal negative feelings Questionnaire Measures
and to avoid direct confrontation (Speilberg-
er et al., 1985). This dichotomy was origi One approach to hostility measurement has
nally thought to represent extremes of one been based on dimensions from omnibus
dimension, but it is clear that this is an over personality inventories. The most prominent
simplification. Measures of these tendencies example is the use of the N EO Personal
load on separate dimensions: Neuroticism ity Inventory (NEO PI; Costa & M cCrae,
and Agreeableness in the FFM (M artin et al., 1992), which operationalizes the FFM . It is
2000 ) as well as the affective and behavioral available in forms that allow for both self-
factors of the component model of hostility reports and peer reports. Congruence be
(Barefoot et al., 1993). In addition, recent tween self-reports and peer reports has been
work has identified other anger response reported to be good and supportive of the
styles such as those mentioned earlier, as division of traits into these dimensions (M c
well as multiple modes of overt expression Crae & Costa, 1987).
(e.g., anger discussion, assertion, and recip Another tactic is the derivation of scales
rocal communication) that reflect a desire to specifically designed to evaluate hostile ten
communicate one’s feelings of anger in non- dencies. These measures can be divided ac
aggressive ways. cording to the three hostility components.
Anger-regulation styles can have impli A large number of instruments have been
cations for social relations and one’s own devised to assess aspects of hostility. There
14. H o stility and P ron en ess to A n g er 213
fore, we do not attempt to be comprehensive The Rotter Interpersonal Trust Scale (Rotter,
in the following descriptions, which concen 1967) consists of 19 items designed to reflect
trate on the most frequently used scales that the belief that one can rely on the promises
are relevant to the health psychology litera and intent of others. It correlates highly with
ture. the Ho scale (Barefoot et al., 1993) and has
been used in a variety of studies to explore
the implications of trust for interpersonal re
Cognitive Aspects
lationships (e.g., Rotter, 1980).
The most widely used measure of hostility Another trust measure is Factor L from
in the health psychology literature is the the Sixteen Personality Factor (16PF) Per
Cook-M edley Hostility Scale (Ho; Cook sonality Inventory, which is derived from
& Medley, 1954). The popularity of this Cattell’s (1946) theory of personality, an
measure is certainly not based on its psy early product of the use of factor analysis
chometric properties, which have repeatedly to derive personality dimensions from trait
been found to be mediocre (Barefoot & Lip- adjectives. Factor L is described as a mea
kus, 1994; Contrada & Jussim, 1992) even sure of vigilance ranging from the extremes
though its reported test-retest reliabilities of suspicious to trusting. The 16PF has been
and internal consistency have been quite high widely used, and its inclusion in important
(Barefoot, Dodge, Peterson, Dahlstrom, & longitudinal studies has permitted the ex
W illiams, 1989; Shekelle, Gale, Ostfeld, & ploration of multiple consequences of suspi
Paul, 1983). In fact, the instrument was not cious tendencies compared to trusting ones.
even originally devised to be a measure of
hostility per se but was empirically derived
A ffective Aspects
to identify items that discriminated between
teachers who had good or bad rapport with A variety of affects may be associated with
their students. The resulting items appeared hostility, but most of the self-report instru
to reflect a hostile interpersonal orienta ments have focused on anger. The most
tion, so the scale was classified as a hostility prominent measure has been from the
measure. The frequent use of the Ho scale Spielberger State-Trait Personality Inven
stems from its ability to predict important tory (Spielberger, Jacobs, Russell, & Crane,
outcomes, including interpersonal stress 1983). The trait component of the scale
and conflict (Smith, Pope, Sanders, Allred, asks respondents to rate the frequency with
& O ’Keefe, 1998), mental well-being (Mao, which they experience anger and annoyance.
Bardwell, M ajor, & Dimsdale, 2003), and The State Anger scale asks similar questions
health outcomes such as coronary heart dis about the respondent’s feelings at the mo
ease and mortality (Smith et al., 2004). ment.
The original Ho scale contains 50 items. The issue of the direction of anger expres
However, a rational analysis identified 11 sion is most frequently assessed via the Spiel
items that did not appear to reflect hostil berger Anger Expression Scale (Spielberger
ity and divided the remaining ones into four et al., 1985). It directly asks the respondents
subsets reflecting cognition (cynicism and to report their tendencies to direct anger in
hostile attributions), affect (hostile affect), ward and suppress it or express it outwardly
and behavioral tendencies (aggressive re toward others by arguing or other overt ac
sponding) (Barefoot et al., 1989). The bulk tions.
of the scale represents cynicism and mis
trust (Smith & Frohm, 1985). The full 50
A ggression
Ho items are often administered, but several
briefer versions have been used to predict The Buss-Durkee Hostility Inventory (Buss
health outcomes (e.g., Boyle et al., 20 0 5 ; & Durkee, 1957) has traditionally been one
Julkunen, Salonen, Kaplan, Chesney, & Sa of the most widely used measures of aggres
lonen, 1994; Surwit et al., 2002). sive tendencies. It contains seven subscales
Two other measures focusing on trust that tend to fall into two dimensions (Bush
should be noted, although they are not used man, Cooper, &c Lemke, 1991). One repre
as widely as the Ho scale in health research. sents components of anger-related affect that
214 III. E M O T IO N A L D IS P O S IT IO N S
correlate with Neuroticism in the FFM , and ings in heart patients and have conceptual
it is most strongly related to the Resentment ized this as a form of denial. This work has
and Suspicion subscales. The other reflects shown that men are more likely than women
aggressive tendencies and is composed of the to underestimate their own hostility. In addi
Assault and Verbal Aggression subscales. tion, there are sizable individual differences
Buss and Perry (1992) noted that the fac in self-awareness, which should also lower
tor structure of the Buss-Durkee measure the congruence between overt behavior and
was somewhat inconsistent across studies self-reports in some people.
and that some of the item interpretations One class of behavioral measures employs
were ambiguous. Therefore they developed coding of interactions during standardized
a new version, the Aggression Question interviews or laboratory tasks. The impe
naire. It contains four subscales: Anger and tus for the development of these methods
Hostility, which correlate with the Neuroti came from the Western Collaborative Group
cism dimension of the FFM , in addition to Study (W CGS; Rosenman, Swan, 8c Car-
Physical Aggression and Verbal Aggression, milli, 1988), which administered structured
which correlate with Agreeableness (Gallo interviews dealing with daily habits and
& Smith, 1998). coded responses based on vocal characteris
Another measure that combines assess tics (e.g., speech rate, loudness) and interac
ment of anger arousal and expression style tion style, as well as content. This procedure
is the Multidimensional Anger Inventory was used to categorize participants on the
(Siegel, 1986). It differentiates between five basis of their Type A behavior, a predictor of
dimensions: ease of anger arousal; range of later coronary disease in this sample. Later
anger-eliciting situations; hostile outlook; scoring schemes separately scored the hostil
anger-in, which has items indicative of ru ity reflected in the respondent’s behavior and
mination; and anger-out. made the scoring procedures more explicit.
The most recent system, the Interpersonal
Hostility Assessment Technique (IHAT;
Measures Based Haney et al., 1996), extends the method de
on Behavioral Observations vised by Chesney, Hecker, and Black (1989),
concentrating on four categories of hostility-
In addition to standard self-report measures, related behaviors. Direct Challenges involve
there is a tradition in this line of research open verbal aggression directed toward the
to base assessments purely on observations interviewer. A more common behavior is an
of the target person’s overt behaviors. There Indirect Challenge, which is a more subtle
are reasons to believe that this strategy will implication of hostility that is judged from
capture information that is not obtained voice stylistics. For example, the statement
from more traditional questionnaire meth “of course” can be said in an agreeable fash
ods (Barefoot &c Lipkus, 1994). One obvi ion or in a tone implying that the answer is
ous barrier to effective self-report measures obvious and the question is stupid. The cat
is the possibility that respondents, especially egory of Irritation is based on indicators of
those who are not trusting, may be reluc negative emotional arousal as judged from
tant to openly admit their hostile tendencies vocal stylistics. The final category, Hostile
given the socially undesirable status of those Withhold-Evade, is coded when the respon
traits. Another is that many of those who dent fails to answer a question and does
have a hostile or combative interaction style so in an antagonistic manner. The sum of
appear to be unaware of this aspect of be scores from these four categories has been
havior, and therefore they may not describe shown to correlate with nonverbal expres
themselves as hostile on a questionnaire. It is sions of negative affect (Brummett et al.,
as if they see their behavior to be a justifiable 1998). More important, it has been shown
reaction to a provoking interaction partner to be associated with the extent of coro
or situation rather than a product of their nary disease in patient samples (Haney et
own personalities. Some investigators (e.g., al., 1996; Siegman, Townsend, Civelek, &
Ketterer et al., 1998) have noticed frequent Blumenthal, 2 0 0 0 ) and to predict coronary
discrepancies between self- and spouse rat mortality in a population sample of men at
14. H ostility and P ron en ess to A n g er 215
high risk (Matthews, Gump, Harris, Haney, M arsh, & Moeller, 1999). Consistent with
& Barefoot, 2004). this, low levels of serotonin function have
Interview-based measures have also been been linked to measures of aggressive behav
devised for the direct study of dominance, ior and impulsivity in clinical and normal
a primary component of the Circumplex samples (Carver & Miller, 2 0 0 6 ; Manuck,
model and a behavioral dimension shown to Flory, Muldoon, & Ferrell, 2002). M ost evi
be important in laboratory studies of cardio dence regarding serotonin shows effects on
vascular reactivity (Smith et al., 2 0 0 4 ; see the measures of overt behavior affected by im
section on social interactions). Houston and pulse control, with less support for links to
colleagues (1997) performed cluster analy more subtle manifestations of hostility seen
ses of speech characteristics observed during in normal interactions.
structured interviews from the W CGS. Peo The role of testosterone in facilitating ag
ple whose vocal styles could be character gressive behavior has also received a great
ized by attempted dominance (e.g., verbally deal of attention (Archer, 2006). Positive as
competitive, with fast speaking rates and sociations between levels of testosterone and
immediate responses) had higher mortality trait aggression have been reported in many
rates over the ensuing 2 2 years of follow-up. studies, and these associations appear to be
Siegman, Townsend, and colleagues (2000) strongest for measures of aggressive domi
found that a rating of dominance behavior nance. Testosterone levels are also associat
was associated with coronary disease in pa ed with higher levels of aggressive behavior
tients undergoing thallium scan testing inde in response to tasks involving competition
pendent of their IHAT scores. or interpersonal challenge. There is also
Another behavioral observation measure evidence that testosterone influences how
ment strategy is to utilize judgments made people assess physical and psychological
by the target person’s peers and family. The threats to their status, a potential mediator
ratings are obviously less standardized and for those associations. Although the focus of
structured than interview or laboratory- much of this research has been men, there is
based systems and are subject to some biases evidence that similar associations are pres
(Barefoot & Lipkus, 1994), but they can ent in women.
capitalize on the extensive experience the in There have been multiple investigations
formant has had with the target in a variety and frequent speculation regarding the pos
of settings. In health research, spouse ratings sible genetic origins of hostility and anger.
of hostility have been better than self-ratings The plausibility of this line of inquiry is
in identifying those with coronary disease bolstered by family studies showing that ge
(Ketterer et al., 2 0 0 4 ; Kneip et al., 2 0 0 4 ; netic influences are likely to account for a
Smith et al., 2 0 0 7 ), especially in men. substantial amount of the variance in hos
tility scores (Weidner et al., 20 0 0 ). Searches
for specific genes associated with hostile ten
Origins o f Hostile Tendencies dencies have been conducted, but the find
ings are complex. As with the studies of ser
Biological U nderpinnings
otonin manipulation, the clearest findings
There has been a considerable amount of have come from studies dealing with genetic
work on possible physiological bases of hos influences on overt aggression and violent
tility, anger, and, particularly, aggression. A behavior (Manuck et al., 2002). The strat
good deal of this effort has focused on the egy of searching for meaningful interactions
processes that influence the levels of the neu between genes and environmental factors
rotransmitter serotonin in the brain (Carver promises to be the most promising course to
& Miller, 20 0 6 ). Experimental manipula pursue in the future (M offitt, Caspi, & Rut
tions of central nervous system serotonin ter, 2005).
have demonstrated decreases in aggressive
tendencies with serotonin enhancement.
D evelopm ental H istory
Serotonin depletion leads to increases in ag
gression, especially in those with preexist Learning based on childhood environment
ing aggressive tendencies (Dougherty, Bjork, plays a significant role in the acquisition of
216 III. E M O T IO N A L D IS P O S IT IO N S
hostile tendencies. Several studies have em (2006) report that supportive parenting
phasized family experiences as a source of successfully reduces hostile attribution and
later hostile tendencies in offspring. An im aggressive tendencies in African American
portant perspective on this literature comes adolescents who are subject to discrimina
from the social-cognitive approach of Dodge tion. This appears to reduce their likelihood
(2006). The initial tenet is that aggression of delinquency and similar problems.
is a universal and natural response to per
ceived threat. One task of socialization is
Trauma
the development of benign attributional
schemas to counteract the tendency to make Aside from childhood stressors that can af
hostile attributions to explain the actions of fect developmental processes, trauma expe
others. These schemas are fostered in chil rienced in adulthood can also result in high
dren by secure attachments with a caregiver, levels of mistrust and other aspects of hostil
modeling of benign attribution habits from ity. It is one of the hallmark symptoms of
others, success experiences, and living in posttraumatic stress disorder (PTSD) result
a culture that values cooperation. Without ing from many types of severe stress (Orth
these experiences, it is more likely that the & Wieland, 20 0 6 ), although the association
child will develop a defensive attitude that is between anger and PTSD is somewhat larger
the precursor to aggressiveness and problem in those exposed to military stressors. Even
behavior. combat veterans who only have subclinical
These principles can be seen in the vo levels of PTSD symptoms show evidence of
luminous literature on childhood environ heightened anger and hostility (Jakupcak et
ments and hostile tendencies. For example, al., 2007). Women with PTSD also report
children who have been abused at home or high levels of hostility (Beckham, Calhoun,
have experienced other significant stressors Glenn, & Barefoot, 2002). In the general
are more likely to have hostile attributional population, PTSD is more prevalent in
tendencies toward others, including peers women than in men and is most frequently
(Price & Glad, 2 0 0 3 ; Turner, Russell, Glov associated with a history of sexual assault
er, & Hutto, 200 7 ). However, much of the (Kessler, Sonnega, Bromet, Hughes, & Nel
literature on family environment and hostil son, 1995).
ity development uses retrospective and/or
subjective measures of family environment,
which carry a large potential for bias. M eth D em ographic Distributions
ods that circumvent those problems have o f Hostility
also been employed. For example, M at
thews, Woodall, Kenyon, and Jacob (1996) Differences in hostility between various de
coded the interactions between fathers and mographic groups reflect many of the influ
their adolescent sons while they discussed ences discussed earlier. M ost of the data for
a disagreement in a laboratory setting. The these comparisons are based on the Ho scale
observed levels of hostile behavior on the because it is the most widely administered
part of both fathers and sons predicted the measure in community samples.
sons’ scores on multiple hostility measures
3 years later after controlling for initial lev
G end er
els.
Dodge’s (2006) approach also predicts One of the strongest consistent demographic
that supportive family environments can be differences is the tendency for women to have
protective against the effects of trauma in lower scores on measures of cynicism and
children. Luecken (2000) compared univer aggression. For example, they have lower Ho
sity students who had experienced the loss scores in every age group of a U.S. national
of a parent with students from intact fam i sample (Barefoot et al., 1991). This pattern is
lies. Those who had experienced loss report seen as an integral part of the female gender
ed more hostility and depression if they had role, consistent with its emphasis on commu
poor family relationships. No difference be nion and positive interpersonal relationships
tween loss groups was present for those with (Helgeson, 1994). Therefore, social learning
supportive families. Simons and colleagues and developmental experiences should influ
14. H o s t ilit y and P ro n e n e s s to A n g e r 217
ence these tendencies. However, women do potent predictor of hostility scores (Harper
not consistently have lower scores on mea et al., 2002), perhaps reflecting the impor
sures of anger or anger expression, indicat tance of that life stage for personality devel
ing that they acknowledge anger experiences opment.
but tend to employ less aggressive modes of Most of the studies cited previously have
expression (Stoney & Engebretson, 1994). used versions of the Ho scale. Although this
establishes an inverse association between
socioeconomic position and cognitive as
A ge
pects of hostility, other associations have
A second common trend is a curvilinear as been found with measures of anger expres
sociation between hostility and age (Bare sion. Lower status groups actually report
foot et al., 1991; Swenson, Pearson, & O s lower scores on the Anger-Out measure,
bourne, 1973). There are large reductions in indicating a reluctance to express these
scores between adolescence and adulthood, feelings overtly (Haukkala, 2002). This is
with levels rising slightly among those above hypothesized to indicate that a position of
60. The elevation in older respondents was power or prestige provides the conditions for
investigated more thoroughly in a sample acceptable anger expression.
of 125 middle-aged and older community
volunteers using multiple hostility measures
Culture
(Barefoot et al., 1993). The positive associa
tion with age was seen most clearly in the Culture influences hostile tendencies through
cognitive measures reflecting cynicism and shared childrearing practices, modeling, and
mistrust, although it was also present in cultural experiences. Thus differences in
the interview-based behavioral assessment. hostile tendencies are not confined to the
In contrast, self-report measures of aggres level of the individual. For example, there are
siveness showed a weak inverse relationship notable differences in emotional respond
with age, and there was no significant effect ing between Western European-based and
for self-reports of more covert experiential Asian-based cultures. In American samples,
aspects of hostility. suppression of emotions can have a number
of negative social outcomes, such as weaker
social attachments and relationship close
Ethnicity
ness (Gross, 2002). In contrast, emotional
There are clear differences in Ho scores expression is more often suppressed in Asian
across ethnic groups, with minorities having cultures, and these negative consequences are
higher scores (Barefoot et al., 1991). This not seen in interactions among individuals
difference is consistently seen across stud sharing Asian values (Butler, Lee, & Gross,
ies and may be due to an environment filled 2 0 07). Consequences of cultural differences
with discrimination and economic challeng in expressiveness can be seen in studies from
es that hinder the development of benign at multicultural societies that compare anger-
tributional styles. Hostile attributions can related physiological reactions of individuals
be seen as a potentially adaptive response varying in ethnicity. For example, both lab
to a threatening and objectively hostile en oratory and ambulatory monitoring studies
vironment. of Singapore residents have found ethnic
variations in blood pressure responses dur
ing daily life and in response to laboratory
Socioeconomic Position
stressors (Bishop & Robinson, 2 0 0 0 ; Enkle-
Multiple studies (e.g., Barefoot et al., 1991; man et al., 2005). Although there have been
Scherwitz, Perkins, Chesney, & Hughes, some variations across studies, the general
1991; Haukkala, 2002) have demonstrated finding is that hostile ethnic Indians have
inverse associations of hostility with income more cardiovascular reactivity to stressors
and education. As with ethnicity, it could be than do Chinese and Malay study partici
argued that cynicism is an understandable pants who are otherwise similar. These find
response to the harsher living conditions ex ings are consistent with the higher rates of
perienced by those groups. Childhood socio cardiovascular disease found in the Indian
economic status appears to be an especially subpopulation.
218 III. E M O T IO N A L D IS P O S IT IO N S
ing social interactions (Smith et al., 2004). of two members with low hostility were not
This not only shows the level of emotional argumentative regardless of the discussion
arousal, but it also makes up part of the ex task. O f most interest was the behavior of
planation for the impaired cardiovascular the couples with one member high and one
health of hostile people (see the following low in hostility. Their behaviors resembled
section). those of the couples composed of two people
A related consequence of this perceptual high in hostility, showing an increase in an
bias is that hostile individuals will tend to tagonistic behavior in response to the high-
see others as less supportive. Indeed, they conflict topic and maintaining it during the
tend to report lower levels of social support final discussion. Therefore the member of
(Benotsch, Christensen, & McKelvey, 1997) the mixed pair with high hostility evidently
and report more social negativity in their elicited conflict behavior from the normally
work environments (M cCann, Russo, Benja nonaggressive partner. Thus the transac
min, &c Andrew, 1997). Even when available, tional theory illustrates a process in which
social support may have less psychological people’s own hostility not only affects their
and physiological impact as well, somewhat interpretations of their social experiences
negating its normally beneficial effects. Lep- but also initiates a scenario of potential re
ore (1995) had participants who varied on ciprocal escalation that enhances the level of
cynicism perform a stressful speaking task stressful encounters.
with or without the presence of a supportive Components of this process have been
confederate. Those who were low in cynicism documented in questionnaire and laborato
showed more benefit from the support, both ry studies, but its consequences can be seen
in ratings of stress and blood pressure levels best by examining the naturally occurring
during the task. These effects on perceived social interactions in the daily lives of people
social support constitute another potential who vary in hostility. For example, Brissette
pathway to the mental and physical health and Cohen (2002) interviewed community
problems associated with high hostility. volunteers on 7 consecutive days about their
The transactional model identifies yet an social experiences and their sleep patterns
other important consequence of this process. during the previous 24 hours. Those high in
The ill intent that is perceived and that leads hostility reported more negative affect as
to anger or other negative affects can ulti sociated with conflict experiences, and their
mately lead to the activation of the behavior sleep was more disturbed on nights following
al component: aggression or related negative days with high levels of conflict. This may
actions. O f course, this can elicit reciprocal indicate a tendency for them to ruminate
negative responses from others, resulting in about those conflicts. It is also important
interactions that are more filled with ani because of the significance of adequate sleep
mosity. The end product is a social environ for both physical and mental well-being.
ment that is both objectively and subjective Even more detailed pictures of the effects
ly less supportive and more stressful. This of hostility on experience can be obtained
process was demonstrated in a laboratory with ambulatory monitoring of cardiovas
study (Smith, Sanders, & Alexander, 1990) cular reactions during the person’s regular
that paired married participants of vary daily activities accompanied by diaries that
ing hostility for a series of discussion tasks ask the participant to describe the nature of
(a low-conflict topic followed by a conflict- the activity that is going on at the time of
inducing topic and then another low conflict the physiological reading. Several studies of
topic) and coded the amount of negative hostility and ambulatory monitoring have
interpersonal behavior that occurred. The been conducted. Brondolo and colleagues
initial discussion did not elicit much conflict (2003) monitored 104 volunteers, taking
behavior, regardless of the composition of blood pressure and heart rate readings every
the couple. Couples composed of two hos 20 minutes during the day. Diaries indicated
tile people reacted to the high-conflict topic whether the person was engaging in social
with more frequent negative interpersonal interaction and his or her affective reactions
behaviors, and the negative interaction tone at the time of the reading. Those high in hos
was maintained into discussion of the subse tility reported fewer social interactions, sug
quent low-conflict topic. Couples composed gesting that they might have a tendency to
220 III. E M O T IO N A L D IS P O S IT IO N S
avoid others or that they are being avoided by studies (Houston & Kelly, 1989; Smith et
others. The interactions they did have were al., 1988). In contrast, Baron and colleagues
more likely to be negative, and the ratings (2007) examined hostility components mea
of negativity were more intense. Positively sured by the Buss-Perry scale to predict
toned interactions were rated as less intense marital adjustment trends over 18 months in
when they did occur. Some, but not all, of 122 couples. The cognitive and anger com
these trends were reflected in blood pressure ponents were both negatively correlated with
readings, with larger diastolic readings for adjustment in cross-sectional analyses, but
highly hostile participants during intense temporal trends were most closely related to
negative interactions. Findings from other anger, especially the wives’ anger.
ambulatory studies generally agree with The findings relating to the differential
these results (e.g., Benotsch et al., 1997), but roles of hostility in marital adjustment of
some found slight differences. For example, men and women are complex. This issue is
Jamner, Shapiro, Goldstein, and Flug (1991) potentially important for understanding the
studied the blood pressures of paramedics. nature of close relationships and for dealing
Those with high Ho scores had higher car with the mental and physical health con
diovascular activity levels during working sequences of marital problems. In general,
hours, especially in contexts that were likely marital relationships have been shown to
to involve interpersonal conflicts. This ef play a larger role in the physical health and
fect was strongest in those with a defensive psychological well-being of men (Kiecolt-
coping style, a tendency to deny or minimize Glaser & Newton, 2 0 0 1 ; Shumaker & Hill,
negative affect. Guyll and Contrada (1998) 1991). More detailed insight into the roles of
compared cardiovascular activity in partici hostility in close relationships can be found
pants varying in hostility during social and in the literature on physiological reactivity
nonsocial activities, finding a positive asso during marital interactions in laboratory
ciation during social encounters primarily in settings (Kiecolt-Glaser & Newton, 2001).
men. Results from studies using this paradigm are
Another way to illustrate the naturally not only useful for understanding emotional
occurring consequences of hostility in daily arousal in the marital context and its poten
life can be found in the literature on marital tial implications for health, but they have
relationships, which are obviously an im also been found to predict subsequent mari
portant component of a person’s well-being. tal dissolution or marital success. Discus
Trust is an essential component of success sions of topics that are the basis of problems
ful close relationships (Rempel, Holmes, in the marriage or experimentally induced
& Zanna, 1985), and the existence of fre disagreements result in elevated physio
quent anger and aggression in a relationship logical indicators of stress, effects that are
is likely to be especially detrimental. The present in samples of both newlyweds and
overall importance of hostility can be dem long-term married couples. M ost studies
onstrated in longitudinal studies of hostility find that these effects are more pronounced
scores in married couples and the course of in women. Hostility of the participants also
their marital satisfaction. Miller, Marksides, plays a role, and the nature of the effect ap
Chiriboga, and Ray (1995) found that the Ir pears to be dependent on the nature of the
ritability subscale of the Buss-Durkee Hos experimental task. When placed in a setting
tility Inventory, a measure of anger prone that emphasizes dominance or agency (e.g.,
ness, was predictive of marital dissolution in convincing the spouse to accept a position
a large sample of M exican Americans over on an issue), men high but not low in hos
an 11-year follow-up. A study of 53 newly tility show heightened physiological reactiv
wed couples followed for 3 years found that ity (Smith & Brown, 1991; Smith & Gallo,
men’s hostility, assessed with the Ho scale, 1999). Women’s hostility did not seem to
was associated with decreases in both their have an impact on their own reactivity, but
own and their wives’ marital satisfaction they did tend to have greater physiological
(Newton & Kiecolt-Glaser, 1995). No such reactions during the high-conflict discus
effect was found for the women’s hostility sions if they were married to men with high
scores. Support for this gender difference hostility scores. This picture is further com
has been obtained from other cross-sectional plicated by the recent suggestion that trait
14. H o s t ilit y and P ro n e n e s s to A n g e r 221
anger scores may be more predictive than to alterations in lipids, glucose metabolism,
hostility scores of women’s physiological re inflammation, and blood pressure (Golden,
sponses (Smith et al., 2004). One plausible 2007; Steptoe, Hamer, & Chida, 2007).
explanation is that men tend to react to chal Health behaviors constitute another po
lenges related to the dimension of agency, tential set of pathways (Bunde &c Suls, 2006).
whereas women respond to disturbances in For example, smoking has been associated
communion (Smith, Gallo, Goble, Ngu, & with hostility measures in several population
Stark, 1998). studies (e.g., Scherwitz et al., 1992; Shekelle
et al., 1983). Part of this association appears
to be due to the difficulties hostile people
H ostility and Physical Health face when trying to quit smoking (Lipkus,
Barefoot, Williams, & Siegler, 1994), per
The link between components of hostility haps because of the role that nicotine can
and health has been hypothesized for many play in affect regulation and the negative
decades, but interest in the topic was acceler affects aroused during cessation attempts.
ated by studies of the Type A behavior pat Jamner, Shapiro, and Jarvik (1999) found
tern as a predictor of heart disease (Rosen- that nicotine patches were especially effec
man et al., 1988) and the identification of tive in reducing anger among more hostile
hostility as its critical component (Williams people, whether or not they were smokers.
& Barefoot, 1988). Subsequently, a large Elevated alcohol consumption among
number of epidemiological, clinical, and highly hostile individuals has also been
laboratory studies have explored the health noted in multiple studies (e.g., Scherwitz
consequences of hostile components and et al., 1992; Shekelle et al., 1983). Boyle,
their mechanisms of action. Much of this Mortensen, Gronbaek, and Barefoot (2008)
work has focused on coronary heart disease observed a high prevalence of an unhealthy
(Smith et al., 2 0 0 4 ), but hostility measures pattern of heavy episodic drinking in those
have also been associated with other health with high Ho scores in addition to higher
outcomes, such as stroke, disability, and total alcohol intake. One possible contribut
total mortality (e.g., Adams, 1994; Kivima- ing factor is that alcohol consumption ap
ki, Vahtera, Koskenvuo, Uutela, & Pentti, pears to be a more effective stress reducer
1998; W illiams, Nieto, Sanford, Couper, in hostile drinkers (Zeichner, Giancola, &
& Tyroler, 2002). The Ho scale has been Allen, 1995).
most frequently used in this work, but other Another relevant lifestyle indicator is body
measures, such as anger indicators (e.g., mass index (BMI). Bunde and Suls (2006)
Williams et al., 2002) and interview ratings note a “fairly robust” positive relationship
(e.g., Matthews et al., 2 0 0 4 ), have also been between BM I and hostility across studies.
successful predictors. There are a number of Haukkala and Uutela (2000) found that the
plausible explanations that could account effect was stronger among women with low
for these phenomena. education. Related phenomena with even
Research cited earlier regarding transac more significant health implications are the
tional processes has shown that the percep associations of hostile characteristics with
tions and behaviors of hostile people tend central adiposity, insulin resistance, abnor
to create more stressful environments. This mal glucose, and lipid functioning, as well
produces higher cardiovascular reactiv as hypertension, although racial and gender
ity in both laboratory and natural settings, differences in the associations have been
resulting in elevated coronary disease risk. noted (Goldbacher & Matthews, 2007; Sur-
Other physiological systems that are re wit et al., 2002). These physiological indica
sponsive to stressors are also affected. For tors are important risk factors for cardiovas
example, high hostiles show enhanced ac cular disease.
tivation of the sympathetic nervous system
and hypothalam ic-pituitary-adrenal cortex
system in response to interpersonal stress Conclusions and Future D irections
(Suarez, Kuhn, Schanberg, W illiams, &
Zimmerman, 1998). The resulting hemody The research based on the transactional
namic and hormonal changes can contribute model demonstrates the reciprocal inter
222 III. E M O T IO N A L D IS P O S IT IO N S
changes between personality, social life, National Institute on Aging and Grant No.
and personal well-being. Social experiences P 0 1 H L 3 7 6 8 7 from the National H eart, Lung,
shape hostile predispositions, which, in turn, and Blood Institute.
influence the person’s social environment
through selection, perception, and their im
pact on the behavior of others. These social R eferen ces
experiences help determine a person’s expe Adams, S. H. (1994). The role of hostility in women’s
rience of stress or tranquility and, coupled health during midlife: A longitudinal study. H ea lth
with coping behaviors, his or her health and P sy ch olog y , 13, 4 8 8 - 4 9 5 .
psychological well-being. Allred, K. D., & Smith, T. W. (1991). Social cognition
in cynical hostility. C o g n itiv e T h era p y a n d R e
One of the most prominent trends in this
sea r ch , 15, 3 9 9 - 4 1 2 .
work has been the demonstration of the rele Archer, J . (200 6). Testosterone and human aggression:
vance of the hostility concept for topics such An evaluation of the challenge hypothesis. N e u r o
as marital relationships, social capital, and s c ie n c e a n d B io h e h a v io r a l R ev iew s , 3 0 , 3 1 9 - 4 1 2 .
physical health. These extensions into areas Barefoot, J . C., Beckham, J. C., Haney, T. L ., Siegler, I.
C., & Lipkus, I. M . (1993). Age differences in hos
with potential practical applications can
tility among middle-aged and older adults. P sy ch o l
continue to fuel the interests of researchers o g y a n d A gin g, 8, 3 - 9 .
in related fields that will provide new per Barefoot, J . C., Dodge, K. A., Peterson, B. L., Dahl-
spectives. Another direction of recent work strom, W. G., & Williams, R. B. (1989). T he C o o k -
Medley Hostility Scale: Item content and ability
that should be pursued is the growing em to predict survival. P s y c h o s o m a tic M ed icin e, 51,
phasis on more complex hypotheses evaluat 46-57.
ing interactions with demographic variables, Barefoot, J. C., & Lipkus, I. M . (1994). Assessment of
other personality characteristics, and situa anger-hostility. In A. W. Siegman & T. W. Smith
tions. The person-situation interaction issue (Eds.), A nger, h o stility a n d th e h e a r t (pp. 4 3 - 6 6 ) .
Hillsdale, NJ: Erlbaum.
is particularly noteworthy. Much of the liter Barefoot, J. C., Peterson, B. L., Dahlstrom, W. G.,
ature, including this chapter, has focused on Siegler, I. C., Anderson, N. B., & Williams, R. B.
the consistencies in hostile predispositions, (1991). Hostility patterns and health implications:
perhaps underestimating the significance Correlates of C o ok -M ed le y scores in a national sur
vey. H ea lth P sy ch olog y , 10, 1 8 - 2 4 .
of situational factors (e.g., Porter, Stone,
Baron, K. G., Smith, T. W., Butler, J . , Nealy-Moore, J.,
& Schwartz, 1999). Their influence should Hawkins, M. W., & Uchino, B. M. (200 7 ). Hostil
not be neglected, and a more complex inter ity, anger, and marital adjustment: Concurrent as
actional approach could help remedy that. sociations with psychosocial vulnerability. J o u r n a l
Such investigations can yield a better un o f B e h a v io r a l M ed icin e, 3 0 , 1 -1 0.
Beckham, J . C., Calhoun, P. S., Glenn, D. M ., & Bare
derstanding of the detailed impact of hostile foot, J. C. (2 002). Posttraumatic stress disorder,
predispositions and lead to more useful so hostility, and health in women: A review of cur
phisticated theoretical explanations. rent research. A n n als o f B eh a v io r a l M ed icin e, 2 4 ,
A caveat is in order. The research pre 219-228.
Benotsch, E. G., Christensen, A. J ., & McKelvey, L.
sented here and the literature in general tend
(1997). Hostility, social support, and ambulatory
to emphasize the negative consequences of cardiovascular activity. J o u r n a l o f B e h a v io r a l M e d
hostile predispositions. However, it should icin e, 2 0 , 1 6 3 - 1 7 6 .
be remembered that positive functions are Bishop, G. D., & Robinson, G. ( 2 00 0). Anger, harass
also served by hostility, at least in its milder ment, and cardiovascular reactivity among Chinese
and Indian men in Singapore. P s y c h o s o m a tic M ed i
forms. Oppositional behavior is often neces cin e, 6 2 , 6 8 4 - 6 9 2 .
sary for the benefit of society, and the ab Boyle, S. H., Mortensen, L., Gronbsek, M ., & Bare
sence of hostility and extreme or unmitigated foot, J. C. (2008). Hostility, drinking pattern, and
communion is also dysfunctional (Helgeson, mortality. A d d ic tio n , 163, 5 4 - 5 9 .
Boyle, S. H., Williams, R. B., M a rk , D. B., Brurn-
1994). The appropriate balance is needed for
mett, B. H., Siegler, I. C., & Barefoot, J . C. (2005).
psychological, social, and physical health. Hostility, age, and mortality in a sample of cardiac
patients. A m er ica n J o u r n a l o f C a rd io lo g y , 9 6 , 6 4 -
66.
A ckn o w ledgm en ts Brehm, J ., & Rah n, W. (1997). Individual-level evi
dence for the causes and consequences of social
capital. A m er ica n J o u r n a l o f P o litica l S cien ce, 41,
T h is work was supported in p art by Grant N o. 999-1023.
R01 H L 5 4 7 8 0 from the National H eart, Lung, Brissette, I., & Cohen, S. (2002). The contribution of
and Blood Institute with cofunding from the individual differences in hostility to the associations
14. H o stility and P ro n e n e s s to A n g e r 223
between daily interpersonal conflict, affect, and Dodge, K. A. (200 6). Translational science in action:
sleep. P erso n a lity a n d S o c ia l P sy ch o lo g y B ulletin , Hostile attributional style and the development of
28,1265-1274. aggressive behavior problems. D e v e lo p m e n t a n d
Brondolo, E., Rieppi, R., Erickson, S. A., Bagiella, E., P sy c h o p a th o lo g y , 18, 7 9 1 - 8 1 4 .
Shapiro, P. A., McKinley, P., et al. (2003). Hostility, Dougherty, D. M ., Bjork, J . M ., Marsh, D., &c Moeller,
interpersonal interactions, and ambulatory blood F. G. (1999). Influence of trait hostility on trypto
pressure. P sy c h o s o m a tic M ed icin e, 65, 1 0 0 3 - 1 0 1 1 . phan depletion-induced laboratory aggression. P sy
Brosschot, J . F., & Thayer, J. F. (1998). Anger inhibi ch ia try R esea rc h , 8 8 , 2 2 7 - 2 3 2 .
tion, cardiovascular recovery, and vagal function: A Enkleman, H. C., Bishop, G. D., Tong, FI. M . W.,
model of the link between hostility and cardiovas Diong, S. M., Why, V. P., Khader, M ., et al. (2005).
cular disease. A n n als o f B e h a v io r a l M ed icin e, 2 0 , The relationship of hostility, negative affect, and
326-332. ethnicity to cardiovascular responses: An ambula
Brummett, B. H., Maynard, K. E., Babyak, M. A., tory study in Singapore. In te r n a tio n a l J o u r n a l o f
Haney, T. L., Siegler, I. C., Helms, M . J., et al. P sy ch o p h y sio lo g y , 5 6 , 18 5-19 7.
(1998). Measures of hostility as predictors of facial Eysenck, H., & Eysenck, M. (1985). P erson ality a n d
affect during social interaction: Evidence for con in d iv id u a l d iffe r e n c e s : A n atu ral s c ie n c e a p p r o a c h .
struct validity. A n n als o f B eh a v io r a l M ed icin e, 2 0 , New York: Plenum Press.
1 6 8 -1 73 . Gallo, L. C., & Smith, T. W. (1998). Construct valida
Bunde, J . , & Suls, J. (20 06 ). A quantitative analysis tion of health-relevant personality traits: Interper
of the relationship between the C o ok-M ed le y H os sonal circumplex and five-factor model analyses of
tility Scale and traditional coronary artery disease the Aggression Questionnaire. I n te r n a tio n a l J o u r
risk factors. H ea lth P sy ch olog y , 2 5 , 4 9 3 - 5 0 0 . n a l o f B e h a v io r a l M ed icin e, 5, 129 -1 47 .
Bushman, B. J., Cooper, H. M ., &c Lemke, K. M. Garssen, B. (2007). Repression: Finding our way in the
(1991). Meta-analysis of factor analyses: An illus maze of concepts. J o u r n a l o f B eh a v io r a l M ed icin e,
tration using the Buss-D urkee Hostility Inventory. 30, 471-481.
P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , 17, Goldbacher, E. M ., & Matthews, K. A. (20 07 ). Are
344-349. psychological characteristics related to risk of the
Buss, A. H., & Durkee, A. (1957). An inventory for as metabolic syndrome?: A review of the literature.
sessing different kinds of hostility. J o u r n a l o f C o n A n n als o f B e h a v io r a l M ed icin e, 3 4 , 2 4 0 - 2 5 2 .
su ltin g P sy ch olog y , 2 1, 3 4 3 - 3 4 9 . Golden, S. H. (200 7). A review of the evidence for a
Buss, A. H., & Perry, M. (1992). The aggression ques neuroendocrine link between stress, depression and
tionnaire. J o u r n a l o f P erson ality a n d S o c ia l P sy diabetes. C u rren t D ia b e te s R e v iew , 3 , 2 5 2 - 2 5 9 .
c h o lo g y , 63, 4 5 2 - 4 5 9 . Gross, J. J . (2002). FImotion regulation: Affective, cog
Butler, FI. A., Egloff, B., Wilhelm, F. H., Smith, N. C., nitive, and social consequences. P sy ch o p h y sio lo g y ,
Erickson, E. A., & Gross, J. J . (2003). The social 39, 2 8 1 - 2 9 1 .
consequences of expressive suppression. E m o tio n , Guyll, M ., & Contrada, R. (1998). Trait hostility and
3, 4 8 - 6 7 . ambulatory cardiovascular activity: Responses to
Butler, E. A., Lee, T. L., & Gross, J. J. (20 07 ). Emotion social interaction. H ea lth P sy ch olog y , 17, 3 0 - 3 9 .
regulation and culture: Are the social consequences Guyll, M., & Madon , S. J. (2003). Trait hostility: The
of emotion suppression culture specific? E m o tio n , breadth and specificity of schema effects. P ers o n a l
7, 3 0 - 4 8 . ity a n d In d iv id u a l D iffe r e n c e s , 3 4 , 6 8 1 - 6 9 3 .
Carver, C. S., & Miller, C. J. (20 06 ). Relations of se Hall, P., & Davidson, K. (1996). The misperception of
rotonin function to personality: Current views and aggression in behaviorally hostile men. C o g n itiv e
a key methodological issue. P sy ch iatry R esea rc h , T h era p y a n d R e se a rch , 2 0 , 3 7 7 - 3 8 9 .
144, 1 -1 5. Haney, T. L., Maynard, K. E., Houseworth, S. J ., Sch
Cattell, R. B. (1946). D escrip tio n a n d m ea su re m en t o f erwitz, L. W., Williams, R. B., &c Barefoot, J. C.
p er so n a lity . Yonkers on Hudson, NY: World Book. (1996). Interpersonal Hostility Assessment Tech
Chesney, M. A., Hecker, M ., & Black, G. A. (1989). nique: Description and validation against the crite
Coronary-prone components of Type A behavior in rion of coronary artery disease. J o u r n a l o f P ers o n
the W CG S: A new methodology. In B. K. Houston ality A ssessm en t, 6 6 , 3 8 6 - 4 0 1 .
&c C. R. Snyder (Eds.), T ype A b e h a v io r p a tte rn : Harper, S., Lynch, J ., Hsu, W. L., Everson, S. A., Hil-
R esea rc h , th eo ry , a n d in te rv e n tio n (pp. 168-188). lemeier, M. M ., Raghunathan, T. E., et al. (2002).
New York: Wiley. Life course socioeconomic conditions and adult
Contrada, R. J ., & Jussim, L. (1992). W hat does the psychosocial functioning. I n te r n a tio n a l J o u r n a l o f
C ook-M ed le y Hostility Scale measure?: In search E p id e m io lo g y , 3 1, 3 9 1 - 4 0 3 .
of an adequate measurement model. J o u r n a l o f A p Haukkala, A. (2002). Socio-economic differences in
p lie d S o c ia l P sy ch olog y , 2 2 , 6 1 5 - 6 2 7 . hostility measures: A population-based study. P sy
Cook , W. W., & Medley, D. M . (1954). Proposed hos c h o lo g y a n d H ea lth , 17, 1 9 1 - 2 0 2 .
tility and pharisaic virtue scales for the M M P I . Haukkala, A., & Uutela, A. ( 2 0 00 ). Cynical hostility,
J o u r n a l o f A p p lied P sy ch olog y , 3 8 , 4 1 4 - 4 1 8 . depression, and obesity: T he moderating role of ed
Costa, P. T., Jr., & Mc Crae, R. R. (1992). N E O P l-R ucation and gender. In te r n a tio n a l J o u r n a l o f Elating
P ro fe ssio n a l M an u al. Odessa, FI.: Psychological D iso rd ers, 27, 1 0 6 - 1 0 9 .
Assessment Resources. Helgeson, V. S. (1994). Relation of agency and commu
Deffenbacher, J . L., Oetting, E. R., Lynch, R. S., & nion to well-being: Evidence and potential explana
Morris, C. D. (1996). T he expression of anger and tions. P sy c h o lo g ic a l B u lletin , 116, 4 1 2 - 4 2 8 .
its consequences. B e h a v io r R e se a rc h a n d T hera p y , Hogan, B. E., & Linden, W. (200 4). Anger response
34, 575-590. styles and blood pressure: At least don’t ruminate
224 III. E M O T IO N A L D IS P O S IT IO N S
egies for investigating interactions between mea nation upon anger, hostile view of relationships, and
sured genes and measured environments. A rchives violence among African American boys. J o u r n a l o f
o f G e n e r a l P sy ch iatry , 6 2 , 4 7 3 - 4 8 1 . H ea lth a n d S o c ia l B eh a v io r, 47, 3 7 3 - 3 8 9 .
Newton, T. L., & Kiecolt-Glaser, J . K. (1995). H os Smith, T. W., 8c Brown, P. W. (1991). Cynical hostility,
tility and the erosion of marriage quality during attempts to exert social control, and cardiovascular
early marriagc. J o u r n a l o f B e h a v io r a l M ed icin e, 18, reactivity in married couples. J o u r n a l o f B e h a v io r a l
601-619. M ed icin e, 14, 5 8 1 - 5 9 2 .
Orth, U., Sc Wieland, E. (200 6). Anger, hostility, and Smith, T. W., 8c Frohm, K. D. (1985). W h a t ’s so un
posttraumatic stress disorder in trauma-exposed healthy about hostility?: Construct validity and
adults: A meta-analysis. J o u r n a l o f C o n su ltin g a n d psychosocial correlates of the Cook and Medley Ho
C lin ic a l P sy c h o lo g y , 74, 6 9 8 - 7 0 6 . scale. H ea lth P sy ch olog y , 4, 5 0 3 - 5 2 0 .
Porter, L. S., Stone, A. A., & Schwartz, J . E. (1999). Smith, T. W., 8c Gallo, L. C. (1999). Hostility and ca r
Anger expression and ambulatory blood pressure: A diovascular reactivity during marital interaction.
comparison of state and trait measures. P s y c h o s o P sy c h o s o m a tic M ed icin e, 61, 4 3 6 - 4 4 5 .
m a tic M ed icin e, 61, 4 5 4 - 4 6 3 . Smith, T. W„ Gallo, L. C „ Goble, L., Ngu, L. Q., &
Price, J. M „ & Glad, K. (2003). Hostile attributional Stark, K. A. (1998). Agency, communion, and ca r
tendencies in maltreated children. J o u r n a l o f A b diovascular reactivity during marital interaction.
n o r m a l C h ild P sy ch olog y , 31, 3 2 9 - 3 4 3 . H ea lth P sy ch olog y , 17, 5 3 7 - 5 4 5 .
Rempel, J. K., Holmes, J. G., & Zanna, M . P. (1985). Smith, T. W., Glazer, K., Ruiz, J. M ., tic Gallo, L. C.
Trust in close relationships. J o u r n a l o f P erson ality (2004). Hostility, anger, aggressiveness, and c o ro
a n d S o c ia l P sy ch olog y , 49, 9 5 - 1 1 2 . nary heart disease: An interpersonal perspective on
Rosenman, R. H., Swan, G. E., & Carmilli, D. (1988). personality, emotion, and health. J o u r n a l o f P er
Definition, assessment, and evolution of the Type A so n a lity , 7 2 , 1 2 1 7 - 1 2 7 0 .
behavior pattern. In B. K. Houston 8c C . R. Snyder Smith, T. W., Pope, M. K., Sanders, J . D., Allred, K.
(Eds.), T ype A b e h a v io r p a tte rn : R e se a rch , th eory , D., 8c O ’Keefe, J. L. (1988). Cynical hostility at
a n d in terv en tio n (pp. 8 - 3 1 ). New York: Wiley. home and work: Psychosocial vulnerability across
Rotter, J . (1967). A new scale for the measurement domains. J o u r n a l o f R e se a rch in P erson ality , 2 2 ,
of interpersonal trust. Jo u r n a l o f P erson ality , 3 5, 525-548.
651-665. Smith, T. W., Sanders, J. D., & Alexander, J . F. (1990).
Rotter, J . (1980). Interpersonal trust, trustworthiness, W hat does the C ook and Medley Hostility Scale
and gullibility. A m eric a n P sy ch olog ist, 3 5 , 1-7. measure?: Affect, behavior, and attributions in the
Scherwitz, L., Perkins, I.., Chesney, M., Sc Hughes, marital context. J o u r n a l o f P erson ality a n d S o c ia l
G. (1991). C ook-M ed le y Hostility Scale and sub P sy ch olog y , 5 8 , 6 9 9 - 7 0 8 .
sets: Relationship to demographic and psychoso Smith, T. W., Uchino, B. N., Berg, C. A., Florsheim,
cial characteristics in young adults in the C A R D IA P., Gale, G., Hawkins, M ., et al. (200 7). Hostile
Study. P sy c h o s o m a tic M ed icin e, 5.3, 3 6 - 4 9 . personality traits and coronary artery calcification
Scherwitz, L., Perkins, L., Chesney, M., Hughes, G., in middle-aged and older married couples: Different
Sidney, S., 8c Manolio, T. A. (1992). Hostility and effects for self-reports versus spouse ratings. P sy
health behaviors in young adults: The C A R D IA c h o s o m a tic M ed icin e, 69, 4 4 1 - 4 4 8 .
Study. A m er ic a n Jo u r n a l o f E p id e m io lo g y , 136, Spielberger, C. D., Jacobs, G., Russell, S. F., 8c Crane,
1 36 -1 4 5 . R. J. (1983). Assessment of anger: The State-Trait
Shekelle, R. B., Gale, M ., Ostfeld, A. M ., 8c Paul, O. Anger Scale. In J . N. Butcher & C. D. Spielberger
(1983). Hostility, risk of coronary heart disease, and (Eds.), A d v a n ces in p e r s o n a lity a s s e ss m e n t (Vol. 2,
mortality. P sy c h o s o m a tic M ed icin e, 4 5, 1 0 9 -1 14 . pp. 1 5 9 -1 8 7 ). Hillsdale, NJ: Erlbaum.
Shumaker, S. A., 8c Hill, D. R. (1991). Gender differ Spielberger, C. D., Joh nson, E. H., Russell, S. F., Crane,
ences in social support and physical health. H ealth R. J ., Jacobs, G. A., Sc Worden, T. J . (1985). The
P sy ch olog y , 10, 1 02-111. experience and expression of anger: Construction
Siegel, J . M . (1986). The Multidimensional Anger In and validation of an anger expression scale. In M.
ventory. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l A. Chesney Sc R. H. Rosenman (Eds.), A n ger a n d
og y , 51, 1 9 1 - 2 0 0 . h o stility in c a r d io v a s c u la r a n d b e h a v io r a l d iso rd er s
Siegman, A. W. (1994). Cardiovascular consequences (pp. 5 - 3 0 ) . New York: McGraw-Hill.
of expressing and repressing anger. In A. W. Sieg Steptoe, A., Hamer, M., 8c Chida, Y. (20 07 ). The ef
man 8c T. W. Smith (Eds.), Anger, h ostility , a n d the fects of acute psychological stress on circulating
h e a r t (pp. 1 7 3 - 1 9 8 ) Hillsdale, NJ: Erlbaum. inflammatory factors in humans: A review and
Siegman, A. W., Kubzansky, I.. D., Kawachi, I., Boyle, meta-analysis. B rain B e h a v io r a n d Im m u n ity , 21,
S., Vokonas, P. S., 8c Sparrow, D. (20 00 ). A prospec 901-912. '
tive study of dominance and coronary heart disease Stoney, C. M ., Sc Engebretson, T. O. (1994). Anger
in the Normative Aging Study. A m er ica n J o u r n a l o f and hostility: Potential mediators of the gender dif
C a rd io lo g y , 8 6 , 145-149. ference in cardiovascular disease. In A. W. Siegman
Siegman, A. W., Townsend, S. T., Civelek, A. C., &c Sc T. W. Smith (Eds.), A nger, h ostility , a n d the h ea rt
Blumenthal, R. S. ( 2 0 00 ). Antagonistic behavior, (pp. 2 1 5 - 2 3 8 ) . Hillsdale, NJ: Erlbaum.
dominance, hostility, and coronary heart disease. Suarez, E. C., Kuhn, C. M., Schanberg, S. M ., W il
P sy c h o s o m a tic M ed ic in e, 6 2 , 2 4 8 - 2 5 7 . liams, R. B., 8c Zimmerman, E. A. (1998). Neuroen
Simons, R. L., Simons, L. G., Burt, C. H., Drummund, docrine, cardiovascular, and emotional responses
H., Stewart, E., Brody, G. H., et al. (200 6 ). Sup of hostile men: The role of interpersonal challenge.
portive parenting moderates the effect of discrimi P sy c h o s o m a tic M ed icin e, 6 0 , 7 8 - 8 8 .
226 III. E M O T IO N A L D IS P O S IT IO N S
Surwit, R. S., Williams, R. B., Siegler, I. C., Lane, J. actional perspective on the five-factor model. In J . S.
D., Helms, M . J ., Applegate, K. L., et al. (2002). Wiggins (Ed.), T h e fiv e - fa c t o r m o d e l o f p e r so n a lity :
Hostility, race, and glucose metabolism in nondia T h e o r e t ic a l p er sp e c tiv e s (pp. 8 8 - 1 6 2 ). New York:
betic individuals. D ia b e t e s C a re, 2 5 , 8 3 5 - 8 3 9 . Guilford Press.
Swenson, W. M ., Pearson, J . S., & Osbourne, D. Williams, J. E., Nieto, F. J ., Sanford, C. P., Couper,
(1973). A n M M P I s o u r c e b o o k : B a sic item , sc a le, D. J . , & Tyroler, H. A. (2002). T he association
a n d p a tte r n d a t a o n 5 0 ,0 0 0 m e d ic a l p a tien ts. M i n between trait anger and incident stroke risk: The
neapolis: University of Minnesota Press. Atherosclerosis Risk in Communities (ARIC) study.
Turner, R. J ., Russell, D., Glover, R ., & Hutto, P. S tr o k e , 3 3 , 1 3 - 2 0 .
( 2 0 07 ). T he social antecedents of anger proneness Williams, R. B., & Barefoot, J. C. (1988). T he emerg
in young adulthood. J o u r n a l o f H ea lth a n d S o c ia l ing role of the hostility complex. In B. K. Houston
B eh a v io r , 4 8 , 6 8 - 8 3 . & C. R. Snyder (Eds.), T ype A b e h a v io r p a tte rn :
Weidner, G., Rice, T., K no x, S. S., Ellison, R. C., Prov R e se a rc h , th eo r y , a n d in terv en tio n (pp. 189-21 1).
ince, M . A., Rao, D. C., et al. ( 2 00 0 ). Familial re New York: Wiley.
semblance for hostility: T he National Heart, Lung, Zeichner, A., Giancola, P. R ., & Allen, J. D. (1995).
and Blood Institute Family He art Study. P s y c h o s o Effects of hostility on alcohol stress-response damp
m a tic M ed icin e, 6 2 , 1 9 7 - 2 0 4 . ening. A lc o h o lis m : C lin ica l a n d E x p e r im e n ta l R e
Wiggins, J. S., & Trapnell, P. D. (1996). A dyadic inter sea rch , 19, 9 7 7 - 9 8 3 .
CHAPTER 15
■ * &. ♦ m » A ♦
Loneliness
J o h n T . C a c io p p o
L o u is e C . H a w k l e y
227
228 III. E M O T IO N A L D IS P O S IT IO N S
of loneliness, individuals made to feel lonely gatherers who chose not to return to share
reported significantly more negative mood their food and offer protection to mother
and lower self-esteem, optimism, social and child (i.e., who felt no loss severing so
skills, social support, sociability, extraver cial/family bonds) may have survived to re
sion, and agreeableness and greater shyness, produce again, but their offspring and, with
anxiety, anger, fear of negative evaluation, them their genes, would have been unlikely
and neuroticism (Cacioppo, Hawkley, et al., to survive to reproduce. In contrast, hunter-
200 6 ). These results identify loneliness as gatherers whose genetic predisposition in
a potential causal factor in characteristics clined them to share food with their families
such as self-esteem, depressive symptomatol may have lowered their own chances of sur
ogy, shyness, and so forth. vival but increased the survival odds of their
A third conceptual approach to loneliness offspring, thereby propagating their genes.
is represented by cognitive discrepancy the O f course, a hunter-gatherer who survives a
ory, which specifies loneliness as the conse famine may then live to have another family
quence of altered social perceptions and at another day, suggesting that no single strat
tributions. Specifically, loneliness is defined egy is necessarily best. Such an evolutionary
as the distress that occurs when one’s so scenario suggests that humans might inherit
cial relationships are perceived as being less differing tendencies to experience loneliness.
satisfying than what is desired (Peplau & Adoption and twin studies among children
Perlman, 1982). From a cognitive discrep and adults have confirmed that loneliness
ancy perspective, it is clear that loneliness has a sizeable heritable component (Booms-
is not synonymous with being alone, nor ma, Willemsen, Dolan, Hawkley, & Caciop-
does being with others guarantee protection pio, 20 0 5 ; McGuire & Clifford, 2000).
from feelings of loneliness (Peplau & Perl
man, 1982). Rather, discrepancies between
ideal and perceived interpersonal relation Measures o f Loneliness
ships produce and maintain feelings of lone
liness. Individual differences in loneliness are typi
A fourth approach derives from an evolu cally measured using paper-and-pencil ques
tionary analysis of loneliness, with an em tionnaires, a number of which are reviewed
phasis on inclusive fitness (Cacioppo, Hawk in Cramer and Barry (1999). Among the mul
ley, et al., 20 0 6 ). This approach calls into tidimensional scales tapping emotional and
question the conceptualization of loneliness social loneliness are the De Jong Gierveld
as an aversive condition without redeeming Loneliness Scale (De Jong Gierveld & Kam-
features and instead conceptualizes loneli phuis, 1985) and the Social and Emotional
ness as an aversive condition that promotes Loneliness Scale for Adults (SELSA; DiTom-
inclusive fitness by signaling ruptures in so maso & Spinner, 1993). These two scales
cial connections to motivate the repair or probe social relational deficits with items
replacement of these connections. For many such as “I have friends to whom I can talk
species, offspring need little or no parenting about the pressures in my life” and “There
to survive and reproduce. H om o sapiens, are plenty of people I can rely on when I have
however, are born to the longest period of problems.” The De Jong Gierveld Loneliness
abject dependency of any species. Simple Scale probes emotional relational deficits
reproduction, therefore, is not sufficient to with items such as “I experience a general
ensure that one’s genes make it into the gene sense of emptiness,” whereas the SELSA
pool. For one’s genes to make it to the gene distinguishes between relational deficits in
pool, the offspring must survive to repro family relationships (e.g., “I feel close to my
duce. Moreover, social connections and the family”) and romantic relationships (e.g., “I
behaviors they engender (e.g., cooperation, have someone who fulfills my needs for in
altruism, alliances) enhance the survival and tim acy”).
reproduction of those involved, increasing The most frequently used instrument is
inclusive fitness. the UCLA Loneliness Scale developed at
Humans walked the earth as hunter- the University of California at Los Angeles
gatherers for tens of thousands of years, (version 3; Russell, 1996). Items probe the
often under conditions of privation. H unter- frequency and intensity of loneliness-related
15. L o n e lin e ss 229
experiences (e.g., “How often do you feel Heritability estimates of complex traits
alone?” “How often do you feel part of a such as loneliness may also change across
group of friends?” and “How often do you the lifespan, as the frequency, duration, and
feel that there are people who really under range of exposure to environmental influ
stand you?”). To avoid response bias, the ences accrues. To address this question, data
terms lonely and loneliness do not appear in from young adult and adult Dutch twins (av
any of the items. Although conceptualized erage age 24 years) in the Netherlands Twin
as a unidimensional scale, factor analyses Register Study were analyzed with genetic
of the UCLA Loneliness Scale have revealed structural equation models, which provide
anywhere from two to five dimensions. estimates of the shared environmental and
Second-order factor analyses, however, unique environmental contributions, as well
have shown a singie overarching loneliness as the genetic contributions (Boomsma,
construct (Hawkley, Browne, & Cacioppo, Willemsen, Dolan, Hawkley, & Cacioppo,
2 0 0 5 ; Russell, 1996) that supports its use as 2005). The estimate of genetic contribu
a unidimensional bipolar measure of lone tions to variation in loneliness in adults was
liness. An abbreviated three-item version of 4 8 % , with the remaining variance explained
this scale has been validated for use in large by unique environmental factors. Thus the
population surveys (Hughes, Waite, Hawk heritability estimates in adults were similar
ley, & Cacioppo, 2004). to those found previously in children. M ore
over, no evidence was found for sex or age
differences in genetic architecture or for
Stability
nonadditive genetic effects.
Temporal stability of loneliness scores is A follow-up longitudinal study of young
relatively high, with test-retest reliabilities Dutch twins at ages 7, 10, and 12 years
of .69, .57, and .51 across 2 , 3, and 5 years, found that the influence of shared family
respectively, in children between 7 and 12 environment increased from .06 and .08 at
years of age (Bartels, Cacioppo, Hudziak, ages 7 and 10 to .35 at age 12, paralleling a
& Boomsma, 2 0 0 8 ); .74 across a 2-week to reduction in heritability estimates from .58
2-month time period in young adults (Ca and .56 at ages 7 and 10 to .26 at age 12
cioppo, Hawkley, et al., 2 0 0 6 ); and from (Bartels et al., 2008). As these children move
.73 to .84 across 1-2 years in middle-aged through adolescence and adapt to new bio
and older adults (Cacioppo, Hughes, Waite, logical and social challenges, environmental
Hawkley, & Thisted, 2 0 0 6 ; Russell, 1996). influences are expected to decline and heri
table dispositions to reemerge to levels ob
served in the young adult and adult twins.
A ntecedents o f Loneliness
H eritability Predictors
If the motivation to form and maintain social Research on the predictors of loneliness is
bonds has evolutionary origins, one might predominantly cross-sectional, and longitu
expect significant genetic contributions to dinal studies have tended to focus on older
loneliness. In a study of adoptive families, adults. These limitations notwithstanding, a
loneliness data were obtained from 69 bio sizeable body of research indicates that so
logically related sibling pairs and 64 unre ciodemographic factors, social roles, quanti
lated pairs when the children were 9, 10, 11, ty and quality of social contact, health, and
and 12 years of age. In a second study, 22 dispositions contribute to individual differ
monozygotic (M Z) twins, 40 dizygotic (DZ) ences in feelings of loneliness.
twins, and 80 full siblings 8 -1 4 years of age
completed a 16-item scale to assess loneli
Sociodemographics
ness in relation to their schoolmates. Results
revealed significant genetic {h2 = 55 and Structural factors such as age, gender, race/
4 8 % , respectively, in Studies 1 and 2) and ethnicity, education, and income constrain
unshared environmental contributions to in opportunities for integration into meaning
dividual differences in loneliness (McGuire ful groups and social roles, and these factors
& Clifford, 2000). contribute to individual differences in lone
230 III. E M O T IO N A L D IS P O S IT IO N S
liness. Age has been associated with loneli the relationship-enhancing benefit of higher
ness, but the shape of that association is a social status and self-esteem associated with
flattened U, not linear as conventional wis this accomplishment (Hawkley, Hughes, et
dom might suggest. Prevalence and intensity al., 2007).
of lonely feelings are greater in adolescence
and young adulthood (i.e., 1 6 -2 5 years of
Social Roles
age) than in any other age group except
the oldest old (i.e., >80 years) (Pinquart & Marriage is well known to protect against
Sorensen, 2003). Findings in longitudinal loneliness, and loneliness is greater among
studies are consistent with those in cross- those who are divorced or never married
sectional studies, and cohort effects have (Dykstra & Fokkema, 2007; Pinquart,
been largely ruled out as an explanation for 2003). Retirement and unemployment also
age effects (Pinquart & Sorensen, 2003). represent loss of social roles, and both
Indeed, in the Americans’ Changing Lives groups are lonelier than the employed (Han-
study of adults 24 years of age and older, sson, Briggs, &C Rule, 1990; Viney, 1985).
age was inversely associated with loneliness Voluntary group membership (e.g., social
even when the loneliness-augmenting effect club, athletic team) (Cattan, White, Bond,
of lost social roles (e.g., marriage, work) was & Learmouth, 2005) and religious/church
held constant (Schnittker, 2007). membership (Johnson & Mullins, 1989) are
Females tend to report slightly greater other roles that have been observed to pro
loneliness than males, but only when the tect against loneliness.
measure includes terms such as lonely or
loneliness (Pinquart & Sorensen, 2003).
Social Contact Quantity and Quality
When examined as a function of marital sta
tus, however, nonmarried men are lonelier Smaller social networks and less frequent in
than nonmarried women (Pinquart, 2003). teractions with friends and family promote
In the United States, African Americans loneliness (Dykstra, van Tilburg, & De
tend to be lonelier than whites (Barg et al., Jong Gierveld, 20 0 5 ; Pinquart & Sorensen,
2 0 0 6 ), although single African American 2003). Accordingly, situational factors that
women were less lonely than Latina and influence the availability of social opportu
white women in the Southern California nities have also been associated with loneli
Social Survey (Tucker & Mitchell-Kernan, ness. For instance, geographical relocation
1998). Cultural differences in loneliness lev predicts loneliness in first-year university
els have also been observed. Chinese students students (Shaver, Furman, & Buhrmester,
at an American university reported greater 1985). Conversely, participation in senior
loneliness than their U.S. counterparts (An center activities protects against loneliness
derson, 1999), an effect that some have ar in older adults living alone (Aday, Kehoe, &
gued is attributable to the Asian collectivist Farney, 20 0 6 ). Contact with friends is more
perspective in the context of an individual important than contact with adult children
istic American society (Goodwin, Cook, & and other family members in preventing
Yung, 2001). loneliness (Pinquart & Sorensen, 2003), and
Greater educational attainment and high the chronic unavailability of social part
er income are associated with less loneliness ners with whom to enjoy social activities
(Pinquart & Sorensen, 2003), but this effect has been associated with greater loneliness
is predominantly indirect and has been at (Rook, 1984).
tributed to larger social networks (Dykstra Social relationship quality is a more po
& De Jong Gierveld, 1999; Lauder, M um tent predictor of loneliness than quantity of
mery, & Sharkey, 20 0 6 ). Holding social social contacts, and this is true of relation
network size constant, however, the attain ships with friends, family, and adult children
ment of a high school diploma continued to (Pinquart & Sorensen, 2003). In addition,
protect against loneliness in our population- although marriage is generally protective,
based sample of middle-aged adults in the only marriages that are close and satisfy a
Chicago Health, Aging, and Social Rela need for a confidant serve to reduce loneli
tions Study (CHASRS), possibly indicating ness (Olson & Wong, 2001).
15. L o neliness 231
intensely to the negatives but also experience to interpersonal or collective social ties than
less of a soothing uplift from the positives did nonlonely participants. It made no dif
(Hawkley, Preacher, & Cacioppo, 2007). ference whether the detail, which was pre
Even when they succeed in eliciting nurtur sented in diary format, was emotionally pos
ing support from a friend or loved one, they itive or negative (Gardner, Pickett, Jeffries,
tend to perceive the exchange as less fulfill & Knowles, 2005). In another study, partici
ing (Hawkley, Burleson, Berntson, & Ca pants who were asked to “relive” a rejection
cioppo, 2003). experience showed greater attention to vocal
The lonely are aware that their social tone in a vocal Stroop task than did partici
needs are not being met, but they perceive pants asked to relive an academic failure ex
that they do not have a great deal of con perience or a neutral experience (the walk to
trol over their ability to fulfill those needs campus that morning) (Pickett, Gardner, &
(Solano, 1987). Tending to be more anxious, Knowles, 2004).
pessimistic, and fearful of negative evalua Greater attention to social cues does not
tion than people who feel good about their ensure greater social sensitivity, however.
social lives, lonely people are more likely We have noted that lonely individuals are
to act and relate to others in ways that are less accurate at decoding facial and postural
anxious, negative, and self-protective, which expressions of emotion (Pickett & Gardner,
leads paradoxically to self-defeating behav 2 0 0 5 ; Pitterman & Nowicki, 2004). In addi
iors (Cacioppo & Hawkley, 2005). For in tion, the participants who relived a rejection
stance, Rotenberg (1994) found that lonely experience were less accurate in decoding
and nonlonely individuals were equally like the meaning of the words in the vocal Stroop
ly to cooperate with a stranger at the outset task (Pickett et al., 2004).
and during the early trials of a prisoner’s di A lack of correspondence between atten
lemma game in which the stranger was play tion and accuracy in responses to social cues
ing a tit-for-tat strategy. As play continued has also been demonstrated in a brain im
and they betrayed their partners, only to aging study of lonely and nonlonely young
find that their partners then betrayed them, adults. When presented with equally arousing
the lonely individuals were especially likely positive and negative pictures of scenes and
to escalate the betrayals than nonlonely in objects (nonsocial stimuli) and people (social
dividuals. stimuli), a set of brain regions often associ
Not only do the lonely contribute to their ated with visual attention and perspective
own negative reality, but others also begin taking varied in response to negative social
to view them more negatively and begin to (in contrast to matched nonsocial) pictures.
act accordingly (Lau & Gruen, 1992). One Relative to the nonlonely, lonely individu
study showed that individuals who were told als showed greater visual cortical activa
that an opposite-gender partner they were tion (consistent with greater attention to the
about to meet was lonely subsequently rated negative social than nonsocial pictures) and
that partner as being less sociable. The in less activation of the temporo-parietal junc
dividuals primed to have these expectations tion (consistent with less attention devoted
also behaved toward their partners in a less to the other person’s perspective). Another
sociable manner than they did toward part set of brain regions, associated with reward
ners whom they expected to be nonlonely systems (i.e., ventral striatum), was found to
(Rotenberg, Gruman, & Ariganello, 2002). be down-regulated in lonely, compared with
Once this negative feedback loop starts roll nonlonely, individuals when viewing posi
ing, the cycle of defensive behavior and neg tive social (in contrast to matched nonsocial)
ative social results spins even further down pictures— results consistent with the finding
hill. In essence, lonely individuals inhabit an that lonely individuals derive less pleasure
inhospitable social orbit that repels others or from viewing positive social circum stanc
elicits their negative responses. es than nonlonely individuals (Cacioppo,
Expecting social rejection, the lonely are Norris, Decety, Monteleone, & Nusbaum,
keenly attuned to cues of social acceptance 2009). This latter finding may bear on the
in their environment. In a test of social finding that lonely individuals find positive
monitoring, lonely participants remembered social interactions during the course of a
a greater proportion of information related normal day less satisfying than do nonlonely
15. L o n e lin e ss 233
individuals (Hawkley, Preacher, & Caciop interpret and respond to ambiguous social
po, 2007). behavior in a more negative, off-putting
One might expect that a lonely person, fashion, thereby confirming their construal
hungry to fulfill unmet social needs, would of the world as threatening and beyond their
be very accepting of a new acquaintance, control. These cognitions, in turn, activate
just as a famished person might take plea neurobiological mechanisms that, with time,
sure in food that was not as tasty as their take a toll on health.
usual fare. Indeed, experimentally increas
ing individuals’ feelings of social isolation
leads to an increase in anthropomorphism Consequences o f Loneliness
that reflects increased efforts to reconnect
Self-R egulation
(Epley, Waytz, & Cacioppo, 2 0 0 7 ). Howev
er, when confronted with an opportunity to Self-regulation refers to the capacity to
form a social connection, studies show that change one’s cognitions, emotions, and/or
the lonely are actually far less accepting of behavior to better meet social standards and
potential new friends than are the nonlonely personal goals. Evidence from young adults
(Rotenberg & Kmill, 1992). Similarly, lonely who performed a dichotic listening task in
students were less responsive to their class dicates that self-regulatory processes are im
mates during class discussions and provided paired in lonely individuals (Cacioppo et al.,
less appropriate and less effective feedback 2000). In the dichotic listening task, partici
than nonlonely students (Anderson & M ar pants are asked to identify the consonant-
tin, 1995). Lonely undergraduates also held vowel pair presented in the left or right ear.
more negative perceptions of their room Typically, performance shows a right-ear ad
mates than did the nonlonely (Wittenberg vantage. In addition, performance is gener
& Reis, 1986), and this perceptual divide ally better for the ear to which participants
widened as one moved from roommates to have been instructed to attend. In our study
suite mates to floor mates to dorm mates of lonely and nonlonely young adults, we ob
(Cacioppo & Hawkley, 2005). served a right-ear advantage and an atten
Time also plays a role in constructing tional instruction advantage, but an inter
these negative “realities.” Researchers asked action between these effects revealed that,
participants to interact with a friend and to although lonely and nonlonely individuals
rate the quality of the relationship and the showed a large attentional shift to the right
communication immediately; after watch ear when so instructed, lonely but not non
ing a videotape of the same social exchange; lonely participants failed to show a left-ear
a few weeks later, after being reminded of advantage when instructed to attend to this
the interaction; and after again watching the ear (Cacioppo et al., 2000).
videotape. At all four measurement points, Experimental manipulations that lead
lonely individuals rated relationship quality people to believe they face a future in so
more negatively than did nonlonely individ cial isolation also increases the challenge
uals. Interestingly, the further in time they of self-regulation (Baumeister & DeWall,
were removed from the social exchange, the 200 5), and this impaired self-regulation has
more negatively they rated it. They were es consequences for mental functioning. Un
pecially negative after the second videotape dergraduate volunteers were provided with
viewing (Duck, Pond, & Leatham, 1994). feedback to induce in them the possibility
When lonely individuals rated the interac that they would experience a future alone
tion soon after it happened, it appears that (e.g., “You’re the type who probably will end
their negative social cognition was reined in up alone. Relationships just won’t last for
by a better understanding of the reasons for you . . . ”), a future belonging (e.g., “You’re
their friend’s behavior. The more time that the type who’ll have rewarding relationships
passed, the more the objective reality suc throughout your life. M ost likely you’ll have
cumbed to the “reality” constructed by the lifelong friendships and a long and happy
lonely individual’s negative social cognition. marriage . . . ”), a future of misfortunes (e.g.,
In sum, lonely individuals are more likely “You’re inherently accident prone. Even if
to construe their world as threatening, to this has not manifested itself in your life
hold more negative expectations, and to so far, you can count on breaking an arm
234 III. E M O T IO N A L D IS P O S IT IO N S
or a leg fairly often . . . ”), or no feedback helplessness and threat. The lonely, both
at all. The future-alone group showed sig young and old, perceived the hassles and
nificantly greater impairment in both speed stresses of everyday life to be more severe
and accuracy on the subsequent Reading than did their nonlonely counterparts. Com
Comprehension Test of the Graduate R e pounding the problem, the lonely found the
cord Exam than either the future-belonging small, social uplifts of everyday life to be less
group or the misfortune control group. Bad intense and less gratifying (Hawkley et al.,
news itself, then, was not enough to cause 2003). The presence of and interaction with
the disruption, only bad news about social other people did not lessen their ratings of
connection. In addition, the mood measure the severity of their everyday stressors.
for the future-alone group showed no indi Stress is not uniformly “bad” but can
cation of emotional distress, suggesting that foster growth and motivate better perfor
any decline in cognitive ability was not a mance. Lonely individuals, however, are
simple matter of being flustered (Baumeister, far less likely than nonlonely individuals
Twenge, &C Nuss, 2002). to see any given stressor as an invigorating
challenge. Instead of responding with opti
mism and active engagement, they tend to
Stress-Related Processes and Outcom es respond with pessimism and avoidance, a
Stress Exposure passive coping strategy that carries its own
costs. Among young adults, the greater the
Surveys of undergraduate students showed degree of loneliness, the more the individual
that lonely and nonlonely young adults do withdrew when faced with stressors. Simi
not differ in their exposure to major life larly, the greater the loneliness, the less the
stressors or in the number of major changes individual sought out emotional support, as
they endured in the previous 12 months (Ca well as instrumental (practical) support (Ca
cioppo et al., 2 0 0 0 ). A “beeper study,” in cioppo, Hawkley, Crawford, et al., 2002).
which these students were asked to sit down Behavioral withdrawal and failure to seek
and record their thoughts and experiences at emotional support are common among
various times during the day, also showed lonely older adults, as well (Hawkley & Ca
that there was no difference in the reported cioppo, 2007).
frequency of hassles or uplifts they experi
ence on an average day, nor in the number
of minor irritants they were confronting H ealth Behaviors
when their beeping wristwatch randomly Poor health behaviors are appealing mecha
interrupted them (Hawkley, Burleson, Bern- nistic candidates for associations between
tson, & Cacioppo, 2003). At least for young loneliness and health. High-calorie, high-fat
adults, then, there was no evidence that lone diets and sedentary lifestyles, for example,
liness increased exposure to objective causes contribute to being overweight or obese,
of stress. However, the number of objective major risk factors for disease in Western
stressors described as “current” had indeed society. In a large cross-sectional survey of
increased among the middle-aged adults in 1,289 adults 18 years and older (mean age
CH A SRS, and the lives of chronically lonely = 4 6 .3 years), the lonely group had a higher
adults involved more objective chronic stres mean BM I and a greater proportion of over
sors than the lives of the nonlonely (Hawk weight/obese individuals than the nonlonely
ley et al., 2008). Moreover, the increasing group did (Lauder, Mummery, Jones, & Ca-
stress load over the course of a lifetime is perchione, 2006).
aggravated by having fewer meaningful re Loneliness differences in physical activity
lationships to provide relief. Loneliness is an have not been observed in studies of young
“added” stress. adults (Hawkley et al., 2003) or in samples
that cover a wide age range from young to
older adults (Lauder et al., 2006). Howev
Stress Perceptions and Coping
er, in our CHASRS sample of middle-aged
Even setting aside the greater number of adults, loneliness was associated with sig
objective stressors in their lives, the lonely nificantly reduced odds of physical activity
express proportionately greater feelings of (OR = 0.65 per SD of loneliness) (Hawkley,
15. L o n e lin e ss 235
Thisted, & Cacioppo, 20 0 7 ). This associa & Cacioppo, 20 0 6 ), consistent with our hy
tion was independent of sociodemographic pothesis of accelerated physiological decline
variables (age, gender, ethnicity, education, in lonely relative to nonlonely individuals.
income), psychosocial variables (depressive
symptoms, perceived stress, hostility, so
Neuroendocrine Functioning
cial support), and self-rated health. M ore
over, deficits in self-regulation, in this case Activity of the hypothalam ic-pituitary-
the diminished tendency of lonely individu adrenocortical (HPA) axis is critical to
als to optimize positive emotions (i.e., poor immune functioning and inflammatory
hedonic emotion regulation), explained the processes, and dysregulation of HPA activ
association between loneliness and physi ity has been associated with loneliness and
cal activity likelihood. Longitudinal analy related psychosocial variables (Hawkley,
ses revealed that loneliness also predicted Bosch, Engeland, M arucha, & Cacioppo,
diminished odds of physical activity in the 2007). Evidence for a loneliness difference
subsequent 2 years (OR = 0.61) and greater in activity of the HPA axis was first report
likelihood of transitioning from physical ac ed by Kiecolt-Glaser and colleagues (1984),
tivity to inactivity (OR = 1.58). These data who observed that lonely nonpsychotic psy
suggest that age-related decreases in physi chiatric inpatients excreted significantly
cal activity among the lonely may exacerbate greater amounts of urinary cortisol than
risk for cardiovascular disease onset and did nonlonely inpatients. More recently,
progression and contribute to an accelerated Steptoe, Owen, Kunz-Ebrecht, and Brydon
physiological decline. (2004) found that lonely individuals showed
a greater 30-minute postawakening increase
in salivary cortisol, and Pressman and col
Physiological Functioning leagues (2005) found that loneliness was as
Cardiovascular Functioning sociated with higher early-morning and late-
night levels of circulating cortisol in young
Blood pressure is a function of cardiac out adult university students.
put (CO) and total peripheral resistance In our study of young adults, we mea
(TPR). In young adults, we found that lone sured catecholamines, adrenocorticotropic
liness was related to differential regulation hormone (ACTH), and cortisol in blood
of systolic blood pressure (SBP). Although samples collected in the morning and again
lonely and nonlonely individuals did not dif in the late afternoon. Analyses revealed that
fer in blood pressure levels, maintenance of only morning levels of ACTH were signifi
blood pressure was attributable to higher cantly higher among lonely than nonlonely
vascular resistance and lower cardiac out students (Cacioppo et al., 2000). We found
put among lonely relative to nonlonely in no loneliness differences in the diurnal pat
dividuals (Cacioppo, Hawkley, Crawford, tern of cortisol secretion or in mean daily
et al., 2 0 0 2 ; Hawkley et al., 2003). Results levels of salivary cortisol, nor did we find
from the Framingham H eart Study indicate differences in HPA reactivity to acute stres
that changes in TPR play a dominant role sors in lonely and nonlonely individuals (Ca
in determining SBP from age 30 until ap cioppo et al., 2000).
proximately age 50 (Franklin et al., 1997). Among older adults in CH ASRS, how
Given the temporal stability of loneliness ever, HPA activity across a 3-day period in
and its substantial heritable component, it is participants’ everyday lives showed an ef
plausible that loneliness-related elevations in fect consistent with a causal role for lone
TP R in early to middle adulthood may lead liness. Diary reports of daily psychosocial,
to higher blood pressure in middle and older emotional, and physical states were com
age. Consistent with this hypothesis, loneli pleted at bedtime on each of 3 consecutive
ness was associated with elevated SBP in a days. Salivary cortisol levels were measured
population-based sample of older adults in at awakening, 30 minutes after awakening,
the CHASRS. Moreover, the association be and at bedtime each day. Multilevel models
tween loneliness and elevated SBP was ex revealed that prior day feelings of loneliness
aggerated in older relative to younger lonely and related feelings of sadness, threat, and
adults in this sample (Hawkley, Masi, Berry, lack of control were associated with a higher
236 III. E M O T IO N A L D IS P O S IT IO N S
cortisol awakening response the next day, ease was more than twice as great in lonely
but morning cortisol awakening responses as in nonlonely individuals (R R scores of
did not predict experiences of these psy 3.2 vs. 1.4, respectively), and this effect was
chosocial states later the same day (Adam, independent of functional physical impair
Hawkley, Kudielka, & Cacioppo, 2006). ments and vascular risk factors and condi
The relevance of this association is particu tions (Wilson et al., 2007). In addition,
larly noteworthy given recent evidence that loneliness was associated with lower cogni
loneliness-related alterations in HPA activity tive ability at baseline and with a more rapid
may occur at the level of the gene. decline in cognition during the 4-year fol
low-up (Wilson et al., 2007). Loneliness has
been associated with poorer self-reported
D N A Transcription Regulation
memory among older black adults (Bazar-
Cortisol can regulate a wide variety of gan & Barbre, 1992) and predicted more
physiological processes via nuclear hor rapid cognitive decline over a 10-year period
mone receptor-mediated control of gene in a Finnish sample of adults 75 years of age
transcription. Cortisol activation of the and older (Tilvis et al., 2004).
glucocorticoid receptor (GR) exerts broad
anti-inflammatory effects by inhibiting Depression
proinflammatory signaling pathways.
Social isolation, however, is associ We have noted that loneliness and depressive
ated with increased risk of inflammation- symptoms are conceptually and empirically
mediated diseases. One possible explana distinct (Cacioppo, Hawkley, et al., 2006).
tion for inflammation-related disease in Nevertheless, levels of loneliness and depres
individuals with high cortisol levels involves sive symptoms covary across the lifespan.
impaired GR-mediated signal transduction Moreover, despite age-group differences in
that prevents the cellular genome from ef loneliness, the association between loneli
fectively “hearing” the anti-inflammatory ness and depressive symptoms appears stable
signal sent by circulating glucocorticoids (i.e., moderately and equivalently positive)
(Cole et al., 200 7 ). Consistent with this across age (Nolen-Hoeksema & Ahrens,
hypothesis, a systematic examination of 2 0 0 2 ).
genome-wide transcriptional alterations in Loneliness has been identified as a risk fac
circulating leukocytes showed increased ex tor for depressive symptoms in longitudinal
pression of genes carrying proinflammatory studies of older adults (Heikkinen & Kaup-
elements and decreased expression of genes pinen, 2004). However, as has been noted,
carrying anti-inflammatory glucocorticoid loneliness is associated with a constellation
response elements in lonely relative to non of demographic and psychosocial risk factors
lonely middle-aged adults (Cole et al., 2007). (e.g., hostility, low social support, perceived
Impaired transcription of glucocorticoid re stress) for depressive symptoms that could
sponse genes and increased activity of proin explain the association between loneliness
flammatory transcription control pathways and depressive symptoms (e.g., Cacioppo,
provide a functional genomic explanation Hawkley, et al., 2006). Recent evidence
for elevated risk of inflammatory disease in from a nationally representative sample of
individuals who experience chronically high adults 54 years old and older revealed that
levels of loneliness. loneliness was associated with more depres
sive symptoms independent of demographic
factors (age, gender, ethnicity, socioeconom
Cognitive F unctioning and D epression ic status, marital status) and related feelings
of hostility, perceived stress, and poor social
Cognitive Functioning
support (Cacioppo, Hughes, et al., 2006).
Social isolation increases the risk of develop Extending these findings, longitudinal data
ing dementia, and this risk extends to those from a population-based sample of 50- to
who perceive themselves to be socially iso 67-year-old adults in CHASRS showed re
lated or lonely. In a 4-year prospective study ciprocal influences between loneliness and
of initially dementia-free older adults (mean depressive symptoms over a 3-year period
age = 80.7 years), the risk of Alzheimer’s dis that again were independent of demographic
15. L o n e lin e ss 237
terleukin-6 in aging women. P ro ceed in g s o f the and loneliness among the elderly. J o u r n a l o f A p
N a tio n a l A c a d e m y o f S cien c es o f th e USA, 102, p lied G e r o n to lo g y , 8, 110 -13 1.
1 87 5 7 -1 8 7 6 2 . Kiecolt-Glaser, J. K., Ricker, D., George, J ., Messick,
Fromm-Reichman, F. (1959). Loneliness. P sy ch iatry , G., Speicher, C. FI., Garner, W., et al. (1984). Uri
2 2 , 1-15. nary cortisol levels, cellular immunocompetency
Gardner, W. L., Pickett, C. L., Jeffries, V., & Knowles, and loneliness in psychiatric inpatients. P s y c h o s o
M . (20 0 5). On the outside looking in: Loneliness m a tic M ed icin e, 4 6 , 1 5 - 2 3 .
and social monitoring. P erson ality a n d S o c ia l P sy Lau, S., 8c Gruen, G. E. (1992). The social stigma of
c h o lo g y B u lletin , 3 1 , 1 5 4 9 - 1 5 6 0 . loneliness: Effect of target person’s and perceiver’s
Goodwin, R., Cook, O., 8c Yung, Y. (20 01). Loneliness sex. P erson ality a n d S o c ia l P sy ch olog y B u lletin , 18,
and life satisfaction among three cultural groups. 182-189.
P ers o n a l R e la tio n s h ip s, 8, 2 2 5 - 2 3 0 . Lauder, W., Mummery, K., Jones, M., 8c Caperchio-
Hansson, R. O., Briggs, S. R., Sc Rule, B. L. (1990). ne, C. ( 2 0 06 ). A comparison of health behaviours
Old age and unemployment: Predictors of perceived in lonely and non-lonely populations. P sy ch olog y ,
control, depression, and loneliness. J o u r n a l o f A p H ealth , a n d M ed icin e, 11, 2 3 3 - 2 4 5 .
p lie d G e r o n to lo g y , 9, 2 3 0 - 2 4 0 . Lauder, W., Mummery, K., 8c Sharkey, S. ( 2 00 6). So
Hawkley, L. C., Bosch, J. A., Engeland, C. G., Maru- cial capital, age and religiosity in people who are
cha, P. T., 8c Cacioppo, J. T. (200 7 ). Loneliness, lonely . J o u r n a l o f C lin ica l N u rsing, 15, 3 3 4 - 3 3 9 .
dysphoria, stress and immunity: A role for cytok Marangoni, C., 8c Ickes, W. (1989). Loneliness: A
ines. In N. Plotnikoff, R. Faith, A. Murgo, 8c R. theoretical review with implications for measure
Good (Eds.), C y to k in e s: Stress a n d im m u n ity (2nd ment. jo u r n a l o f S o c ia l a n d P erso n a l R e la tio n s h ip s,
ed., pp. 6 7 - 8 5 ) . Boca Raton, FL: C R C Press. 6, 9 3 - 1 2 8 .
Hawkley, L. C., Browne, M. W., 8c Cacioppo, J . T. McGuire, S., 8c Clifford, J. (20 00 ). Genetic and en
(2005). How can I connect with thee?: Let me count vironmental contributions to loneliness in children.
the ways. P sy c h o lo g ic a l S cien ce, 16, 7 9 8 - 8 0 4 . P sy c h o lo g ic a l S cien ce, 11, 4 8 7 - 4 9 1 .
Hawkley, L. C., Burleson, M. H., Berntson, G. G., 8c Moustakas, C. FI. (1961). L o n elin es s. Englewood
Cacioppo, J . T. (2 003). Loneliness in everyday life: Cliffs, NJ: Prentice-Hall.
Cardiovascular activity, psychosocial context, and Nolen-Hoeksema, S., 8c Ahrens, C. (200 2). Age differ
health behaviors. J o u r n a l o f P erson ality a n d S o c ia l ences and similarities in the correlates of depressive
P sy ch olog y , 8 5 , 1 0 5 - 1 2 0 . symptoms. P sy ch o lo g y a n d A ging, 17, 1 1 6 - 1 2 4 .
Hawkley, L. C., 8c Cacioppo, J. T. ( 2 0 07 ). Aging and Ohayon, M. M . (200 5 ). Prevalence and correlates of
loneliness: Downhill quickly? C u rren t D ire ctio n s in non restorative sleep complaints. A rch iv es o f in te r
P s y c h o lo g ic a l S c ien c e, 16, 187 -1 91 . n al M ed icin e, 165, 3 5 - 4 1 .
Hawkley, L. C., Hughes, M. E., Waite, L. J., Masi, C. Olson, K. L., 8c Wong, E. H. (2001). Loneliness in
M ., Thisted, R. A., 8c Cacioppo, J. T. (2008). From marriage. F am ily T h era p y , 2 8 , 1 0 5 - 1 1 2 .
social structure factors to perceptions of relationship Peplau, L. A., 8c Perlman, D. (1982). Perspectives on
quality and loneliness: The Chicago Health, Aging, loneliness. In I.. A. Peplau 8c D. Perlman (Eds.),
and Social Relations Study. J o u r n a l o f G e r o n t o lo y : L o n e lin e s s : A s o u r c e b o o k o f cu rren t th eo ry , r e
S o c ia l S c ien c es, 6 3 B , S 3 7 5 - S 3 8 4 . s e a r ch a n d th era p y (pp. 1-20). New York: Wiley.
Hawkley, I.. C., Masi, C. M ., Berry, J . D., 8c Caciop Pickett, C. L., 8c Gardner, W. L. (2005). The social
po, J. T. (200 6 ). Loneliness is a unique predictor of monitoring system: Enhanced sensitivity to social
age-related differences in systolic blood pressure. cues as an adaptive response to social exclusion.
P sy c h o lo g y a n d A ging, 2 1, 1 5 2 - 1 6 4 . In K. D. Williams, J. P. Forgas, 8c W. von Hippel
Hawkley, L. C., Preacher, K. J., 8c Cacioppo, J. T. (Eds.), T h e s o c ia l o u tc a s t: O stra cism , s o c ia l e x c lu
(20 0 7). Multilevel modeling of social interactions sio n , re jec tio n , a n d b u lly in g (pp. 2 1 3 - 2 2 6 ) . New
and mood in lonely and socially connected individu York: Psychology Press.
als: T he MacArthur Social Neuroscience Studies. In Pickett, C. L., Gardner, W. L., 8c Knowles, M . (2004).
A. D. Ong 8c M . H. M . van Dulmen (Fids.), O x Getting a cue: The need to belong and enhanced
f o r d h a n d b o o k o f m e t h o d s in p o s itiv e p s y c h o lo g y sensitivity to social cues. P erson ality a n d S o c ia l
(pp. 5 5 9 - 5 7 5 ) . New York: Oxford University Press. P sy ch o lo g y B u lletin , 3 0 , 109 5 -11 07 .
Hawkley, L. C., Preacher, K., J ., Waite, L. J ., 8c C a Pinquart, M . (2003). Loneliness in married, widowed,
cioppo, J. T. ( 2 0 07 ). P erce iv ed lo n elin e ss a n d the divorced, and never-married older adults. J o u r n a l
sa lu b r ity o f sle ep . Manuscript submitted for pub o f S o c ia l a n d P ers o n a l R ela tio n s h ip s, 2 0 , 3 1 - 5 3 .
lication. Pinquart, M ., 8c Sorensen, S. (2003). Risk factor
Hawkley, L. C., Thisted, R. A., 8c Cacioppo, J . T. (in for loneliness in adulthood and old age: A meta
press). Loneliness predicts reduced physical activity: analysis. In S. P. Shohov (Ed.), A d v a n ce s in p s y c h o l
Cross-sectional and longitudinal analyses. H ea lth og y resea rch (Vol. 19, pp. 111- 143). Hauppauge,
P sy ch olog y . NY: Nova Science.
Heikkinen, R., 8c Kauppinen, M . (200 4 ). Depressive Pitterman, H., 8c Nowicki, S. ( 2 00 4 ). A test of the
symptoms in late life: A 10-year follow-up. A rch iv es ability to identify emotion in human standing and
o f G e r o n t o lo g y a n d G e ria tric s , 3 8 , 2 3 9 - 2 5 0 . sitting postures: The Diagnostic Analysis of N o n
Hughes, M. E., Waite, L. J . , Hawkley, L. C., 8c C a verbal Accuracy— 2 Posture Test (DANVA2-POS).
cioppo, J . T. (20 04 ). A short scale for measur G e n e tic , S o cia l, a n d G e n e r a l P sy ch o lo g y M o n o
ing loneliness in large surveys: Results from two g ra p h s, 130, 1 4 6 - 1 6 2 .
population-based studies. R e se a rc h on A ging, 2 6 , Pressman, S. D., Cohen, S., Miller, G. E ., Barkin, A.,
655-672. Rabin, B. S., tk Treanor, J. J. (2005). Loneliness, so
Joh nson, D. P., 8c Mullins, L. C. (1989). Religiosity cial network size, and immune response to influenza
240 III. E M O T IO N A L D IS P O S IT IO N S
vaccination in college freshmen. H ea lth P sy ch olog y , skills, and loneliness. In S. Duck 8c D. Perlman
24, 2 97 -30 6 . (Eds.), U n d erstan d in g p e r s o n a l r e la tio n sh ip s: An
Rogers, C. R. (1961). Ellen West— and loneliness. R e in terd iscip lin a ry a p p r o a c h (pp. 1 9 3 - 2 1 9 ) . T ho u
v iew o f E x is ten tia l P sy ch o lo g y a n d P sychiatry, 1, sand O ak s, CA: Sage.
94-101. Solano, C. H. (1987). Loneliness and perceptions of
Rook, K. S. (1984). Promoting social bonding: Strat control: General traits versus specific attributions.
egies for helping the lonely and socially isolated. J o u r n a l o f S o c ia l B e h a v io r a n d P erson ality , 2(2),
A m er ic a n P sy c h o lo g ist, 3 9 , 1 3 8 9 - 1 4 0 7 . 201-214.
Rook, K. S. (1987). Social support versus companion Spiegel, K., Leproult, R., 8c Van Cauter, E. (1999). Im
ship: Effects on life stress, loneliness, and evalua pact of sleep debt on metabolic and endocrine func
tions by others. J o u r n a l o f P erson ality a n d S o c ia l tion. L a n c e t , 3 5 4 , 1 4 3 5 -1 43 9 .
P sy c h o lo g y , 5 2 , 1 1 3 2 -1 1 47 . Steptoe, A., Owen, N., Kunz-Ebrecht, S. R., & Bry-
Rotenberg, K. (1994). Loneliness and interpersonal don, L. (200 4 ). Loneliness and neuroendocrine,
trust. J o u r n a l o f S o c ia l a n d C lin ica l P sy ch olog y , cardiovascular, and inflammatory stress responses
1 3, 1 5 2 - 1 7 3 . in middle-aged men and women. P s y c h o n e u r o e n d o
Rotenberg, K. J . , Gruman, J . A., 8c Ariganello, M. crin o lo g y , 2 9 , 5 9 3 - 6 1 1 .
(2002). Behavioral confirmation of the loneliness Tilvis, R. J ., Kahonen-Vare, M . H., Jolkkonen , J.,
stereotype. B a sic a n d A p p lied S o c ia l P sy ch olog y , Valvanne, J ., Pitkala, K. H., & Strandberg, T. E.
24, 81-89. (2004). Predictors of cognitive decline and m ortal
Rotenberg, K. J ., 8c Kmill, J . (1992). Perception of ity of aged people over a 10-year period. J o u r n a ls
lonely and non-lonely persons as a function of in o f G e r o n t o lo g y S eries A : B io lo g ic a l S c ien ces a n d
dividual differences in loneliness. J o u r n a l o f S o c ia l M e d ic a l S cien ces, 5 9, M 2 6 8 - M 2 7 4 .
a n d P erso n a l R e la tio n s h ip s, 9, 3 2 5 - 3 3 0 . Tucker, M. B., & Mitchell-Kernan, C. (1998). Psy
Russell, D. W. (1996). U CLA Loneliness Scale (Version chological well-being and perceived marital op
3): Reliability, validity, and factor structure. J o u r portunity among single African American, Latina
n al o f P erso n a lity A s sessm en t, 6 6 , 2 0 - 4 0 . and white women. J o u r n a l o f C o m p a ra tiv e F am ily
Russell, D. W., Cutrona, C. E., de la M o ra, A., 8c Wal S tu dies, 2 9 , 5 7 - 7 2 .
lace, R. B. (1997). Loneliness and nursing home ad Viney, L. L. (1985). “They call you a dole bludger”:
mission among rural older adults. P sy ch o lo g y a n d Some experiences of unemployment. J o u r n a l o f
A ging, 12, 5 7 4 - 5 8 9 . C o m m u n ity P sy ch olog y , 13, 3 1 - 4 5 .
Savikko, N., Routasalo, P., Tilvis, R. S., Strandberg, Wallhagen, M . I., Strawbridge, W. J ., Shema, S. J .,
T. E., 8c Pitkala, K. H. (20 05 ). Predictors and sub Kurata, J ., 8c Kaplan, G. A. (2001). Comparative
jective causes of loneliness in an aged population. impact of hearing and vision impairment on subse
A rch iv es o f G e r o n t o lo g y a n d G e ria tric s , 4 1, 2 2 3 - quent functioning. J o u r n a l o f the A m er ic a n G e r ia t
2 33. rics S ociety , 4 9, 1 0 8 6 - 1 0 9 2 .
Schnittker, J . (200 7). Look (closely) at all the lonely Weiss, R. S. (1973). L o n e lin e s s : T h e e x p e r ie n c e o f
people: Age and social psychology of social support. e m o tio n a l a n d s o c ia l is o la tio n . Cambridge, M A :
J o u r n a l o f A gin g a n d H ea lth , 19, 6 5 9 - 6 8 2 . M I T Press.
Shaver, P., 8c Brennan, K. A. (1991). Measures of Wilson, R. S., Krueger, K. R ., Arnold, S. E., Schneider,
depression and loneliness. In J. P. Robinson, P. J . A., Kelly, J . F., Barnes, L. L., et al. (20 07 ). Lon e
R. Shaver, 8c L. S. Wrightsman (Eds.), M easu res liness and risk of Alzheimer’s disease. A rch iv es o f
o f p er s o n a lity a n d s o c ia l p s y c h o lo g ic a l a ttitu d es G e n e ra l P sychiatry, 6 4 , 2 3 4 - 2 4 0 .
(pp. 1 9 5 - 2 8 9 ) . San Diego, CA: Academic Press. Wittenberg, M . T., 8c Reis, H. T. (1986). Loneliness,
Shaver, P., Furman, W., 8c Buhrmester, D. (1985). social skills, and social perception. P erso n a lity a n d
Transition to college: Network changes, social S o c ia l P sy ch o lo g y B u lletin , 12(1), 1 2 1 - 1 3 0 .
CHAPTER 16
j
Affect Intensity
R a n d y J. L a rsen
H istory o f A ffect Intensity days, these two measures of daily mood in
tensity correlated .60 to .77 across persons
Affect intensity refers to individual differ in their samples. Moreover, a mean daily
ences in the typical intensity with which emotional intensity score could be calculated
people experience their emotional respons (by averaging positive and negative intensity
es (Larsen & Diener, 1987). The construct scores) that in turn exhibited high test-retest
also includes affective variability, such that reliability and that correlated in interesting
not only are persons high on affect inten ways with peer reports of emotionality, with
sity more emotionally reactive but also, over parental ratings, and with various other cri
time, their emotional states vary more widely terion variables. It appeared to be a mean
as they react to ongoing life events. The con ingful individual-difference characteristic
struct generalizes over emotions, such that, and one not clearly identified as such in ex
for example, people who experience their isting taxonomies of personality.
positive emotions more strongly will, over Only a few prior studies had examined in
time, generally experience their negative dividual differences in constructs related to
emotions more strongly as well. The charac intensity of emotional response. One impor
teristic highlights that folk notion that “the tant study was that published by Weissman
higher you go up when you are up, the lower and Ricks (1966), which examined the daily
you go down when you are down.” moods of Harvard and Radcliff students
Research on affect intensity began in the using ESM. They identified two aspects of
m id-1980s, when Larsen and colleagues (e.g., individual differences in daily affect; mean
Larsen & Diener, 1985) began conducting mood level over time and mean variability
daily studies of mood and emotion using the over time. A person’s amount of mood vari
experience-sampling method (ESM). When ability, indexed by a within-subject standard
examining global daily mood plotted for in deviation on mood measures over time,
dividuals over several months, they noticed would be a natural consequence of having an
that participants who exhibited wide swings intense emotional response system. A second
upward in positive mood on good days also important paper was written by Underwood
showed wide swings downward in negative and Froming (1980), who were interested in
mood on bad days. In fact, when they cal trait-like characteristics of mood and who
culated the mean positive mood on positive developed a questionnaire measure of mood
days and mean negative mood on negative level and mood reactivity. However, the
241
242 III. E M O T IO N A L D IS P O S IT IO N S
mood reactivity scale was never validated AIM . Incremental validity of the EIS over the
against daily mood measures or laboratory AIM has not been documented, and validity
or field measures of emotional reactivity and evidence for the EIS is sparser than it is for
so remains rarely used or cited. the AIM . Another scale, the Affect Intensity
Larsen and Diener’s (1987) early work Questionnaire (EIQ — Elliot, Sherwin, H ar
assessed emotional intensity using ESM , kins, & Marm arosh, 1995; Harkins, Gram-
calculating affect intensity scores based on ling, & Elliot, 1990) is a visual analog scale
the average distance each participant’s daily with 18 items that asks the participants to
moods deviated from the expected values. rate the relative intensities of distinct affects
Several important observations were made that they experience. This scale seems most
based on these data, including the fact that useful for assessing state, rather than trait,
the frequency with which people experienced affect. Two other drawbacks of the EIQ are
their positive and negative emotions was in that the psychometrics of this measure are
dependent of their intensity (Diener, Larsen, influenced by the ipsatizing effects of the in
Levine, & Emmons, 1985). Affect intensity structions to rate emotions relative to each
also correlated with a cluster of other vari other. In addition, this scale remains unpub
ables, including ratings of the importance of lished.
life events and life goals (Emmons & King, A third measure is the Intensity and Time
1989; Larsen, Diener, & Emmons, 1986). Affect Survey (ITAS— Diener, Fujita, &
However, using ESM to assess affect inten Seidlitz, 1991; Lucas, Diener, & Larsen,
sity has drawbacks. Primary among these is 2 0 0 3 ; Schimmack & Diener, 1997), which
the inordinate amount of time and effort it was developed in tandem with another affect
takes to obtain repeated measures of mood intensity measure called the Scenario Rating
on enough occasions to calculate a reliable Task (SRT; Schimmack & Diener, 1997). The
estimate of mean affect intensity for each ITAS is an adjective-rating task, employing
participant. 24 emotion terms, in which the participant is
asked: “How intensely do you typically ex
perience X , if you experience X ? ” (where X
M easurem ent o f A ffect Intensity is one of the 24 emotions). In examining the
predictive validity correlates of several affect
Because of the need for an efficient and eco intensity measures, the ITAS showed lower
nomical measure of trait affect intensity, validity coefficients than either the A IM or
Larsen (1984) constructed and validated a the SRT (Schimmack & Diener, 1997). The
questionnaire measure called the Affect In SRT presents participants with 20 standard
tensity Measure (AIM). The scale construc ized scenarios and asks them to imagine
tion strategy, including item generation, se being in each of these situations, much like
lection, and refinement into a final 40-item the EIS. However, for each of the SRT sce
measure, is described in Larsen and Diener narios, the participant rates 10 emotions on
(1987). That report also includes prelimi how much of each he or she thinks will be
nary reliability and validity information as evoked by the imaginary scenarios. The SRT
well, some of which I review later. is thus a long and repetitive instrument (re
Since the AIM was originally published quiring 2 0 0 ratings) and is based on respon
(Larsen, 1984), at least four other measures dents’ hypothetical responses to imagined
of affect intensity have been developed. The situations. It does, however, exhibit validity
Emotional Intensity Scale (EIS; Bachorowski correlations that are comparable to the much
& Braaten, 1994) has 30 items that each ask shorter and more economical A IM (Schim
the participant to imagine him- or herself in mack & Diener, 1997). Whereas the SRT has
a specific emotionally evocative situation, not been published, the ITAS is reproduced
then to indicate which of several responses in Lucas and colleagues (2003).
(which vary on intensity) they are most likely Because the predominant measure of af
to have in that scenario. This scale correlates fect intensity remains the A IM , this chapter
.45 with the AIM (Bachorowski & Braaten, focuses primarily on this measure. The AIM
1994) and exhibits a pattern of correlations has been translated into several languages
with third variables that is very similar to the (e.g., German, Spanish, Portuguese, Ital
16. A f f e c t Intensity 243
about the same, the participants with high responses by increasing the perceived im
affect intensity report stronger emotional portance of events. Schimmack and Diener
reactions to those events than participants (1997) demonstrate that affect intensity is
with low affect intensity. correlated with the importance ratings of life
To examine how individuals with high events, and they argue that the attribution
affect intensity come to react so differently of importance to events is a likely cause of
to the same kinds of events compared with affect intensity. Diener, Colvin, Pavot, and
individuals with low affect intensity, Larsen, Allman (1991) also demonstrate, across five
Diener, and Cropanzano (1987) conducted studies, that the importance one attaches to
a thought-sampling study while exposing an event strongly influences the intensity of
participants to emotionally evocative im emotional reactions to that event.
ages. They proposed that affect intensity
would be associated with a distinct pattern
of cognitive operations that would be pres Correlates and Consequences
o f A ffect Intensity
ent while viewing the emotional images. The
theoretical notion was that these cognitive Physiology
operations would lead individuals to inter
pret or construe emotion-provoking stimuli Emotional experience depends in part on
in a manner that intensifies the affective perceived physiological changes. Several
response to those stimuli. Larsen and col researchers have therefore examined affect
leagues found that individuals with high af intensity in relation to perceptions of physi
fect intensity engaged in significantly more ological activity. One interesting study re
personalizing cognition and more general ported by Chwalisz, Diener, and Gallagher
izing cognition than those with low affect (1988) examined affective reactions in per
intensity. Personalizing cognition refers to sons with spinal cord injuries, who have
the tendency to relate to an event by seeing limited perception of their bodily states.
it as self-relevant or focusing on the personal Participants with greater autonomic feed
meanings for oneself. So a person might see back (i.e., lower spinal cord injury) reported
an image of a child wounded in a war and more intense emotions than participants
start thinking about a time when he or she with weaker autonomic feedback. However,
was hurt as a child. Generalizing cognition participants with very high lesions, who had
refers to abstracting from a single event to almost no autonomic feedback, still report
arrive at broad conclusions that are not war ed the experience of emotions, but at a lower
ranted. For example, seeing an image of a intensity level. Such findings suggest that
child wounded in war, a person might start the perception of autonomic arousal may
thinking about how war is horrible and that not be necessary for emotional experience.
human nature at its core is dark and destruc However, increased perception of autonom
tive. People high in affect intensity, relative ic arousal may enhance the felt intensity of
to those low in it, tended to both personalize emotional experience.
and generalize more often, and they did this Blascovich and colleagues (1992) pro
to both positive and negative emotional im vide another perspective on the perception
ages (relative to neutral). These findings were of physiological arousal in relation to trait
replicated in a study by Dritschel and Teas- affect intensity. The authors report three
dale (1991) using a sample of middle-aged separate studies of individual differences
British women. Larsen, Billings, and Cutler in visceral self-perception assessed using
(1996) conceptually replicated these effects a standard heartbeat detection paradigm.
by having participants generate informative Although the AIM was unrelated to actual
descriptions of life events, finding that the cardiac arousal, it was negatively related to
descriptions of participants with high affect perceived cardiac arousal in all three stud
intensity contained significantly more gener ies. These findings suggest that individuals
alizing and more references to arousal and with high affect intensity have relatively
personal feeling states than the descriptions diminished visceral awareness of their own
of participants with low affect intensity. cardiac activity. These results are discussed
The cognitive style of personalizing and in terms of how individuals with high af
generalizing most likely intensifies affective fect intensity may not become aware of their
16. A f f e c t In te n sity 245
emotional reactions until those reactions be line arousal and/or their reactivity to stimu
come quite strong. As such, these individu lation. Consequently, the fourth postulate is
als would require stronger emotional stimu that some individuals will need more stimu
lation before they engaged in self-regulation lation than others to reach their optimal
to dampen their emotional reactions. Lar levels and some will need less stimulation.
sen (2 0 0 0 ; Larsen et al., 1996) presents a The theory predicts individual differences
control-theory model of emotion regula in stimulation-seeking behavior, mainly to
tion, with individual differences in the self compensate for underreactivity and/or lower
perception of physiological arousal playing levels of baseline arousal. This homeostatic
an important role. theory of arousal regulation has existed in
Vanman, Dawson, and Brennan (1998) personality theory in various forms for some
report similar findings of diminished physi time (e.g., Eysenck, 1967; Gale, 1986; Geen,
ological reactivity on the part of partici 1983; Zuckermann, 1979).
pants with high affect intensity. This study M ost of the research on arousal regulation
examined the eyeblink startle reflex to af has focused on two sources of stimulation
fect-laden images. Loud auditory tones were that are sought out to compensate for under
presented quasi-randomly while participants reactivity. One source is behavior: either so
viewed a series of affective images. The stan cializing, heightened activity level, or sensa
dard finding is that, when viewing negative tion seeking. In fact, both Eysenck’s theory
slides, the eyeblink reaction to the auditory of extraversion and Zuckerman’s early the
startle probe tends to be stronger than it is ory of sensation seeking were based on the
to positive or neutral images. However, this notion of individual differences in baseline
eyeblink startle effect was significantly di arousal and the management of arousal level
minished for participants high in affect in through the regulation of behavioral activi
tensity, suggesting that individuals high in ties (Eysenck, 1967; Zuckermann, 1979).
affect intensity are less easily aroused by the Extraverted behavior is seen as an attempt
startle probe. to maximize stimulation input through so
Larsen, Diener, and Emmons (1986) also cial activity in order to compensate for a rel
report negative correlations between affect atively underaroused condition at baseline.
intensity and measures of peripheral physiol Introverts, on the other hand, avoid social
ogy. Both resting galvanic skin response (the stimulation (as well as intense stimulation
number of spontaneous spikes in a 1-minute in general) in order to avoid increasing their
interval) and resting heart rate were found already relatively overaroused condition at
to correlate negatively with the AIM (r - baseline.
-.3 1 and - .2 6 , respectively). These negative A second mechanism of arousal regulation
associations suggest that individuals high is through sensory stimulation. Some indi
in affect intensity, when placed in a quiet, viduals exhibit dampened reactivity to sen
stimulus-reduced environment, are physi sory stimulation. Theories of this individual
ologically less aroused relative to the par difference have variously been called stimu
ticipants low in affect intensity. These find lus intensity modulation theory (Barnes,
ings, and those in the preceding paragraph, 1976; Petrie, 1967), reducer-augmenter
are consistent with basic notions of arousal theory (Herzog, Williams, & Weintraub,
regulation theory, which I now briefly de 1985; Sales, 1971, 1972), and strength of
scribe. the nervous system theory (Pavlov, 1957;
Strelau, 1982, 1985). All refer to the ten
dency of some people to react less strongly
An Arousal Regulation Theory
to sensory stimuli, as, for example, in indi
o f Affect Intensity
vidual differences in pain tolerance. Low-
This theory has a few basic postulates. The reactive persons should be motivated to seek
first is that, for any given task, there exists an out stronger forms of stimulation, whereas
optimal level of arousal for completing the high-reactive persons, those who are more
task; the second is that individuals will seek sensitive, should seek to avoid strong sen
a common optimal level of arousal in a given sory stimulation. Research testing these pre
situation (Hebb, 1955). A third postulate is dictions generally find support in that low-
that individuals differ with respect to base sensory-reactive persons do exhibit a greater
246 III. E M O T IO N A L D IS P O S IT IO N S
need for stimulation (Herzog et al., 1985; lation (85 dB intermittent white noise and
Mishara & Baker, 1981), are bored easily bright flashing lights) on the proofreading
and are motivated to seek out stronger forms performance of participants who scored
of stimulation (Larsen & Baggs, 1986), and high or low on the affect intensity dimension
have higher levels of activity and socializing (Larsen, Zarate, & Dare, 1986). We found
(Petrie, 1967; Sales, 1971) and a tendency that strong sensory stimulation actually im
to abuse illicit stimulant and consciousness- proved the performance of participants high
altering drugs (Kohn, Barnes, &c Hoffman, on the affect intensity dimension, whereas
1979). participants low in affect intensity showed
Larsen (1984; Larsen &c Diener, 1987) a decline in performance when going from
suggested that emotion might be a third normal to high stimulation conditions. In
source of stimulation that could play a role another study participants were asked how
in arousal regulation. If this is true, then in they would perform in a situation while
dividuals with high affect intensity should they were emotionally aroused (e.g., being
display diminished physiological reactivity, angry when having to do homework, feeling
a hypothesis consistent with the findings de nervous while taking a test, feeling jealous
scribed in the previous section. Moreover, if while having to work on a term paper). We
the regular experience of intense emotions found that participants low in affect intensi
is a compensatory strategy for overcoming ty reported that the emotion would interfere
low levels of baseline arousal or diminished with or disrupt their performance, whereas
reactivity, then affect intensity should corre persons high in affect intensity thought that
late with other individual differences related having the emotional stimulation would ac
to arousal regulation, such as extraversion, tually facilitate their performance. Further
sensation seeking, and sensory reducing. research on how emotions can facilitate or
Such correlations have been reported in the impair performance, as well as individual
literature (e.g., Dritschel & Teasdale, 1991; differences in these kinds of effects, is an
Larsen & Diener, 1987; Larsen, Diener, 8c important topic for future research. One in
Emmons, 1986; M aio & Esses, 2 0 0 1 ; Ruch, teresting observation I have made over the
Angleitner, & Strelau, 1991). Also, both years is that persons high on affect intensity,
questionnaire and psychophysical measures while acknowledging that their emotions
of sensory reducing have been found to cor sometimes get them into trouble, neverthe
relate negatively with the AIM (Larsen & less like their intense emotional lifestyle and
Zarate, 1991). The study by Larsen and generally do not want to change.
Zarate (1991) also demonstrated that people
use emotions to compensate for diminished
Emotion Regulation
arousal. In this study we induced boredom
in participants for 35 minutes, then offered Whereas arousal regulation refers to felt
them the choice of participating in an emo levels of energy and activation, emotion
tion manipulation study or a questionnaire regulation refers to self-control attempts to
study. Participants who chose to undergo the modulate hedonic tone or specific emotion
emotion manipulation experience scored sig al reactions. By up-regulating felt arousal
nificantly more in the reducing direction on through strong emotions, persons with high
a measure of sensory reducing-augmenting. affect intensity may appear low on emotion
In a study of desired affect, Rusting and regulation. Moreover, due to its relation to
Larsen (1995) showed that most people desire emotional reactivity and variability, affect
more pleasant and positive emotions, though intensity likely is related to low levels of
affect intensity correlated significantly with emotional control. Several researchers (e.g.,
the desire for stronger felt arousal. The Hunt, 1993; Goldsmith &C Walters, 1989)
arousal regulation theory of affect intensity have found that persons high in affect in
generates a variety of interesting predictions tensity express their emotions more and are
concerning the behavioral and experiential more socially expressive and sensitive (Flett,
implications of emotion-provoking situa Blankstein, Bator, & Pliner, 1989). When
tions for individuals high versus low in af people high in affect intensity engage in
fect intensity. For example, in one study we suppression as a coping style, they are espe
examined the effects of high sensory stimu cially likely to experience distress or depres
16. A f f e c t Intensity 247
sion (Lynch, Robins, Morse, & MorKrause, happen), the net effect on long-term happi
2001). Cheavens and colleagues (2005) have ness is nil.
argued that attempts to suppress emotions
can actually backfire, resulting in stronger
Psychopathology
emotions that are even more difficult to reg
ulate. The connection between affect intensity and
Other researchers have examined beliefs various forms of psychopathology has been
and expectancies about the self-regulation of an active area of research. One disorder re
emotion. For negative emotions, affect inten ceiving much attention is borderline person
sity is associated with the expectation of di ality disorder (BPD), which is characterized,
minished ability to regulate negative moods in part, by extreme emotional instability.
(Flett, Blankstein, &c Obertynski, 1996). Af Bland, Williams, Scharer, and Manning
fect intensity correlates negatively with per (2004) showed that women with BPD scored
ceived emotional self-control, though it is higher on affect intensity, though the effect
unrelated to perceived self-control in other was particularly strong for the Negative
areas of life or to generalized self-control Intensity subscale (consistent with the idea
expectancies (Flett et al., 1989). Research that BPD is related to deficient anger man
suggests that such beliefs in diminished self- agement). A relationship between BPD and
control of emotions are veridical. Eisenberg affect intensity has also been found by other
and Okun (1996) showed that, in stressful researchers (e.g., Yen, Zlotnick, &c Costello,
circumstances, individuals with high nega 2002). Henry and colleagues (2001) provides
tive affect intensity engage in fewer emotion a strong test of this relationship by examin
regulation behaviors and experience more ing affect intensity in BPD compared with
personal distress. An exploratory yet inter other disorders of affect, including bipolar
esting report on rapid eye movement (REM ) disorder. They report that affect intensity is
sleep and affect intensity (Nofzinger et al., elevated in BPD relative to other disorders.
1994) reported a positive correlation be In terms of etiological factors, Rosenthal,
tween affect intensity and the amount and Cheavens, Lejuez, and Lynch (2005) showed
density of R E M sleep patterns. They argue that elevated affect intensity also was relat
that the intense experience of emotions in ed to a (self-reported) history of childhood
the daytime is carried over into sleep, result abuse among persons with BPD.
ing in elevated phasic R E M sleep, which they BPD is also related to self-harm, and at least
see as an indicator of autonomic instability. one study (Gratz, 2006) has shown that, in a
Several researchers have shown that affect nonclinical sample of adult women, the AIM
intensity is unrelated to overall happiness or subscales discriminated women with a his
life satisfaction (e.g., Chamberlain, 1988; tory of self-harming behavior from women
Diener, Colvin, et al., 1991; Larsen & Di with no history of self-harm. In particular,
ener, 1987). Although counterintuitive given high negative affect intensity and low posi
the preceding discussion, there may be sev tive affect intensity distinguished women
eral reasons for this finding. First, the expe high in self-harm (illustrating the utility of
rience of intense emotions may be a compen considering subscales, in addition to the
satory mechanism in providing desired levels total score, when using the AIM). Others
of heightened arousal. Although high affect studies have found elevated affect intensity
intensity comes with the cost of wear and among persons with a history of suicidal
tear on the autonomic nervous system and behavior (lancu et al., 1999). Lynch, Cheav
distress when things do not go well, it may ens, Morse, and Rosenthal (2004) found
satisfy a more basic need to up-regulate felt that, although affect intensity was elevated
arousal. A second reason affect intensity may in persons with a suicidal history, this rela
be unrelated to happiness is that, because tionship was moderated by emotional sup
happiness is the ratio of long-term positive pression, such that persons with high affect
to negative affect (Larsen & Prizmic, 2008) intensity were more likely to be at risk for
and because persons with high affect inten suicide when they also chronically inhibit
sity do have strong positive emotional reac their emotional reactions.
tions when good events happen (along with Flett and Hewitt (1995) took a broad
strong negative reactions when bad events band approach to personality disorders by
248 III. E M O T IO N A L D IS P O S IT IO N S
administering the M illon Clinical Multi- eralizing from events. They also found that
axial Inventory (M illon, 1983), along with this cognitive style was stable over time and
the A IM , in a sample of adult psychiatric consistent across situations and that it oper
patients. Affect intensity was found to cor ated similarly for men and women.
relate positively with indices of BPD, as well A study by Sheldon (1994) found that af
as with passive-aggressive personality, and fect intensity discriminated between art and
negatively with compulsive-conforming per science graduate students, with art students
sonality. Affect intensity also correlated with scoring significantly higher on affect inten
symptom measures of poor adjustment, so sity than science students. Affect intensity
matization, hypomania, alcohol abuse, and was assessed at the start of their training,
psychotic thinking. The authors conclude so it is likely that affect intensity differ
that affect intensity may contribute to a va ences existed prior to exposure to training
riety of forms of psychopathology, primarily in these respective fields. Sheldon suggests
through diminished self-control of emotion that the cognitive style associated with af
and poor inhibition (Flett & Hewitt, 1995). fect intensity lends itself to an interest in
A variety of other forms of psychopathol art more than in science. Moreover, he sug
ogy have also been related to affect intensity. gests that artists and scientists face differ
For example, Day and Wong (1996) found ent social norms regarding the expression
that persons high in psychopathy (or anti of emotion, with artists being encouraged
social character traits) have lower affect in to exaggerate, dwell on, and express their
tensity and exhibit less intense emotional re emotional reactions and scientists encour
actions to everyday life events than persons aged to downplay theirs. His findings sug
low in psychopathy. Also, not surprisingly, gest that individual differences in such tem
affect intensity is associated with being at peramental factors as affect intensity, and
risk for anxiety and panic disorder. At least their associated cognitive styles, may under
one study has shown that persons high in af lie vocational choices.
fect intensity are at risk for substance abuse, Another cognitive style concerns event
most likely in attempts to self-medicate for appraisal. If an event is appraised as very
emotional suppression (Thorberg & Lyvers, important, then affective reactions to the
200 6 ). And finally, as might be imagined, outcome of that event will be more intense
extremely low affect intensity is associated than if the event were viewed as less impor
with alexithymia, a characteristic deficiency tant. Indeed, if you want to know what is
in understanding, processing, or describ important to a person, you might proceed
ing emotions (Iancu et al., 1999; Jacob & by inquiring about the kinds of events that
Hautekeete, 1999; Ritz, 1994). Alexithymia provoke the strongest emotions. Along these
is characterized by difficulty in identifying lines, Emmons and King (1989) reported
and describing feelings, constricted imagi that the importance ratings attached to life
nation and paucity of fantasy, and an exter goals and strivings were associated with
nally oriented cognitive style (Taylor, Bagby, individual differences in affect intensity.
& Parker, 1997). Although not classified as Moreover, individuals high in affect inten
a mental disorder, alexithymia is a trait that sity had more differentiated goals, that is,
places people at risk for developing disor more strivings that were unrelated to each
ders, as well as making people less respon other. Individuals with high affect inten
sive to various psychological treatments. sity want all sorts of things out of life, even
though their goals may be in conflict (e.g.,
to have a high-powered career, a loving and
Cognition and Emotion
committed marriage, lots of interesting hob
Because cognitive and emotional processes bies, and a large family). Moreover, indi
are linked, it is likely that individual differ viduals with high affect intensity had fewer
ences in one are related to, or perhaps even discrete plans for how they might achieve
driven by, individual differences in the other. their goals. In other words, their goal struc
As mentioned earlier, Larsen and colleagues ture was relatively shallow, with many dis
(Larsen et al., 1987, 1996) reported that af crete goals but fewer concrete plans for ways
fect intensity is associated with a cognitive they might realize those goals. Similarly, a
style of personalizing events and overgen study by Dance, Kuiper, and M artin (1990)
16. A f f e c t Intensity 249
demonstrated that affect intensity is associ private self-consciousness and the social-
ated with a higher number of distinct self stimulation facet of affiliation motivation
relevant roles, as assessed in a role-sorting (Blankstein, Flett, Koledin, & Bortolotto,
task. It may be that affect intensity is related 1989), and trait arousability (Mehrabian,
to high self-concept complexity (Linville, 1995). One study examined emotional intel
1985). ligence in relation to affect intensity (Engel-
berg & Sjoberg, 20 0 4 ), wherein the M ayer-
Salovey-Caruso Emotional Intelligence Test
Personality and Demographic Correlates
(M SCEIT; Mayer, Salovey, & Caruso, 2001),
Far and away the personality variables most which has been highly criticized in the litera
frequently found to correlate with affect in ture (e.g., Larsen & Lerner, 2 0 0 6 ), showed
tensity are extraversion and neuroticism (e.g., no correlations with affect intensity or with
Dritschel & Teasdale, 1991; Kardum, 1999; the criterion behavior of accuracy in the as
Larsen & Diener, 1987; M cFatter, 1998). sessment of mood experienced by others.
Both of these personality variables correlate In terms of demographics, a consistent
positively and moderately with affect inten finding is that women score higher than men,
sity. The reason most likely is that extraver at least among young adult samples (Fujita,
sion (E) is related to a disposition to respond Diener, & Sandvik, 1991; Goldsmith &
with stronger positive emotional reactivity Walters, 1989; Seidlitz & Diener, 1998; W il
and neuroticism (N) with a disposition to liams & Barry, 2003). The gender difference
respond with negative emotional reactivity tends to get smaller with age, such that, by
(as found in experimental studies of labora late middle age, men and women are no lon
tory mood induction procedures; see Larsen ger significantly different (Diener, Sandvik,
& Ketelaar, 1989, 1991; Rusting & Larsen, & Larsen, 1985). Although men and women
1997, 1 9 9 8 ,1 9 9 9 ; Zelenski & Larsen, 1999, both decline on affect intensity with age,
2002). If personality space is defined by the women decline faster. Looking at gender
orthogonal dimensions of E and N, then af roles, Jakupcak, Salters, Gratz, and Roemer
fect intensity is a vector that is located half (2003) found that stereotypically masculine
way between them. The incremental validity men report even lower levels of affect inten
of affect intensity over E and N concerns the sity than men with more modern gender-role
focus on affective reactions for these two attributes. The stereotype of women as the
constructs. Whereas the construct defini more emotional gender appears to have a
tion of N has always contained reference kernel of truth, at least when it comes to self-
to affect, particularly anxiety and fear, the report measures of affect intensity among
construct definition of E has not, until very young adult women. The constructive aspect
recently, made much reference at all to the of this gender difference is that women also
affective associates of this trait. Moreover, report more intense positive emotions, such
because E and N are unrelated, the distribu as enthusiasm and joy, compared with men
tion of persons in the two-dimensional space (Fujita et al., 1991).
defined by these constructs is normally dis In terms of age trends, after it peaks in
tributed around any vector passing through adolescence, affect intensity appears to
the origin of the space. This means that the drop with age (Diener, Sandvik, & Larsen,
affect intensity dimension represents, at the 1985). Many others have also shown that
high end, persons who are high on both pos subjective emotional experiences go down
itive and negative emotional reactivity— or, with age, particularly for negative emotions
in other words, persons who have both high (e.g., Carstensen, Pasupathi, Mayr, & Nes-
approach motivation an d high avoidance selroade, 2 0 0 0 ; Gross et al., 1997). Studies
motivation (Larsen & Augustine, 2 0 0 8 ) or of aging and emotion have also examined
are highly sensitive to both cues of reward physiological measures of emotional re
an d cues of punishment (Zelenski &c Larsen, activity, and these studies have also docu
1999). mented decreased reactivity to emotional
Other personality variables have also been stimuli among older adults (e.g., Levenson,
studied in relation to affect intensity, includ Carstensen, Friesen, & Ekman, 1991; Lev
ing self-esteem variability (Oosterwegel, enson, Carstensen, & Gottm an, 1994). A re
Field, H art, & Anderson, 2001), public and cent study by Mather and colleagues (2004)
250 III. E M O T IO N A L D IS P O S IT IO N S
examined amygdala activation during ex agement was related to affective traits, in
posure to positive and negative images, cluding affect intensity, though the effects of
with older participants showing diminished these traits on conflict behaviors were fully
amygdala activation to negative, relative to mediated by state affect on the day of the
positive, stimuli. conflict. Given that other people are a fre
quent source of emotion, understanding the
implications of individual differences in af
Applications o f Research on A ffect Intensity
fect intensity for social relations and within
One applied aspect receiving some attention social organizations is an important topic
concerns individual differences in response for further research.
to advertising appeals. Some advertisers Social justice research often examines
target emotional reactions, whereas oth how people react to the behaviors of others
ers appeal to facts in their advertisements. that are perceived as fair or unfair. Given
Chang (2006) reviews the literature on af that such reactions often contain a strong
fect intensity within consumer research and affective component, van den Bos, M aas,
discusses several mechanisms whereby indi Waldring, and Semin (2003) hypothesized
vidual differences in affect intensity might that affect intensity would be related to an
influence how people respond to advertis exaggerated response to unfairness. In two
ing materials— for example, persons with studies, they found that people high in af
high affect intensity might be more likely fect intensity show strong affective reactions
to elaborate on positive emotional appeals, following the experience of outcome and
more likely to respond to appeals that prom procedural unfairness. Participants with low
ise to relieve negative affect, and so forth. affect intensity exhibited weak to no unfair
M oore, Harris, and Chen (1995) present em ness effects, leading the authors to suggest
pirical data from two experiments showing that, for them, actual fairness may not be an
that participants high, compared with low, important aspect of social justice concerns.
in affect intensity are more responsive to
emotional advertising appeals and showed
no differences in response to nonemotional Conclusions
appeals. In a later study, Moore and Homer
(2000) showed that participants with high Affect intensity is a construct that refers to
affect intensity responded with significantly individual differences in the characteristic
stronger emotions in response to affectively magnitude of emotion reactions. It general
charged advertising appeals and that affect izes to both positive and negative affect, as
intensity predicts arousing lifestyle activity well as to specific emotions. It implies emo
preferences. Moore and Harris (1996) also tional variability over time, as individuals
demonstrated that the effects of emotional react strongly to various hedonic events in
advertising appeals, both positive and nega their lives. Several measures of affect inten
tive, were stronger for participants high in sity have been developed, though the one
affect intensity than for those low in it. They with the most validity evidence and the lon
argue that the relation between affect inten gest research track record is the AIM . The
sity and responding to advertising appeals, A IM exhibits desirable psychometric prop
as well as attitudes toward the ads, are medi erties, has been translated into a number of
ated by emotional responses. languages, and exists in a short form.
Weiss, Nicholas, and Daus (1999) discuss The broad theoretical appeal of the affect
affective variables in organizational behav intensity construct is likely due to several
ior contexts. They report a study of affect things. One is the existence of a sound mea
in the workplace that found that affect in sure with good validity evidence. Another is
tensity predicted heightened variability in the explosion of research on affect and emo
mood on the job, consistent with other stud tion that occurred in the 1990s and early
ies of affect intensity and mood variability. 20 0 0 s. A third reason has to do with using
Rhoades, Arnold, and Jay (2001) examined individual-differences measures to test vari
affective traits during episodes of organi ous theories. For example, if some phenom
zational conflict in an experience sampling enon is theorized to be driven by affect, or
study of business employees. Conflict man if affect is the underlying mechanism, then
16. A f f e c t Intensity 251
individual differences in the phenomenon processes. One process concerns the link on
might be related to individual differences in the right side between organism and response
affect intensity. For example, a researcher and implies that the individual difference is
might theorize that a certain attitude effect in the response magnitude or the response
relies on affect for its impact. If this is true, output side of the equation. Throughout
then individual differences in affective reac most this chapter, I have been treating affect
tivity should predict individual differences intensity as though it were due to this part of
in the attitude effect. As a different example, the formulation. However, another possibil
a researcher might hypothesize that affect ity is that the individual difference is due to
produces a narrowing of attention. If this is the link on the left side, between the stimu
true, then individual differences in affect in lus and the organism. This component refers
tensity should predict individual differences to the stimulus sensitivity, or threshold-for-
in the narrowing of attention. In this way, activation side of the formulation. In a few
affect intensity can be a useful tool for test places in this chapter I have treated affect in
ing broad theories that posit an important tensity as though this process might also be
role for affect in producing some main-effect involved, for example, when talking about
phenomenon. affect intensity as reactivity to life events.
Similarly, if there is a theory about some Distinguishing these component parts of the
causal mechanism involved in affect, then affect system is important for understanding
that mechanism might relate to individual the mechanisms of affect and will also con
differences in affect intensity. For example, tribute to our understanding of the nature of
if personalizing cognitions are thought to affect intensity as an individual difference.
produce stronger affective responses, then
persons with characteristically stronger af
fective responses (i.e., those high in trait af A ckn o w ledgm en t
fect intensity) should display more personal
Preparation of this chapter was supported in part
izing cognitions. If the mechanism is truly by Grant N o . R 0 1 - A G 0 2 8 4 1 9 from the National
causal, then manipulating the mechanism Institute on Aging.
should diminish affect intensity such that a
person high in affect intensity would begin
to react more like a person low in affect in R eferences
tensity. The idea of testing general theories
with individual-difference measures is an Bachorowski, J ., & Braaten, E. B. (1994). Emotional
intensity: Measurement and theoretical implica
interesting and effective application of per tions. P erson ality a n d In d iv id u a l D iffe r e n c e s , 27,
sonality psychology to the broader questions 19 1 -1 9 9 .
of psychology in general. Barnes, G. E. (1976). Individual differences in percep
A final question about the nature of indi tual reactance: A review of the stimulus intensity
modulation individual difference dimension. C a n a
vidual differences in affect is implicit in the
d ian P sy c h o lo g ic a l R ev iew , 17, 2 9 - 5 2 .
material covered earlier. The question con Bland, A. R., Williams, C. A., Scharer, K., & M a n
cerns the locus and interpretation of individ ning, S. (200 4). Emotion processing in borderline
ual differences in affect intensity. M ost ex personality disorders. Issu es in M en tal H ea lth
perimental studies of affect intensity involve N ursing, 2.S, 6 5 5 - 6 7 2 .
Blankstein, K. R., Elett, G. L., Koledin, S., & Borto-
the manipulation or measurement of some lotto, R. (1989). Affect intensity and dimensions of
stimulus, typically a mood induction or the affiliation motivation. P erson ality a n d In d iv id u a l
hedonic value of some life event. Then emo D iffe r e n c e s , 10, 1 2 0 1 - 1 2 0 3 .
tional responses are assessed and examined Blascovich, J ., Adlin, R., Brennan, K., Coad, M . L..,
Hughes, P., Kelsey, R. M ., et al. (1992). Affect in
for predictable individual differences. This
tensity and cardiac arousal. J o u r n a l o f P erson ality
can be displayed in the typical stimulus- a n d S o c ia l P sy ch olog y , 6 3 , 1 6 4 -17 4.
organism-response model: Bryant, F. B., Yarnold, P. R., & Grimm, L. G. (1996).
Toward a measurement model of the Affect Inten
S —> O —> R sity Measure: A three-factor structure. J o u r n a l o f
R esea rch in P erson ality , 3 0 , 2 2 3 - 2 4 7 .
Carstensen, L. L., Pasupathi, M ., Mayr, U., & Nessel-
This simple formulation suggests that the roade, J. R. ( 2 00 0). Emotional experience in every
locus of individual differences in affective day life across the adult life span .J o u r n a l o f P ers o n
response could originate from two different a lity a n d S o c ia l P sy ch olog y , 79, 6 4 4 - 6 5 5 .
252 III. E M O T IO N A L D IS P O S IT IO N S
Chamberlain, K. (1988). On the structure of subjec Flett, G. L., & Hewitt, P. L. (1995). Criterion validity
tive well-being. S o c ia l In d ic a to r s R e se a rc h , 2 0 , and psychometric properties of the Affect Intensity
581-604. Measure in a psychiatric sample. P erson ality a n d
Chang, C. (200 6). Context-induced and ad-induced In d iv id u a l D iffe r e n c e s , 19, 5 8 5 - 5 9 1 .
affect: Individual differences as moderators. P sy Fujita, F., Diener, E., & Sandvik, E. (1991). Gender
c h o lo g y a n d M arketin g , 2 3 , 7 5 7 - 7 8 2 . differences in negative affect and well-being: The
Cheavens, J. S., Daughters, S. B., Kosson, D., Lejuez, case for emotional intensity. J o u r n a l o f P erson ality
C. W., Lynch, T. R., Nowak, J ., et al. (2005). An a n d S o c ia l P sy ch olog y , 61, 4 2 7 - 4 3 4 .
analogue investigation of the relationships among Gale, A. (1986). Extraversion-introversion and spon
perceived parental criticism, negative affect, and taneous rhythms of the brain: Retrospect and pros
borderline personality disorder features the role of pect. In J . Strelau, F. Fareley, & A. Gale (Eds.),
thought suppression. B e h a v io u r R e se a rch a n d T h e r T h e b io lo g ic a l b a se s o f p e r s o n a lity a n d b e h a v io r
ap y , 4 3 , 2 5 7 - 2 6 8 . (pp. 2 5 - 4 2 ) . Washington, DC: Hemisphere.
Chwalisz K., Diener, E., & Gallagher, D. (1988). Auto Geen, R. G. (1983). T he psychophysiology of
nomic arousal feedback and emotional experience: extraversion-introversion. In J. T. Cacioppo & R.
Evidence from the spinal cord injured. J o u r n a l o f E. Petty (Eds.), S o c ia l p sy c h o p h y s io lo g y (pp. 3 9 1 -
P erso n a lity a n d S o c ia l P sy ch o lo g y , 5 4, 8 2 0 - 8 2 8 . 416). New York: Guilford Press.
Dance, K., Kuiper, N. A., & Martin, R. (1990). Inten Geuens, M., & de Pelsmacker, P. (2 002). Developing a
sity of affect, role self-concept, and self-evaluative short Affect Intensity Scale. P sy c h o lo g ic a l R e p o r ts ,
judgments. P sy c h o lo g ic a l R e p o r ts , 67, 3 4 7 - 3 5 0 . 9 1, 6 5 7 - 6 7 0 .
Day, R., & Wong, S. (1996). Anomalous perceptual Goldsmith, R. E., & Walters, H. (1989). A validity
asymmetries for negative emotional stimuli in the study of the Affect Intensity Measure. J o u r n a l o f
psychopath. J o u r n a l o f A b n o r m a l P sy ch olog y , 105, S o c ia l B eh a v io r a n d P erson ality , 4, 1 3 3 - 1 4 0 .
648-652. Gratz, K. L. (200 6 ). Risk factors for deliberate self-
Diener, E., Colvin, C. R., Pavot, W. G., &c Allman, A. harm among female college students: The role and
(1991). The psychic costs of intense positive affect. interaction of childhood maltreatment, emotional
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 61, inexpressivity, and affect intensity/reactivity. A m er
492-503. ican J o u r n a l o f O r th o p s y c h ia tr y , 76, 2 3 8 - 2 5 0 .
Diener, E., Fujita, F., & Seidlitz, L. (1991). M an u al Gross, J. J ., Carstensen, L. L ., Pasupathi, M ., Tsai, J.,
f o r th e In ten sity a n d T im e A ffe c t S u rvey (IT A S). Skorpen, C. G., & Hsu, A. Y. C. (1997). Emotion
Unpublished manuscript, University of Illinois at and aging: Experience, expression, and control.
Urbana-Champaign. P sy ch o lo g y a n d A ging, 12, 5 9 0 - 5 9 9 .
Diener, E., Larsen, R. J., Levine, S., & Emmons, R. A. Harkins, S. W., Gramling, S., & Elliott, T. (1990). T h e
(1985). Intensity and frequency: Dimensions under A ffe c t In ten sity Q u es tio n n a ir e. Unpublished manu
lying positive and negative affect. J o u r n a l o f P er script, Virginia Commonwealth University.
so n a lity a n d S o c ia l P sy ch olog y , 4 8 , 1 2 5 3 - 1 2 6 5 . Hebb, D. O. (1955). Drives and the CN S (conceptual
Diener, E., Sandvik, E., & Larsen, R. J . (1985). Age nervous system). P sy c h o lo g ic a l R ev iew , 6 2 , 2 4 3 -
and sex effects for emotional intensity. D e v e lo p 2 54.
m e n ta l P sy c h o lo g y , 2 1 , 5 4 2 - 5 4 6 . Henry, C., Mitropoulou, V., New, A. S., Koenigsberg,
Dritschel, B. H., & Teasdale, J. D. (1991). Individual H. W., Silverman, J ., & Siever, L. J. (2001). Affec
differences in affect-related cognitive operations tive instability and impulsivity in borderline per
elicited by experimental stimuli. B ritish J o u r n a l o f sonality and bipolar II disorders: Similarities and
C lin ic a l P sy ch olog y , 3 0 , 151 -1 60 . differences. J o u r n a l o f P sy ch iatric R e se a rch , 3 5 ,
Eisenberg, N., Sc Okun, M. S. (1996). The relations of 307-312.
dispositional regulation and emotionality to elders’ Herzog, T., Williams, D. M ., 8c Weintraub, D. J.
empathy-related responding and affect while volun (1985). Meanwhile, back at personality ranch: The
teering. J o u r n a l o f P erson ality , 6 4 , 1 5 7 -1 83 . augmenters and reducers ride again. J o u r n a l o f P er
Elliott, T. R., Sherwin, E., Harkins, S. W., & Mar- s o n a lity a n d S o c ia l P sy ch olog y , 4 8 , 1 3 4 2 - 1 3 5 2 .
marosh, C. (1995). Self-appraised problem-solving Hunt, M . G. (1993). Expressiveness does predict well
ability, affective states, and psychological distress. being. S ex R o les, 2 9 , 147-169.
J o u r n a l o f C o u n selin g P sy ch olog y , 4 2 , 1 0 5 - 1 1 5 . lancu, I., Horesh, N., Offer, D., Dannon, P. N., Lep-
Emmons, R. A., & King, L. A. (1989). Personal striving kifker, E., & Kotler, M . (1999). Alexithymia, affect
differentiation and affective reactivity. J o u r n a l o f intensity and emotional range in suicidal patients.
P erson ality a n d S o c ia l P sy ch o lo g y , 5 6 , 4 7 8 - 4 8 4 . P sy ch o th er a p y a n d P sy c h o s o m a tic s , 6 8 , 2 7 6 - 2 8 0 .
Engelberg, E., & Sjoberg, L. (200 4 ). Emotional intel Jaco b, S., & Flautekeete, M . (1999). Alexithymia is
ligence, affect intensity, and social adjustment. P er associated with a low self-estimated affective in
s o n a lity a n d In d iv id u a l D iffe r e n c e s , 37, 5 3 3 - 5 4 2 . tensity. P erson ality a n d In d iv id u a l D iffer en c es , 27,
Eysenck, H. J. (1967). T h e b io lo g ic a l b a sis o f p e r s o n 125-133.
ality . Springfield, IL: Thomas. Ja kupca k, M ., Salters, K., Gratz, K. L., & Roemer, L.
Flett, G. L., Blankstein, K. R., & Obertynski, M. (2003). Masculinity and emotionality: An investi
(199 6). Affect intensity, coping styles, mood regula gation of men’s primary and secondary emotional
tion expectancies and depressive symptoms. P ers o n responding. S ex R o le s, 4 9, 1 1 1 - 1 2 0 .
a lity a n d In d iv id u a l D iffe r e n c e s , 2 0 , 2 2 1 - 2 2 8 . Kardum, I. (1999). Affect intensity and frequency:
Flett, G. L., Blankstein, K. R., Bator, C., & Pliner, P. Their relation to mean level and variability of posi
(1989). Affect intensity and self-control of em o tive and negative affect and Eysenck’s personality
tional behaviour. P erson ality a n d In d iv id u a l D if traits. P erso n a lity a n d In d iv id u a l D iffe r e n c e s , 2 6 ,
fe r e n c e s , 10, 1-5. 33-47.
16. A f f e c t In te n sity 253
Kohn, P. M ., Barnes, G. E., & Hoffm an, F. M . (1979). In d iv id u a l d iffe r e n c e s in e m o tio n a l re a ctiv ity a n d
Drug-use history and experience seeking among p e r fo r m a n c e u n d er stress. Paper presented at the
adult male correctional inmates, j o u r n a l o f C o n annual meeting of the Midwestern Psychological
su ltin g a n d C lin ic a l P sy ch olog y , 47, 7 0 8 - 7 1 5 . Association, Chicago.
Larsen, R. J. (1984). Theory and measurement of affect Levenson, R. W., Carstensen, L. I.., Friesen, W. V., &
intensity as an individual difference characteristic. Ekm an, P. (1991). Emotion, physiology, and expres
D iss er ta tio n A b stra cts In te rn a tio n a l, 8 5 , 2 2 9 7 B . sion in old age. P sy ch o lo g y a n d A ging, 6, 2 8 - 3 5 .
(UMI No. 8 4-22 11 2). Levenson, R. W., Carstensen, L. L., &c Gottman, J. M.
Larsen, R. J. (1987). The stability of mood variability: (1994). The influence of age and gender on affect,
A spectral analytic approach to daily mood assess physiology, and their interrelations: A study of long
ments. jo u r n a l o f P erson ality a n d S o c ia l P sy c h o l term marriages. J o u r n a l o f P erson ality a n d S o cia l
og y , 5 2 , 1 1 9 5 - 1 2 0 4 . P sy ch olog y , 67, 5 6 - 6 8 .
Larsen, R. J. (20 00 ). Toward a science of mood regula Linville, P. W. (1985). Self-complexity and affective
tion. P s y c h o lo g ic a l In q u iry , J 1, 1 2 9 - 1 4 1 . extremity: D o n ’t put all of your eggs in one cogni
Larsen, R. J ., & Augustine, A. A. (200 8). Basic person tive basket. S o c ia l C o g n itio n , 3(1), 9 4 - 1 2 0 .
ality dispositions related to approach and avoidance: Lucas, R. E., Diener, E., & Larsen, R. J. (2003). M ea
Extraversion/neuroticism, BIS/BAS, and positive/ suring positive emotions. In S. J . Lopez & C. R.
negative affect. In A. J . Elliot (Ed.), H a n d b o o k o f Snyder (Eds.), P ositiv e p s y c h o lo g ic a l a s se ss m e n t: A
a p p r o a c h a n d a v o id a n c e m o tiv a tio n (pp. 1 5 1-1 64 ). h a n d b o o k o f m o d e ls a n d m ea su res (pp. 2 0 1 - 2 1 8 ).
M ah w ah, NJ: Erlbaum. Washington, DC: American Psychological Associa
Larsen, R. J., & Baggs, D. W. (1986). Some psy tion.
chophysical and personality correlates of the Strelau Lynch, T. R., Cheavens, J . S., Morse, J. Q., & Rosen
Temperament Inventory. P erson ality a n d In d iv id u a l thal, M. Z. (200 4 ). A model predicting suicidal
D iffe r e n c e s , 7, 5 6 1 - 5 6 5 . ideation and hopelessness in depressed older adults:
Larsen, R. J., Billings, D. W., & Cutler, S. E. (1996). T he impact of emotion inhibition and affect inten
Affect intensity and individual differences in infor sity. A gin g a n d M e n ta l H ea lth , 8, 4 8 6 - 4 9 7 .
mational style, jo u r n a l o f P erson ality , 6 4 , 185— Lynch, T. R., Robins, C. J ., Morse, J . Q., & MorKrause,
207. F.. D. (2001). A mediational model relating affect
Larsen, R. J ., & Diener, E. (1985). A multitrait- intensity, emotion inhibition, and psychological dis
multimethod examination of affect structure: He tress. B e h a v io r T h era p y , 3 2 , 5 1 9 - 5 3 6 .
donic level and emotional intensity. P erson ality a n d Maio, G. R., & Esses, V. M . (2001). The need for af
In d iv id u a l D iffer en c es , 6, 6 3 1 - 6 3 6 . fect: Individual differences in the motivation to ap
Larsen, R. J . , & Diener, E. (1987). Affect intensity as proach or avoid emotions. J o u r n a l o f P erson ality ,
an individual difference characteristic: A review. 69, 5 8 3 - 6 1 5 .
J o u r n a l o f R e se a rc h in P erson ality , 2 1, 1-39. Mather, M ., Canli, T., English, T., Whitfield, S. L.,
Larsen, R. J ., Diener, E., & Cropanzano, R. S. (1987). Wais, P., Ochsner, K. N., et al. (200 4). Amygdala
Cognitive operations associated with individual dif activity in response to emotional pictures in older
ferences in affect intensity, j o u r n a l o f P erson ality adults. P sy c h o lo g ic a l S cien ce, 15, 2 5 9 - 2 6 3 .
a n d S o c ia l P sy ch olog y , 5 3 , 7 6 7 - 7 7 4 . Mayer, J. D., Salovey, P., & Caruso, D. R. (2001).
Larsen, R. J ., Diener, E., & Emmons, R. A. (1986). T ec h n ic a l m a n u a l f o r th e M a y e r - S a lo v e y - C a r u s o
Affect intensity and reactions to daily life events. E m o tio n a l In tellig en ce T est V .2.0. Toronto, O n
Jo u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 51, tario, Canada: M H S.
'803-814. McFatter, R. M. (1998). Emotional intensity: Some
Larsen, R. J., & Ketelaar, T. (1989). F’ xtraversion, neu components and their relations to extraversion and
roticism, and susceptibility to positive and negative neuroticism. P erson ality a n d In d iv id u a l D iffe r e n c
mood induction procedures. P erson ality a n d In d i es, 2 4 , 7 4 7 -75 8 .
v id u a l D iffe r e n c e s , 10, 1 2 2 1 - 1 2 2 8 . Mehrabian, A. (1995). Theory and evidence bearing
Larsen, R. J., & Ketelaar, T. (1991). Personality and on a scale of trait arousability. C u rren t P sy ch o lo g y :
susceptibility to positive and negative emotional D e v e lo p m e n ta l, L ea rn in g , P erson ality , S o c ia l, 14,
states. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l 3-28.
og y , 61, 1 3 2 - 1 4 0 . Millon, T. (1983). T h e M illon C lin ica l M u ltia x ia l In
Larsen, R. J ., &c Lerner, C. ( 2 0 06 ). Emotional intel v en tory m a n u a l (3rd ed.). Minneapolis, M N : Na
ligence and mood regulation following the attack of tional Computer Systems.
September 11. In A. Delle Fave (Hd.), D im en s io n s Mishara, B. L., & Baker, A. H. (1981). Individual dif
o f w e ll-b ein g : R e se a rc h a n d in terv en tio n (pp. 4 8 9 - ferences in stimulus intensity modulation in the el
511). M ilano, Italy: FrancoAngeli. derly. In te r n a tio n a l J o u r n a l o f A gin g a n d H u m an
Larsen, R. J ., & Prizmic, Z. (2008). Regulation of D e v e lo p m e n t, 13, 2 8 5 - 2 9 5 .
emotional well-being: Overcoming the hedonic Moore, D. J., & Harris, W. D. (1996). Affect inten
treadmill. In M. Eid & R. J . Larsen (Eds.), T h e s c i sity and the consumer’s attitude toward high impact
e n c e o f su b je c tiv e w e ll-b e in g (pp. 2 5 8 - 2 8 9 ) . New emotional advertising appeals .J o u r n a l o f A d v e rtis
York: Guilford Press. ing, 2 5 , 3 7 - 5 0 .
Larsen, R. J., & Zarate, M . A. (1991). Extending re- Mo ore, D. J., Harris, W. D., & Chen, H. C. (1995).
ducer/augmenter theory into the emotion domain: Affect intensity: An individual difference response
The role of affect in regulating stimulation level. to advertising appeals. J o u r n a l o f C o n s u m e r R e
P erso n a lity a n d In d iv id u a l D iffer en c es , 12, 7 1 3 - sea rch , 2 2 , 1 5 4 - 1 6 4 .
72 3 . Moore, D. J., & Homer, P. M . ( 2 0 0 0). Dimensions of
Larsen, R. J ., Zarate, M. A., & Dare, T. (1 98 6, May). temperament: Affect intensity and consumer life
254 III. E M O T IO N A L D IS P O S IT IO N S
C o g n it iv e D is p o s it io n s
CHAPTER 17
Openness to Experience
R o bert K . M cC rae
A n g e l in a R . S u t i n
257
258 IV. C O G N IT IV E D IS P O S IT IO N S
sensitivity, are still not represented by single time in the 20s and then gradually declining
trait adjectives in English. Lay conceptions (e.g., McCrae et al., 2005a).
of Openness are often confounded with in It is useful to distinguish Openness from
terpersonal openness (Sneed, M cCrae, & constructs with which it might be confused,
Funder, 1998). It is therefore understandable particularly intelligence.2 Although adjec
that there are different conceptualizations of tive Intellect scales include such terms as
Openness among experts (De Raad & Van perceptive, analytical , and intelligent, and
Heck, 1994). although they correlate well with Openness,
In this chapter we adopt the view of Open the association of Openness with measured
ness operationalized in the Revised NEO intelligence is modest and specific. Correla
Personality Inventory (N EO -PI-R; Costa tions around .40 are found with measures of
& M cCrae, 1992a), but in general there divergent thinking, which is often thought to
are substantial correlations among different underlie creativity (M cCrae, 1987). Open
measures of Openness, including the Open ness scores were associated (rs ~ .30) with
ness scale of the Big Five Inventory (BFI; performance on verbal and facial emotion
Benet-Martfnez & John, 1998), and Gold recognition tasks for both Caucasians and
berg’s (1990) adjective-based Intellect scales. African Americans (Terracciano, M erritt,
(However, the fifth factor in the Five-Factor Zonderman, &C Evans, 2003). Noftle and
Personality Inventory [Hendriks, Hofstee, Robins (2007) reported an overall correla
& De Raad, 1999] is called Autonomy and tion of .26 between Openness and the verbal
is only modestly related to Openness; De score on the Scholastic Aptitude Test, but
Fruyt, M cCrae, Szirmak, & Nagy, 2004.) only .05 with the math score. Higher verbal
The N EO -PI-R has facet scales for Open scores may reflect more and broader read
ness to Fantasy, Aesthetics, Feelings, Ac ing among students high in Openness rather
tions, Ideas, and Values. Highly open people than greater native ability.
are thus seen as imaginative, sensitive to Finally, it will be useful to discuss the
art and beauty, emotionally differentiated, relation of Openness to some of the other
behaviorally flexible, intellectually curi constructs discussed in this book. Open
ous, and liberal in values. Closed people are ness is inversely, and rather strongly, related
down-to-earth, uninterested in art, shallow to Authoritarianism/Dogmatism: Trapnell
in affect, set in their ways, lacking curiosity, (1994) reported correlations of from - .2 9
and traditional in values.1 M ost psycholo to -.6 3 between N EO -PI-R Openness facet
gists would judge the high pole of this di scales and Right Wing Authoritarianism,
mension to be desirable, because most psy with the largest correlation unsurprising
chologists are themselves high in Openness ly with Openness to Values. To the extent
(Staudinger, M aciel, Smith, & Bakes, 1998), that aggression is related to authoritarian
but among laypeople there is a strong cor ism (weakly; see Carnahan & M cFarland,
relation between their social desirability rat 2 0 0 7 ), we would expect authoritarians to be
ings of Openness and their own self-reports antagonistic as well as closed.
(Konstabel, 2 0 0 7 ): Open people admire Need for Closure (Webster & Kruglanski,
openness, closed people despise it. 1994), the desire for definite and final an
Like the other basic factors, Openness is swers, is also related to low Openness (r -
strongly heritable, and the covariation of - .4 2 , N - 84, p < .001; Costa & M cCrae,
Openness facets to define the factor appears 1998) but is unrelated to Agreeableness (r =
at the genetic level as well as the phenotypic - .0 8 , n.s.). Instead, this construct includes a
level (Yamagata et al., 2 0 0 6 )— that is, peo preference for order and predictability that
ple who are intellectually curious also tend gives it an association with Conscientious
to be imaginative and artistically sensitive in ness (r = .42, p < .001). Thus people prone
part because the same genes help shape these to seizing on the first idea offered and then
three traits. Like the other basic factors, freezing on this solution (Kruglanski &
Openness shows high levels of differential Webster, 1996) are in general uninterested
stability across the adult lifespan (Terraccia- in exploring alternative possibilities, keep
no, Costa, &C M cCrae, 2 0 0 6 ), but it shows ing their views simple and uncluttered.
a distinctive pattern of maturational trends, Other people pursue ideas vigorously,
increasing from early adolescence until some being high on both Openness and Consci
17. O penness to E x p e rie n ce 259
entiousness. Such people score high on Need We do not mean to suggest that these
for Cognition (Cacioppo & Petty, 1982; Sad- constructs are equivalent to O; they differ
owski & Cogburn, 1997; P. D. Trapnell, per both in their associations with other factors
sonal communication, November 9, 2007). and in their specific content that gives each
Need for Cognition is most directly relevant a unique focus of convenience. However, if
to 0 5 : Ideas, ’ but it is related to most facets measures of all of them were factored to
of Openness (Berzonsky &C Sullivan, 1992). gether, it is likely that a first general factor
Remarkably, a PsycINFO search found 474 would be defined chiefly by Openness. The
entries for “Need for Cognition” and 1,032 social consequences of Openness, to some
for “Openness to Experience,” but only 6 degree, include the social consequences of
that included both terms. The Need for Cog Authoritarianism, Need for Closure, and so
nition scale was created by social psycholo on.
gists and has been used widely in experimen
tal studies, whereas Openness is employed
in correlational studies in the personality Individual Social Interactions
literature. Petty, Brinol, Loersch, and Mc-
Person Presentation and Perception
Caslin (Chapter 21, this volume) should
give readers an idea of how Openness might Do open people express their Openness in
function if it were included as a moderator ways that other people can detect? Are oth
variable in social-psychological experiments. ers able to recognize these cues accurately,
For example, research by D ’Agostino and or do lay observers have intuitive ideas
Fincher-Kiefer (1992) suggests that highly about what behaviors reflect Openness that
open people would be less susceptible to the may not be diagnostic of the individual’s ac
correspondence bias, that is, to misattribute tual level of Openness? Can multiple observ
behavior to dispositional rather than situ ers come to consensus on whether another is
ational causes open? And are they accurate? The person-
Tetlock, Peterson, and Berry (1993) re perception literature addresses each of these
ported that Integrative Complexity (a form questions and paints a broad picture of how
of cognitive complexity in which people tend Openness is manifested in daily living and
to consider a range of possibilities before interpersonal interactions and how others
coming to a conclusion) showed positive as perceive these cues.
sociations with Myers-Briggs Type Indica Open individuals express their creativity,
tor Intuition, Adjective Check List Creative intellectual curiosity, and need for variety in
Personality, and California Psychological characteristic ways across a variety of me
Inventory Flexibility— all known correlates diums. They are verbally fluent, humorous,
of Openness (McCrae & Costa, 1997). and expressive in interpersonal interactions
Kensinger (1996) scored Thought Com (Sneed, M cCrae, &t Funder, 1998). When
plexity from definitions given in response going about their daily lives, these individu
to 11 words (see Kreitler & Kreitler, 1990) als use fewer third-person pronouns and
and found that it was associated with total past-tense verbs and spend more of their
Openness (r = .36, N = 60, p < .05) and with time in restaurants, bars, and coffee shops
0 2 : Aesthetics (r = .30) and especially 0 5 : (Mehl, Gosling, & Pennebaker, 2006).
Ideas (r = .51, p < .01). Given that open individuals have both artis
Given the association of Openness with tic and intellectual proclivities, it is not sur
emotion recognition (Terracciano et al., prising that these interests are expressed in
2003), one might guess that it would also how they present themselves to the world.
be related to emotional intelligence, and For example, on their personal Web pages,
there are some data supporting a mod open individuals choose to highlight their
est association (Schulte, Ree, & Carretta, own creative and work projects and present
2004). Finally, one of the variables classi information that expresses their emotions
fied as a motivational disposition, Sensa and personal opinions (Marcus, Machilek,
tion Seeking, has an Experience Seeking & Schiitz, 2006). These same proclivities are
subscale that is clearly related to Openness manifested in their working and living spac
(Zuckerman, Kuhlman, Joirem an, Teta, & es. Their love of novelty and originality is
K raft, 1993). evident here: Open individuals decorate both
260 IV. C O G N IT IV E D IS P O S IT IO N S
their offices and bedrooms in distinctive and Openness coupled with more reliable mea
unconventional ways, and, consistent with sures, has found considerable consensus.
their intellectual interests, own and display This is true across a variety of sources of
varied books and magazines (Gosling, Ko, zero-acquaintance information: Observers
Mannarelli, & M orris, 2002). agree on Openness when judging personal
Observers are fairly good at picking up websites (Vazire & Gosling, 2004), top-10
on these behavioral indicators of Openness. song lists (Rentfrow & Gosling, 20 0 6 ), and
For example, perceivers judge individuals offices and bedrooms (Gosling et al., 2002).
who speak fluently, initiate humor, and are Compared with the other traits in the FFM ,
expressive to be high on Openness (Sneed et Openness and Extraversion typically show
al., 1998). Individuals who use fewer past- similar levels of consensus, and consensus
tense verbs and who frequent restaurants, on both remains high as acquaintanceship
bars, and coffee shops are perceived as being increases (Borkenau, Mauer, Riemann,
open (Mehl et al., 2 0 0 6 ), as are individu Spinath, & Angleitner, 2004). A slightly dif
als with websites that have links to work/ ferent pattern emerges for virtual acquain
personal projects and that express personal tanceships. In chatrooms, there is moder
opinions (Marcus et al., 20 0 6 ). Likewise, ate consensus on Openness for one-on-one
perceivers use the distinctiveness of both chats— albeit lower than consensus on Ex
office space and bedrooms to judge the in traversion and Agreeableness— but this con
habitant’s level of Openness (Gosling et al., sensus disappears when chatting in a group
2002). Observers appear relatively adept at rather than one on one (Markey & Wells,
recognizing many behavioral cues diagnos 2002). Although there were no differences in
tic of Openness. the amount of text written in the two con
Yet lay perceivers also have their own ditions, consensus may have decreased be
ideas about what behaviors are indicative of cause the content of the text may have been
Openness that are not necessarily diagnos more superficial during group interactions
tic; that is, lay conceptions can be inaccu and thus less diagnostic.
rate. For example, observers judge individu Across these varied contexts, consensus
als who have highly decorated, cheerful, and among observers tends to be higher than ac
colorful offices to be open, whereas these curacy: Others can agree on whether they
office characteristics are largely unrelated believe a person is open, but they may not
to the individual’s actual level of Openness be right (perhaps because shared lay concep
(Gosling et al., 2002). Likewise, using big tions of the cues of Openness are not always
words in everyday speech is perceived to be correct). Accuracy also depends on the task
a sign of Openness, when in fact Openness observed; some tasks are more diagnostic of
is unrelated to this speech characteristic. Openness than others. Open individuals are
On personal Web pages, perceivers judge imaginative and creative people, and observ
individuals who post many pictures and re ers are more accurate when judging Open
veal much personal information to be open ness from tasks that allow these qualities to
(Marcus et al., 2 0 0 6 ), and in chatrooms, the be expressed rather than from highly struc
number of topics discussed and number of tured tasks (Borkenau et al., 2004).
self-deprecating remarks are taken as signs Finally, perceivers in laboratory studies
of Openness, whereas Openness is unrelated form an impression of Openness very quick
to these behaviors (Rouse & Haas, 2003). ly that is resistant to change. From observing
This discrepancy, of course, raises the as little as 5 seconds of a getting-to-know-
question of how accurately others can infer you conversation, perceivers can make attri
Openness. Multiple judges do agree with butions about Openness. Although accuracy
each other on the individual’s level of Open ratings are generally lower for Openness
ness, which suggests that lay conceptions of than for the other traits in this context, ac
Openness are not idiosyncratic. Although curacy does not vary as a function of slice
early research addressing this question length— it takes a very narrow sliver of time
found little consensus among observers at for a perceiver to form a judgment of Open
zero acquaintance (Kenny, Albright, Malloy, ness (Carney, Colvin, & H all, 2007). And
& Kashy, 1994), more recent research, per once this impression is formed, it is not eas
haps because of better conceptualizations of ily changed. Openness is a low-maintenance
17. O penness to E x p e rie n ce 261
trait (Kammrath, Ames, & Scholer, 2007). These men and women tend to be high on
That is, initial impressions can be resis Absorption and low on Traditionalism, two
tant to reevaluation. In contrast to traits scales from the Multidimensional Personal
such as Agreeableness and Conscientious ity Questionnaire (M PQ) closely related to
ness, which require frequent confirmatory Openness (Johnson, M cGue, Krueger, &c
evidence to maintain the judgment, impres Bouchard, 2004). They may find fulfillment
sions of Openness are relatively impervious in other types of relationships and activities
to disconfirming evidence; information that and, without a strong internal need to con
contradicts the initial Openness impression form to the expectations of society, pursue
tends to be disregarded. Once an individual these interests instead of potential mates.
is tagged as being open (or closed), regard Whether single, dating, or married, peo
less of the amount of evidence to the con ple have a good idea of what they want in
trary, the impression sticks. Kammrath and their ideal partners-—often someone like
colleagues (2007) suggested that lay concep themselves, particularly on Openness. When
tions of both Openness and ability may con contemplating the ideal mate, single individ
tribute to stable impressions of Openness. uals prefer partners who strongly resemble
Specifically, people equate Openness with them on Openness, with Agreeableness and
ability and perceive ability as stable; thus Extraversion coming in a distant second
people are less sensitive to disconfirming and third, respectively (Figueredo, Sefcek,
evidence. & Jones, 2006). A similar pattern holds
These laboratory studies of person per for both dating couples and newlyweds, al
ception are complemented by correlational though at the stage of marriage, similarity
studies, in which agreement among observer on Conscientiousness becomes slightly more
ratings and between ratings and self-reports important than a match on Openness (Bot-
can be studied among people who have win, Buss, & Shackelford, 1997). And re
known each other, not for seconds or min gardless of their own personalities, women
utes, but for up to 70 years (Costa & M c in particular value mates who are open and
Crae, 1992b). Such studies typically show dominant (Botwin et al., 1997). Taking an
that length of acquaintance increases cross evolutionary perspective, Botwin and col
observer agreement over the course of weeks leagues (1997) suggested that women prefer
or months (Kurtz & Sherker, 2003). Among these qualities because they are the most
long-term acquaintances, cross-observer strongly associated with resource acquisi
correlations for Openness, typically .40 to tion.
.60, are similar to those found for other fac Despite these clear preferences, most peo
tors (Connolly, Kavanagh, & Viswesvaran, ple settle for much less. Some studies find
2007). This level of agreement is seen in no correlation between ratings of an ideal
studies around the world (McCrae et al., partner and ratings of an actual partner
2004). (Figueredo et al., 20 0 6 ); others find a moder
ate correlation at best (Botwin et al., 1997).
Although we can build the ideal mate in our
M arriage and Fam ily
minds, the constraints of reality typically
In any relationship, dynamics of the interac force compromise. In the end, other factors,
tion are shaped, in part, by the personalities such as physical attractiveness, proximity, or
of the individuals involved. Although true availability, may be more important than the
for any dyadic interaction, most evidence ideal personality.
comes from research on romantic relation But people do want a partner with a simi
ships and married couples. At each stage, lar personality, and it is important to ask to
from deciding whether to get married to what extent individuals succeed in finding
parenting, Openness shapes these choices, such a match. This question is of consider
interactions, and consequences. able interest to behavioral geneticists, who
Marriage is a normative and expected typically assume no assortative mating in
event; there is often considerable social pres calculating estimates of heritability. That is,
sure to “find someone, settle down, and start they presume that an open man would be
a family.” Yet, despite this pressure, some just as likely to marry a closed woman as an
choose to remain single and never marry. open woman. Researchers have now docu
262 IV. C O G N IT IV E D IS P O S IT IO N S
mented couples’ similarity on a variety of at dates. This trait similarity among married
tributes, from intelligence to social attitudes couples appears to come from initial choice
to personality. In one large-scale study of rather than convergence over time. People
newlyweds, Watson and colleagues (2004) with the same values and intellectual pur
found the highest similarity correlations for suits seek each other out rather than mold
age, religiousness, and political conserva each other into their likenesses over time.
tism (mean r = .71), lower correlations for Openness not only influences mate selec
education and intelligence (mean r = .43), tion, but it also shapes relationship quality,
and virtually no correlation for any of the conflict interactions, and daily life within
FFM personality traits (mean r = -.0 3 ). the family. Although people maintain that
But given that Openness is strongly related they want someone similar to themselves on
to political conservatism, religiosity, and ed Openness and are somewhat successful in
ucation, one would expect some evidence of finding a similar mate on this trait, similar
assortative mating for this trait. And indeed, ity does not necessarily imply relationship
despite somewhat mixed findings, similarity satisfaction. Nemechek and Olson (1999),
on Openness emerges more often than not. for example, found that partners who were
Neyer and Voigt (2004), for example, found similarly conscientious had higher levels of
significant correlations for both Openness (r marital adjustment but that similarity on
= .25) and Conscientiousness (r = .39), but Openness was unrelated to adjustment. Even
not for Neuroticism, Extraversion, or Agree discrepancies between ideal partner person
ableness. Similar findings are summarized in ality and actual partner personality do not
McCrae (1996). Biases such as age, gender, predict dissatisfaction (Botwin et al., 1997).
education, and assessment method may con In contrast, degree of Openness, rather
tribute to these inconsistent findings. than similarity, is associated with satisfac
Recently, McCrae and colleagues (2008) tion in both serious dating relationships (e.g.,
analyzed trait similarity using both self- Neyer & Voigt, 2004) and among married
reports and spouse ratings of personality in couples (e.g., Donnellan, Conger, & Bryant,
married couples across four cultures, con 2004). Interestingly, husbands’ and wives’
trolling for these potential biases. Consistent Openness contributes to different aspects of
with previous research, similarity correla relationship satisfaction. For both husbands
tions for the broad domains were generally and wives, husbands’ level of Openness is
modest, and Openness had the largest cor related to satisfying relationships overall
relation (mean r for Openness across the (Botwin et al., 1997; Neyer & Voigt, 2004)
three cultures = .22). Facet-level analyses and well-adjusted marriages (Bouchard,
revealed that couples were drawn together Lussier, & Sabourin, 1999). Wives’ level of
on some aspects of Openness more than Openness, however, is unrelated to marital
others. Across the different cultures, Open adjustment (Neyer Sc Voigt, 2004). On the
ness to Values consistently showed the most flip side, wives’, but not husbands’, level of
evidence for trait similarity: liberals seek out Openness is related to the couple’s sexual
other liberals, whereas conservatives seek satisfaction (Donnellan et al., 2004). Don
out other conservatives. Part of this pairing nellan and colleagues speculated that Open
is likely a matter of convenience; these two ness is related to sexual satisfaction because
types of people inhabit very different social open individuals are motivated to seek out
worlds. In addition, their differing ideologies new and varied experiences; open wives may
would likely be a continued source of argu be more willing to explore new and varied
ment and conflict within the relationship. sexual experiences, which may translate into
Although lower in magnitude, individuals greater sexual satisfaction for both part
also tend to marry partners who are similar ners.
to themselves on 0 2 : Aesthetics (McCrae et Conflict between two people, however,
al., 2008). In the early stages of dating, to get is inevitable, and communication is often
to know each other, couples may engage in touted as the key to maintaining a healthy,
shared interests, such as going to art muse satisfying relationship. Flow individuals ap
ums or the symphony. If one partner adores proach (or avoid), work through, and re
the arts, whereas the other one is bored stiff, solve conflict has major implications for
the relationship may last only one or two the health of the relationship. The flexibil
17. O penness to E x p e rie n ce 263
ity, perspective-taking ability, and willing who score higher on conventionalism report
ness to tolerate differences of opinion of less marital distress (Snyder, Mangrum, &
open people may facilitate communication W ills, 1993). Furthermore, among middle-
and reduce conflict. Open men and women aged women, divorce is associated with a
have a constructive communicative style in more liberal/radical political orientation
which they actively negotiate conflicts while (Fahs, 2007). Both conventionalism and po
recognizing the other’s perspective. That is, litical ideology have been associated with
both members of the couple face the con Openness, and these findings suggest that
flict, freely express their feelings, and work the relation between Openness and relation
together toward resolution. In contrast, ship satisfaction and length may be a com
closed women prefer to avoid discussion or plex one.
change activities when conflict occurs. And, Finally, Openness shapes daily life within
regardless of their own Openness, men per the family, particularly when it comes to
ceive conflict interactions with closed wives parenting. Closed individuals value obedi
as characterized by demand-withdraw: The ence and deference to authority without
wife criticizes, complains, and demands question, whereas open individuals are more
change, and in response the husband avoids open-minded, tolerant, and willing to listen
the conflict by being silent or walking away to opposing arguments. These characteris
(Heaven, Smith, Prabhakar, Abraham, & tics are readily apparent in their different
M ete, 20 0 6 ). With these types of interaction parenting philosophies. In interactions with
styles, it is hardly surprising that closed in their children, open parents are emotionally
dividuals typically have less satisfying rela expressive and warm, and they encourage
tionships. children to voice their opinions. In contrast,
In addition to communication, effective closed parents demand obedience, expect
coping is also important to the health of the their children to follow their rules without
relationship. When faced with marital dif question, and limit their children’s autonomy
ficulties, both husbands and wives high in (Metsapelto & Pulkkinen, 2003). The con
Openness engage in problem-focused coping sequence of these different parenting styles
(Bouchard, 2003). That is, they try to iden may be evident in their children’s behavior:
tify the cause of the relationship stress and Open parents are less likely to report child
then actively work to change the identified misbehavior as a major daily stressor (Lee-
elicitor. Open individuals may be com fort Baggley et al., 2005). It is possible, however,
able with this strategy because of their natu that open parents are more tolerant of child
ral ability to find novel solutions to problems misbehavior rather than actually having
and their willingness to try new approaches more well-behaved children.
when old ones fail. In contrast, when faced
with interpersonal stress, closed individuals
Strangers and F riends
employ distancing coping strategies, such as
ignoring the problem or refusing to become The social consequences of Openness for
emotionally involved (Lee-Baggley, Preece, interpersonal interactions are not limited to
& DeLongis, 2005). These individuals are romantic relationships and the family. Open
uncomfortable with strong emotional reac and closed individuals have different styles
tions and may employ distancing techniques of interacting with the world that influence
as a preemptive strategy against such experi how they interact with strangers, the types
ences. These strategies are not without conse of friends they seek out, and how those rela
quence, however, and their relative effective tionships are maintained. Open and closed
ness may be observable by others. Donnellan individuals differ in their political orienta
and colleagues (2004), for example, found tions, beliefs about religion, and intellectual
that independent observers judged open men interests. These characteristics may influ
and women to have interactions that were ence friendships for at least two reasons.
less negative while discussing their relation First, people tend to meet each other when
ships. enjoying shared interests; thus a foreign-film
In some contexts, however, low Openness buff and a NASCAR fanatic are not likely to
may be related to more beneficial outcomes. cross paths often. Second, politics and reli
Following therapy, for example, couples gion are often sources of great conflict when
264 IV. C O G N IT IV E D IS P O S IT IO N S
strongly held convictions differ. Constant ris, M ilich, & Georgesen, 2003). Teasing a
argument does not make a good basis for stranger, even if playfully, violates norm ex
friendship. pectations and may create a novel situation
Across the five factors, correlations be that closed people find uncomfortable.
tween friends tend to be modest at best Low Openness has likewise been associ
(Berry, W illingham, & Thayer, 2 0 0 0 ). Simi ated with other problems in interpersonal
larity correlations for Openness, however, functioning. In evaluating their interper
are clearly the largest (r = .35). Similar to sonal interactions, these individuals endorse
romantic partners, individuals tend to seek items related to difficulty in perspective tak
out friends who share similar interests. As ing, being easily persuaded by others (pre
McCrae (1996) pointed out, “open people sumably those in positions of authority),
are bored by the predictable and intellec and losing their sense of self when interact
tually undemanding amusements of closed ing with strong-minded others (Gurtman,
people; closed people are bored by what they 1995). And just as these characteristics influ
perceive to be the difficult and pretentious ence conflict and communication in couples’
culture of the open” (p. 331). Given these interactions, they also affect interactions be
different orientations to the world, open and tween friends. In a diary study, for example,
closed individuals are unlikely to voluntarily closed individuals had more conflicts with a
spend enough time with each other to devel close friend over a 4-week period than did
op a lasting friendship. open individuals. In response to the conflict,
In addition to studying the basis for closed friends were more likely to engage in
friendship, it is also of interest to ask how passive-aggressive strategies, whereas open
Openness shapes casual interactions among friends adopted a forgive-and-forget strat
strangers and its role in interpersonal inter egy. Also similar to couples, these strategies
actions between friends. When getting ac do not go unnoticed; friends get more irritat
quainted, open individuals spend more time ed with closed friends than with open ones
looking at their interaction partners and less (Berry et al., 2000).
time talking about themselves. Observers The relation between Openness and
to these conversations mistake this greater conflict, however, takes a different course
visual attention as an indication of relation among college roommates than between
ship quality (Berry & Hansen, 20 0 0 ). Yet friends; in this case, open individuals are
Openness is unrelated to perceived interac more likely to have conflict with their room
tion quality in either spontaneous interac mates (Bono, Boles, Judge, & Lauver, 2002).
tions in same-sex dyads (Berry & Hansen, Unlike friendships, students typically have
20 0 0 ) or in getting-to-know-you conversa little choice in their roommates, and a mis
tions in opposite-sex interactions (Berry & match on Openness may be one source of
Miller, 2001). Open individuals are curious conflict. And, indeed, conflict was unrelated
and attentive to the world around them and, to Openness when roommates had similar
in the process of getting to know somebody mean levels and conflict was marginally re
new, their curiosity may lead them to look lated to mean-level differences in Openness
more intently at their interaction partners as between roommates. At both ends of the
they take them in and try to figure them out. continuum, like-minded individuals may
This nonverbal cue, however, does not facili understand each other better and feel more
tate high-quality interactions. comfortable as roommates. Roommates mis
Closed individuals are sensitive to appro matched on Openness, in contrast, may be
priate social interactions between strangers likely to butt heads if one is unconventional
and react strongly when norm expectations and emotional and the other conservative
are violated. In one study, for example, com and stoic. In addition, what might be fun
pared with the control condition, closed argumentativeness for an open individual
participants became less friendly after being may amount to a serious conflict for a closed
teased by a confederate, and their narratives individual. For both reasons, there may be
of the interaction with the confederate were less conflict when roommates are matched
less positive. For open participants, in con on Openness.
trast, being teased did not influence their Finally, one great benefit of a close rela
interaction with the teaser (Bollmer, H ar tionship is the support that can be provided
17. O penness to E x p e rie n ce 265
by the other during times of great stress. retail assistants. Higher elevations of Open
Openness is associated with both the type ness (as well as Agreeableness and Consci
and frequency of support offered to others. entiousness) were associated with better
Open individuals reciprocate emotional sup performance (Neuman, Wagner, & Chris
port, whereas closed individuals reciprocate tiansen, 1999). Taggar (2000) analyzed data
instrumental support (Knoll, Burkert, & from 94 teams at both individual and team
Schwarzer, 2006). Once open individuals re levels and found that Openness had no effect
ceive emotional support from a friend, they at the individual level but that the greater the
easily return the favor, which likely deepens proportion of team members high in Open
the emotional bond between them. In con ness was, the better the performance was. An
trast, instrumental support is more concrete analysis of the specific behaviors responsible
and costly; perhaps closed individuals feel for good performance suggested that open
indebted and thus more compelled to re members contributed by generating ideas,
ciprocate. Instrumental support, although promoting free discussion, and synthesizing
costly for the individual, is often more ben team efforts. Openness has also been found
eficial for the recipient due to the practical to promote emergent leadership— the ability
application of the support. Thus, in times of to take charge of a leaderless group (Kickul
distress, when concrete solutions are needed, & Neuman, 2000).
closed individuals may provide more useful A meta-analysis of job performance and
support. These different approaches to sup team personality found advantages for
port likely affect the nature and closeness of teams higher in Openness, but only in field
the friendship over time. studies, not laboratory studies (Bell, 2007),
Taken together, these findings demonstrate suggesting that it is the long-term effects of
how Openness shapes interpersonal interac Openness that are noticeable. Another meta
tions, from casual interactions to long-term analysis sorted studies by the kind of task
committed relationships. Open and closed involved, using Holland’s (1985) vocational
individuals tend to develop lasting relation typology. Predictably, team-level Openness
ships with like-minded individuals, and sub predicted success in Investigative tasks (An
sequently these pairings have implications derson, 2006) but was unrelated to success in
for a variety of outcomes, from relationship Social, Conventional, or Enterprising tasks.
satisfaction to conflict resolution to parent LePine (2003) examined the effect of intro
ing to social support. Clearly, an individual’s ducing an unforeseen change— a breakdown
experiential orientation to the world pro of communication— in a simulated military
foundly affects his or her interaction with “command and control” task. Teams high in
the people in it. Openness (and low in C2: Order, C3: Duti
fulness, and C6: Deliberation) adapted to the
new situation more readily and successfully.
Openness in W ork Groups Bing and Lounsbury (2000) studied perfor
mance of managers of Japanese companies
In the past decade, industrial/organizational operating in the United States; presumably
psychologists have taken an interest in the because they could handle the complexi
effects of personality traits on team per ties of cross-cultural interactions, managers
formance. Although teams with high mean high in Openness were rated higher in per
levels of Conscientiousness tend to perform formance.
well in many situations, results are much However, high Openness also presents
more mixed for Openness. High team-level problems for groups. For example, G. H.
Openness is generally advantageous, but Kickul (2000) found that Openness was
often only for certain kinds of tasks or with negatively related to goal clarity (presumably
in certain contexts. And in some respects, because people high in Openness kept gener
Openness interferes with the work of the ating new ideas), and Lun and Bond (2006)
group. found that it interfered with achieving rela
An early study of team personality el tionship harmony in a work group (perhaps
evation (mean level) and variability (within- because members high in Openness were too
team variance) examined customer service individualistic). A study of 22 0 individuals
and task completion ratings for 82 teams of in 45 teams also found that Openness (like
266 IV. C O G N IT IV E D IS P O S IT IO N S
low Agreeableness) was inversely related sample, Ekehammar and Akrami (2007) ex
to peer-rated social role behavior— that is, amined correlations of generalized prejudice
how well group members got along (Stew (a composite of ethnic prejudice, sexism, ho
art, Fulmer, & Barrick, 2005). In another mophobia, and prejudice against people with
study, employees high in Openness were low mental disabilities) with N EO -PI-R scales.
in organizational loyalty, especially if they At the domain level, the strongest correla
lacked resources (Moss, M cFarland, Ngu, & tions (both -.4 9 ) were with Openness and
Kijowska, 2006). Agreeableness; at the facet level, the stron
At least one finding relates to the team vari gest were A6: Tender-Mindedness (-.61) and
ability in Openness. Given the frequent an 0 6 : Values (-.55), which are considered at-
tagonism between individuals high and low titudinal facet scales. However, prejudice
in Openness and their very different working was also inversely related to Openness to
styles and goals, it is perhaps not surprising Fantasy, Aesthetics, Feelings, and Actions,
that a meta-analysis found that homogeneity rs = - .2 5 to -.4 9 , N = 170, p < .05.
with respect to Openness led to better group Flynn (2005), in studies of white Ameri
performance, at least among professional cans, found that Openness is associated
teams. Presumably the best results— and the with lower racial prejudice, more favorable
highest levels of morale— would be obtained judgments of a fictional black character, and
by choosing teams uniformly high in Open more favorable assessments of black inter
ness to deal with changeable situations and viewees, and attributed this in part to the
investigative tasks and teams uniformly low willingness of people high in Openness to
in Openness to deal with well-structured, consider stereotype-disconfirming informa
conventional tasks. tion. Duriez and Soenens (2006) found that
racism was related to low Openness (and
low Agreeableness) among Belgian ado
Social and Political Effects lescents. Given the strong, consistent, and
theoretically expectable associations of low
The quotations from Agnew and Kierkeg Openness with prejudice and racism, it is ex
aard that open this chapter illustrate not traordinary that, of 11,015 items found in a
only the substantive differences between PsycINFO search on “prejudice or racism,”
closed and open people in social attitudes only 10 involved Openness. Social psycholo
but also the affective tone: Both sides hold gists have overlooked one of the key deter
the other in contempt. Agnew famously de minants of one of their most-studied phe
clared that so-called intellectuals were “an nomena.
effete corps of impudent snobs,” whereas There has been much less research on
Kierkegaard clearly regarded his fellow reverse prejudice, but Lecci and Johnson
citizens as Philistines. There is, however, a (2008) reported the intriguing finding that,
subtle asymmetry in these characterizations. among American blacks, in addition to the
Agnew, spokesman for the Silent Majority, expectable inverse association with Agree
assures us that Americans, preferring order ableness, there is a small (r = .15) but sig
to freedom, share his values. By contrast, nificant positive correlation between Open
the Danish existentialist emphasizes his iso ness and antiwhite attitudes. Perhaps it was
lation, distinguishing himself from “them.” Openness that led the Black Panthers to
Open people prize uniqueness and individu challenge the racist status quo in mid-20th-
ality (Dollinger, Ross, & Preston, 2002) at century America.
the cost of some social alienation; closed There is ample evidence that Openness
people are loyal and patriotic, strongly iden is inversely related to authoritarianism, as
tified with their own kind. Open people root well as to other, less extreme forms of so
for the underdog; closed people support fa cial conservativism. Van Hiel, Kossowska,
vorites (W ilkinson, 2007). and Mervielde (2000) examined left-right
The dark side of closed people’s in-group political ideology in Belgium and Poland.
loyalty is their intolerance for out-groups, A right-wing ideology score was defined in
characterized by Agnew as “Yippies, Hip part by preference for nationalist parties
pies, Yahoos, Black Panthers, lions and ti over socialist and green parties, as well as by
gers . . . the whole damn zoo.” In a Swedish general conservative political beliefs. This
17. O penness to E x p e rie n ce 267
index was inversely related to 0 6 : Values in ing the citizens of some cultures as being, in
each of four samples (rs = - .3 7 to - .6 4 , p < general, more open than others— although
.001), but it was also more modestly related differences between cultures were generally
to each of the other facets in one or more small compared with the ubiquitous individ
of the samples. For example, O l: Fantasy, ual differences within cultures.
which shares no obvious content with mea Which cultures are most open? O f the 28
sures of ideology, showed correlations rang cultures with both self-report and observer-
ing from - .2 0 to - .3 9 , all p < .05, in the Bel rating data, the highest mean Openness
gian samples. scores were found for French-speaking
In a later study, Van Hiel and Mervielde Switzerland, Serbia, Austria, Germany,
(2004) related Openness to separate mea and German-speaking Switzerland, with
sures of cultural and economic conserva T-scores of 53 to 59. The lowest-scoring
tism. Cultural conservatism was related to countries were Croatia, Spain, Hong Kong,
Openness and all its facets, but economic Malaysia, and India, with T-scores of 46 to
conservatism was unrelated to total Open 49. It is surely puzzling that Serbia scores
ness and only weakly related to 0 2 : Aesthet so much higher than Croatia, but the other
ics (r = -.1 9 ) and 0 6 : Values (r = -.1 5 ). The findings make a certain sense: Modern, pro
strongest personality predictor of economic gressive, well-educated countries are higher
conservatism was low Agreeableness (r = in Openness than are traditional cultures.
- .2 3 , p < .001). Economic conservatives may The United States was near average on ag
be mean, but they are not necessarily closed. gregate Openness.
Economic conservatism is presumably based It is possible to move past simple impres
on ideology and self-interest; cultural con sions about these sets of cultures by con
servatism is psychological rather than ideo ducting culture-level analyses, relating ag
logical (cf. Van Hiel & Mervielde, 2004) gregate Openness levels to other features
and seems to reflect the preference of closed of nations. M cCrae (2002) showed that
individuals for simple, stable, and familiar Openness was significantly related to three
beliefs and values. In a Polish study, cultural of Hofstede’s (2001) dimensions of culture:
but not economic conservatism was found to low Power Distance, high Individualism,
be heritable (Oniszczenko & Jakubowska, and high Masculinity. The first two of these
2005). associations were replicated in the observer-
rating study (McCrae et al., 2005b). Thus
people from cultures with high mean levels
A ggregate Openness and Culture of Openness prefer egalitarian to hierarchi
cal social structures and focus on themselves
C ross-C ultural Analyses
as individuals rather than on the groups
Within the past few years, cross-cultural to which they belong. McCrae and col
studies have suggested that nations differ leagues (2005b) examined country scores on
systematically in mean levels of personality Schwartz’s (1994) values survey and found
traits (but see Poortinga, van de Vijver, & positive correlations of Openness with Af
van Hemert, 2 0 0 2 , for a critical view of that fective Autonomy, Intellectual Autonomy,
claim). McCrae (2002) assembled self-report and Egalitarian Commitment values and a
N EO -PI-R data from 36 cultures, and M c negative correlation with Conservatism.
Crae and colleagues (2005b) gathered Based on data from the World Values Sur
observer-rating N EO -PI-R data from 51 vey, Inglehart and Norris (2003) identified
cultures. Mean aggregate personality scores two broad dimensions: Survival versus Self-
were calculated for each culture. Across the Expression and Traditional versus Secular-
two datasets, convergent culture-level corre Rational. Openness was significantly related
lations were significant for 4 of the 5 factors to Secular-Rational values (r = .34, N = 42,
and 26 of the 30 facets. In particular, corre p < .05) and showed a trend toward a posi
lations for Openness facets ranged from .44 tive association with Self-Expression [r =
for 0 4 : Actions to .75 for 0 6 : Values, with .29, N = 42, p < .10). Traditional cultures
a correlation of .50 for total Openness. Thus are guided by religion and tend to reject
different samples using different methods of abortion, divorce, and euthanasia— values
measurement generally concurred in describ shared by closed individuals. Survival val
268 IV. C O G N IT IV E D IS P O S IT IO N S
ues are generally found in poorer countries, differences in mean levels of traits in differ
where material prosperity is a major concern. ent cultures might be due to artifacts, such as
Cultures with a history of high economic problems in the translation, culture-specific
development have citizens more concerned response styles, or inadequacies in sampling.
about tolerance, imagination, and personal But a demonstration of the validity of ag
fulfillment— goals more congenial to open gregate personality scores has recently been
individuals. provided by Rentfrow, Gosling, and Potter
In general, these are sensible correlates (2008), who used Internet data on the BFI
and suggest that associations found on the collected from over 6 0 0 ,0 0 0 respondents to
individual level may also be found on the compare mean personality trait levels in the
culture level. This is by no means always 50 U. S. states and the District of Columbia.
the case. Cultures high in 0 6 : Values show Here language and national culture were
higher use of the drug Ecstasy (McCrae & held constant, and Rentfrow and colleagues
Terracciano, 2 0 0 8 ), but a study at the in argued that their sample was broadly rep
dividual level in the Netherlands found no resentative. Yet mean-level differences still
difference in Openness levels between those appeared and, in general, made sense. They
who did and did not use the drug (instead, found the highest aggregate level of Open
Ecstasy users were higher on Extraversion ness in Washington, D.C.— which joined
and lower on Conscientiousness; ter Bogt, only Massachusetts in favoring McGovern
Engels, & Dubas, 20 0 6 ). The widespread and Shriver over Nixon and Agnew in the
use of Ecstasy is presumably limited to 1972 presidential election— followed by
wealthy nations, and Openness is correlat New York, Oregon, and Massachusetts.
ed with per-capita gross domestic product. Lowest in Openness were Alabama, Alas
Thus effects of aggregated personality traits ka, Wyoming, and North Dakota. Rent
may be inflated or masked by other culture- frow and colleagues correlated these scores
level variables. with state-level indicators and showed that
However, one association that unfortu Openness was positively related to favorable
nately seems to hold on both levels is that aggregate attitudes toward legalizing mari
between low Openness and H IV stigmatiza juana, abortion, and same-sex marriage;
tion. In Russia and the United States, stig to the proportion of the state population
matization was associated with low Open employed in occupations related to the arts
ness and especially low 0 6 : Values (McCrae and entertainment and to computers and
et al., 20 0 7 ). In the self-report study, black mathematics; and, unexpectedly, to the per-
South African, Zimbabwean, Indian, and capita robbery and murder rates. Openness
Malay cultures scored lowest on 0 6 : Values was negatively related to spending time in a
(McCrae, 2002). In both South Africa and bar or tavern and to attending church. Espe
Zimbabwe, the AIDS epidemic has been fu cially in a democracy, personality traits can
eled by indifference or denial on the part of have a dramatic effect on collective behav
the governments. In India, where at least 2 ior: Oregon was the first state to decrimi
million people are living with H IV infection, nalize marijuana, and Massachusetts recog
“the HIV epidemic is misunderstood and nizes same-sex marriages.
stigmatised among the Indian public. People The chief question remaining is how traits
living with H IV have faced violent attacks; come to be associated with features of cul
been rejected by families, spouses and com ture. Do cultural practices promote the de
munities; been refused medical treatment; velopment of certain traits, or do common
and even, in some reported cases, denied the traits stimulate the evolution of cultural insti
last rites before they die” (AVERT, 2007). tutions? Until quite recently, anthropologists
Fortunately, having learned from the expe and psychologists would have reflexively pre
rience in Africa, both Malaysia and India sumed that culture shapes personality. But
have programs in place to educate the public the strong evidence of predominant genetic
about H IV infection risks. Given the public’s influence on individual differences within
perceptions, anonymous testing and confi cultures makes it reasonable to suppose that
dential treatment ought to be emphasized. the distribution of personality-related trait
Critics of this line of research (e.g., Poort- alleles may differ across nations and give rise
inga et al., 2002) have argued that apparent to different mean personality profiles. Over
17. O penness to E x p e rie n ce 269
the course of centuries, these collective per Riemann, & Livesley, 1998); the large corre
sonality differences may have been one influ lation between Need for Cognition and 0 5 :
ence on culture (M cCrae, 200 4 ). Both sides Ideas (r = .78; Berzonsky & Sullivan, 1992)
have been argued as an explanation for the virtually guarantees that Need for Cogni
links between aggregate traits and Hofstede tion must be substantially heritable. Again,
dimensions (Hofstede & M cCrae, 20 0 4 ), but we know that Openness reaches it highest
there is very little empirical evidence. Some mean level during the early 2 0s— a fact that
support for environmental influences comes surely affects the generalizability of find
from the changes in values that accompany ings from experiments on college students.
changes in economic development (Inglehart Social psychologists are not accustomed to
& Norris, 2003). Some evidence for genetic thinking about the long-term implications of
effects comes from studies of isolated popula their findings, but the longitudinal stability
tions (Ciani, Capiluppi, Veronese, & Sartori, of Openness suggests that patterns of behav
2007). Perhaps the most informative designs ior observed in students may persevere for
are acculturation studies, in which members decades. How might awareness of that fact
of an ethnic group move from one culture to reshape theories of social behavior?
another. Do they retain their ethnic profile This volume is dedicated to integrating
or come to resemble the citizens of the host the topics of individual differences, most
culture? One such study has been reported often studied by personality psychologists,
(M cCrae, Yik, Trapnell, Bond, & Paulhus, and the social behavior that is the focus of
1998); it found that Chinese born in Hong social psychology. Because of its deep intra
Kong scored about one-half standard devia psychic basis and its widespread social con
tion lower in Openness than ethnic Chinese sequences, Openness to Experience may be
born in Canada (an acculturation effect), a particularly useful construct on which to
but that these Canadian-born Chinese Ca center a dialogue between these two fields.
nadians still scored significantly lower than
European Canadians on 0 3 : Feelings and
0 6 : Values (ethnicity effects). Both ethnic A ck n o w led gm en ts
ity and acculturation may affect mean trait
levels for Openness. As Rentfrow, Gosling, P r e p a r a t i o n o f t h is c h a p t e r w a s su p p o r te d by
and Potter (2008) pointed out, genetic and the In tram u ral R esearch P ro g ram , N atio n al I n
environmental influences are likely to be st itu te s o f H e a l t h , N a t i o n a l In s t it u t e o n A g ing .
mutually reinforcing: Open people may be R o b e r t R . M c C r a e rec eives ro y a lt ie s f r o m th e R e
inclined to move to Massachusetts, and the vi sed N E O P e r s o n a l i ty In ventory .
1. R eca ll that the facets of O pen ness are only loo se
Conclusion ly related and thus that individuals may be high
in some facets and low in others. As a group,
Concepts related to Openness— such as the people o f India are high in O pen ness to A e s
Authoritarianism, Need for Closure, and thetics and low in Openness to Values ( M c C r a e ,
Integrative Complexity— have long been 2 0 0 2 ) , as were, perhaps, T. S. Eliot and Ezra
Pound. T h o s e pioneers o f m odern poetr y ended
employed by social psychologists, but usu
up, respectively, as an o r th o d o x A nglican and
ally without a good grasp of their relation to a pro pagand ist for M u ssolin i. O r perhaps their
basic personality traits. There is considerable Openness to Values was so high that they qu es
advantage to construing such scales as indi tioned and rejected the tenets of co nven tional
cators of Openness, because a great deal is liberalism.
known about the origins, development, and 2 . Spiro Agnew studied chem istry at J o h n s H o p
correlates of that factor. For example, there kins University before o b ta in ing a law degree.
3. By co nven tion, the 3 0 facet scales of the N E O -
appear to be no studies on the heritability of
P I-R are designated by the fa cto r initial, a facet
Need for Cognition, but there are many that num ber from 1 to 6 , and the facet nam e. T h e
show that Openness to Experience, and in names o f facets of O pen ness are un derstood to
particular Openness to Ideas, have a strong include “ Openness t o , ” so th a t 0 5 : Ideas is read
genetic basis (e.g., Jang, M cCrae, Angleitner, “ O pen ness to Ideas.”
270 IV . C O G N IT IV E D IS P O S IT IO N S
V. E., Rukavishnikov, A. A., Senin, I. G., et al. Oniszczenko, W., Sc Jakubo wska, U. (2 005). Genetic
(200 4). Consensual validation of personality traits determinants and personality correlates of sociopo
across cultures. J o u r n a l o f R e se a rc h in P erson ality , litical attitudes in a Polish sample. T w in R e se a rch ,
38, 179-201. 8, 4 7 - 5 2 .
McCra e, R. R ., Costa, P. T., Jr., M artin , T. A., Oryol, Piedmont, R. L., Sc Aycock, W. (20 07 ). An historical
V. E ., Senin, I. G., Sc O ’Cleirigh, C. (200 7 ). Person analysis of the lexical emergence of the Big Five per
ality correlates of H IV stigmitization in Russia and sonality adjective descriptors. P erson ality a n d In d i
the United States. J o u r n a l o f R e se a rc h in P e r s o n a l v id u al D iffe r e n c e s , 4 2 , 1 0 5 9 - 1 0 6 8 .
ity, 41, 1 9 0 - 1 9 6 . Poortinga, Y. H., van de Vijver, F., & van Hemert, D.
McCra e, R. R., Martin, T. A., Hrebickova, M ., Ur- A. (2002). Cross-cultural equivalence of the Big
banek, T., Boomsma, D. I., Willemsen, G., et al. Five: A tentative interpretation of the evidence. In
(200 8). Personality trait similarity between spous R. R. M cCra e Sc J . Allik (Eds.), T h e F iv e-F a c to r
es in four cultures. J o u r n a l o f P erson ality , 76(5), M o d e l o f p e r s o n a lity a c r o s s cu ltu res (pp. 2 7 3 - 2 9 4 ) .
1 1 3 7 -1 1 6 4 . New York: Kluwer Academic/Plenum.
Mc Crae, R. R ., & Terracciano, A. (20 08 ). The Five- Rentfrow, P. J ., Sc Gosling, S. D. (2 006). Message in a
Factor Model and its correlates in individuals and ballad: The role of music preferences in interperson
cultures. In F. J . R. Van de Vijver, D. A. van Hemert, al perception. P sy c h o lo g ic a l S cien ce, 17, 2 3 6 - 2 4 2 .
Sc Y. H. Poortinga (Eds.), M u ltilev el a n a ly ses o f in Rentfrow, P. J . , Gosling, S. D., & Potter, J. (20 08 ). A
d iv id u a l a n d cu ltu re (pp. 2 4 9 - 2 8 3 ) . Mahwah, NJ: theory of the emergence, persistence, and expression
Erlbaum. of geographical variation in psychological charac
M cCra e, R. R., Terracciano, A., Sc 79 Members of the teristics. P ersp ectiv es on P sy c h o lo g ic a l S cien ce, 3,
Personality Profiles of Cultures Project. (2005a). 339-369.
Personality profiles of cultures: Aggregate personal Rouse, S. V., Sc Haas, H. A. (2 003). Exploring the ac
ity traits, j o u r n a l o f P erso n a lity a n d S o c ia l P sy c h o l curacies and inaccuracies of personality perception
og y , 89, 4 0 7 - 4 2 5 . following Internet-mediated communication. J o u r
Mc Crae, R. R., Terracciano, A., Sc 79 Members of the n a l o f R e se a rc h in P erson ality , 37, 4 4 6 - 4 6 7 .
Personality Profiles of Cultures Project. (2005b). Sadowski, C. J., Sc Cogburn, H. E. (1 997). Need for
Universal features of personality traits from the ob Cognition in the Big Five structure. J o u r n a l o f
server’s perspective: Data from 50 cultures. J o u r n a l P sy ch o lo g y : In terd iscip lin a ry a n d A p p lied , 131,
o f P erson ality a n d S o c ia l P sy ch olog y , 8 8 , 5 4 7 - 5 6 1 . 307-312.
McCra e, R. R., Yik, M . S. M ., Trapnell, P. D., Bond, M. Schulte, M. J ., Ree, M . J . , Sc Carretta, T. R. (2 004).
H., & Paulhus, D. L. (1998). Interpreting personal Emotional intelligence: No t much more than g and
ity profiles across cultures: Bilingual, acculturation, personality. P erson ality a n d In d iv id u a l D iffe r e n c
and peer rating studies of Chinese undergraduates. es, 37, 1 0 5 9 - 1 0 6 8 .
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 74, Schwartz, S. H. (1994). Beyond individualism/collec
1041-1055. tivism: New cultural dimensions of values. In U.
Meh l, M. R., Gosling, S. D., & Pennebaker, J . W. Kim, H. C. Triandis, C. Kagitcibasi, S.-C. Choi,
(20 0 6). Personality in its natural habitat: M an ife s Sc G. Yoon (Eds.), In d iv id u a lis m a n d c o llec tiv ism :
tations and implicit folk theories of personality in T h eo r y , m e th o d , a n d a p p lic a tio n s (pp. 8 5 - 1 1 9 ).
daily life. J o u r n a l o f P ers o n a lity a n d S o c ia l P sy Thousand Oaks, CA: Sage.
c h o lo g y , 9 0 , 8 6 2 - 8 7 7 . Sneed, C. D., Mc Crae, R. R ., & Funder, D. C. (1998).
Metsapelto, R. L., & Pulkkinen, L. (2003). Personal Lay conceptions of the Five-Factor Model and its
ity traits and parenting: Neuroticism, Extraversion, indicators. P erson ality a n d S o c ia l P sy ch o lo g y B u l
and Openness to Experience as discriminative fac letin , 2 4 , 1 1 5 - 1 2 6 .
tors. E u r o p ea n j o u r n a l o f P erson ality , 17, 5 9 - 7 8 . Snyder, D. K., Mangrum, L. F., Sc Wills, R. M . (1993).
Moss, S. A., McFarland, J., Ngu, S., Sc Kijowska, A. Predicting couples’ response to marital therapy:
( 2 0 06 ). Maintaining an open mind to closed indi A comparison of short- and long-term predictors.
viduals: The effect of resource availability and lead J o u r n a l o f C o n su ltin g a n d C lin ica l P sy ch olog y , 61,
ership style on the association between Openness to 61-69.
Experience and organizational commitment, j o u r Staudinger, U. M., Maciel, A. G., Smith, J., & Baltes,
n a l o f R e se a rc h in P erson ality , 4 1 , 2 5 9 - 2 7 5 . P. B. (1998). What predicts wisdom-related perfor
Nemechek, S., Sc Olson, K. R. (1999). Five-factor per mance?: A first look at personality, intelligence, and
sonality similarity and marital adjustment. S o cia l facilitative contexts. E u r o p ea n J o u r n a l o f P ers o n a l
B e h a v io r a n d P erson ality , 27, 3 0 9 - 3 1 8 . ity, 12, 1-17.
Neuman, G. A., Wagner, S. H., Sc Christiansen, N. Stewart, G. L., Fulmer, I. S., Sc Barrick, M . R. (2005).
D. (1999). The relationship between work-team An exploration of member roles as a multilevel link
personality composition and the job performance ing mechanism for individual traits and team out
of teams. G r o u p a n d O rg a n iz a tio n M a n a g em en t, comes. P erso n n el P sy ch olog y , 5 8 , 3 4 3 - 3 6 5 .
24, 28-45. Taggar, S. ( 2 0 00 ). Personality, cognitive ability and be
Neyer, F. J ., Sc Voigt, D. (20 04 ). Personality and so haviour: The antecedents of effective autonomous
cial network effects on romantic relationships: A work teams. D iss er ta tio n A b stra cts In te rn a tio n a l,
dyadic approach. E u r o p ea n J o u r n a l o f P erson ality , 60(9A), 3 43 8.
18, 2 7 9 - 2 9 9 . ter Bogt, T. F. M ., Engels, R. C. M. F2., Sc Dubas, J. S.
Noftle, E. E., & Robins, R. W. (20 07 ). Personality pre (20 0 6). Party people: Personality and M D M A use
dictors of academic outcomes: Big Five correlates of of house party visitors. A d d ictiv e B eh a v io r s, 31,
GPA and SAT scores, j o u r n a l o f P erson ality a n d 1240-1244.
S o c ia l P sy ch olog y , 9 3, 1 1 6 - 1 3 0 . Terracciano, A., Costa, P. T., Jr., Sc M cCrae, R. R.
17. O penness to E x p e rie n ce 273
(20 0 6). Personality plasticity after age 30. P ers o n a l J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 87,
ity a n d S o c ia l P sy c h o lo g y B u lletin , 3 2 , 9 9 9 - 1 0 0 9 . 123-132.
Terracciano, A., Merritt, M., Zonderm an, A. B., & Watson, D., Klohnen, E. C., Casillas, A., Simms, E.
Evans, M . K. (2003). Personality traits and sex N., Haig, J ., & Berry, D. S. (20 04 ). M atch makers
differences in emotion recognition among African and deal breakers: Analyses of assortative mating
Americans and Caucasians. A n n als o f the N ew in newlvwed couples. J o u r n a l o f P erson ality , 72,
York A c a d e m y o f S c ien c es, 1 0 0 0 , 3 0 9 - 3 1 2 . 1029-1068.
Tetlock, P. E., Peterson, R. S., &c Berry, J. M . (1993). Webster, D. M ., & Kruglanski, A. W. (1994). Indi
Flattering and unflattering personality portraits vidual differences in need for cognitive closure.
of integratively simple and complex managers. J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 67,
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 6 4, 1049-1062.
500-511. Wilkinson, T. J. (20 07 ). Individual difference and
Trapnell, P. D. (1994). Openness versus intellect: A sport fans: W ho roots for the underdog? D is s e r ta
lexical left turn. E u r o p e a n J o u r n a l o f P erson ality, tion A b stra cts In te r n a tio n a l, 67(8B), 4750.
8, 2 7 3 - 2 9 0 . Yamagata, S., Suzuki, A., Ando, J., On o, Y., Kijima,
Van Hiel, A., Kossowska, M ., & Mervielde, 1. (2 000). N., Yoshimura, K., et al. (20 06 ). Is the genetic
The relationship between Openness to Experience structure of human personality universal?: A cross-
and political ideology. P erson ality a n d In d iv id u a l cultural twin study from North America, Europe,
D iffe r e n c e s , 2 8 , 741 -7 51 . and Asia. J o u r n a l o f P erson ality a n d S o c ia l P sy
Van Hiel, A., & Mervielde, I. (20 04 ). Openness to E x c h o lo g y , 9 0 , 9 8 7 - 9 9 8 .
perience and boundaries in the mind: Relationships Zuckerman, M., Kuhlman, D. M., Joirem an, J ., Teta,
with cultural and economic conservative beliefs. P., & Kraft, M. (1993). A comparison of three struc
J o u r n a l o f P erson ality , 7 2 , 6 5 9 - 6 8 6 . tural models for personality: The Big Three, the Big
Vazire, S., &c Gosling, S. D. (200 4 ). e-Perceptions: Per Five, and the Alternative Five. J o u r n a l o f P ers o n a l
sonality impressions based on personal websites. ity a n d S o c ia l P sy ch olog y , 65, 7 5 7 - 7 6 8 .
CHAPTER 18
A d r ia n F u r n h a m
274
18. L o cu s o f C o n tro l and A ttrib u tio n Style 275
ing separate groups of low-anxiety externals ston & Wallston, 1981). Tremendous situa
and high-anxiety externals (Dawkins & tion specificity and intraindividual variation
Furnham, 1989). may exist across domains of activity and set
Another approach to the problem of de tings. In this sense, a person may be internal
fensive externality may be to include ques with regard to one set of activities or actions
tionnaire items that have been shown to and external with regard to another. This
elicit a fairly uniform internal attribution. seeming contradiction may be both perfectly
If the experimental participant makes an explicable in terms of the person’s life expe
external attribution on such an item, there riences and highly adaptable.
may be good reason to infer that this re
sponse is defensively motivated. The possi
S elf versus Other
bility that defensive processes may influence
locus-of-control judgments should also be People may maintain different locus-of-
considered in connection with the preceding control belief systems for themselves and
discussion of reinforcement valence. That is, for others (Furnham & Steele, 1993; Gurin,
a person may verbally indicate that a certain Gurin, Lao, & Beattie, 1969). For example,
reinforcement or outcome is not important an individual’s profile of expectancy beliefs
as a defense against expected failure. Such may show that he or she has an internal locus
a reaction might be described as “defensive of control when assessing other people’s be
undervaluation.” This tendency goes hand havior but an external locus of control with
in hand with the tendency toward defensive respect to him- or herself. Or a person may be
externality. an instrumentalist with regard to his or her
own beliefs and behaviors but react to others
as if they were fatalists or at least at the mercy
H e a lt h y —U n h ea lth y D ic h o to m y
of forces beyond their control. Furthermore,
Much research starts with the assumption people may hold person-specific as well as
that having an internal locus of control is situation-specific locus-of-control beliefs,
good, adaptive, and healthy, but that it is constituting multiple sets of beliefs that may
bad or maladaptive to have an external locus overlap in differing degrees. Thus they may
of control. Indeed, an overwhelming body of hold internal locus-of-control beliefs about
evidence suggests that internality is gener themselves but external locus-of-control be
ally associated with more desirable charac liefs about their families or vice versa. The
teristics and behaviors than externality. For picture takes on added complexity when we
example, internals are more likely to display consider that some of these beliefs are more
many sorts of healthy and adaptive behaviors or less defensively motivated.
at school, work, and play (Lefcourt, 1991).
However, it is questionable to assume that
Attributions o f Cause, Responsibility,
only positive attributes and actions are as
and Blame
sociated with internality. As people with an
internal locus of control tend to take respon The locus-of-control and attribution-of-
sibility for the consequences of their actions cause literatures clearly overlap. The major
more readily than externals, they are more difference between these concepts is that
likely to experience lower self-esteem when whereas attributional measures are con
faced with failure. They may also respond cerned with the causes o f past events , locus
to uncontrollable events less well than exter of control measures are concerned mainly
nals. Therefore, externality may sometimes with the expectation o f future events. Hence
be associated with altruism and collectivist the paradox that although an external as
attitudes and internality with selfish and cription can be made about the physical or
ruggedly individualistic modes of action. psychological cause of an event— for ex
Not all conceptualizations of locus of con ample, failure to achieve a goal— an inter
trol involve only the internal and external nal attribution may be made concerning
dimensions. More complex conceptualiza responsibility if the consequence could have
tions provide a means of exploring the inter been foreseen. Causation and responsibility
relationships of different varieties of internal should therefore be treated as separate but
and external locus-of-control beliefs (Wall- related concepts. Locus of control is fre
18. L o cu s o f C o n tro l and A ttrib u tio n Style 277
quently associated with perceived cause but (Ashkanasy, 1985). Both Rotter (1966) and
not responsibility (for future events), and Franklin (1963) reported that the Rotter
the two operate quite differently. Locus-of- Internal-External Locus of Control Scale
control beliefs are partially the product of was unidimensional, but a number of later
causal attributional beliefs about past events studies did not replicate their results. Gurin
and should be distinguished conceptually and colleagues (1969) and Sanger and W alk
from both causal beliefs and responsibility er (1972) reported two factors involving per
beliefs. sonal control and control ideology, whereas
Mirels (1970) and Cherlin and Bourque
(1974) found two different factors (general
Stability and Temporality
control and political control). Collins (1974),
People may hold more internal beliefs for who separated item pairs for the analysis, re
outcomes that are temporally distant because ported four factors, and Schneider and Par
they are separated from present actions by sons (1970) isolated five.
various intervening or confounding events. Empirical and theoretical analyses by Lev-
Thus locus-of-control beliefs for events im enson (1981) suggested that the inconsisten
minent in the short term may differ from cies and inadequacies of the I -E Scale would
beliefs for events expected to occur only in be improved by making a distinction in the
the long term. This possibility may affect external scale between believing in power
the stability of the locus of control patterns, ful others who control the world and believ
which may be stable or unstable depending ing that the world is merely unordered and
on what is being predicted. This issue has unpredictable. In the former case, a poten
not been addressed in the use of the locus- tial for control exists, whereas in the latter
of-control scales. it does not. On this basis, Levenson devel
oped the Internal, Powerful Others, Chance
(IPC) Scales, which have been extensively
Cause, Effect, and Reciprocity
used. Similarly, Wallston, Wallston, and De
To what extent do locus-of-control beliefs de Vellis (1978) used these dimensions in their
termine attributional style, or to what extent revised multidimensional health locus-of-
are locus-of-control beliefs determined by control scale (Marshall, Collins, & Crooks,
experiences that shape attributions? Various 1990).
cycles of influence have been proposed such O ’Brien (1981) attempted to clarify the
that pessimistic or optimistic attributional issue further by introducing two positions be
styles are likely to become self-perpetuating. tween internal and external locus-of-control
Positive successful life experiences probably beliefs. He proposed four dimensions: inter
increase internal locus-of-control beliefs nals (who believe in internal control across
through optimistic attributions that, in turn, all situations), realists (whose internal and
may increase confidence, initiative, and mo external beliefs vary as a function of the do
tivation and lead to more successful experi main or situation they consider), structural
ences. The opposite may occur with nega ists (whose external beliefs stress societal de
tive, unsuccessful life experiences that leave terminants of behavior), and fatalists (who
people feeling at the mercy of powerful and see all outcomes as dependent on luck, fate,
hostile forces beyond their control, thereby or chance).
increasing external locus of control. Although several efforts have been made
to distinguish different varieties of exter
nal locus-of-control beliefs, few researchers
Methodological Issues
have attempted to subdivide the internal be
Three primary methodological issues apply lief pattern— Bradley, Brewin, Gamsu, and
to both the locus-of-control and attribution Moses (1984) and Furnham, Sadka, and
literatures. Brewin (1991) being exceptions. Internality
(or instrumentality) may result from either
effort or ability. That is, by exercising suf
Dimensionality
ficient salient effort, one may control out
A major issue concerns whether the locus-of- comes; or, simply by virtue of one’s ability,
control measure is uni- or multidimensional certain outcomes can be controlled.
278 IV. C O G N I T I V E D I S P O S I T I O N S
In their review of the extant locus-of- cesses and failures) in their lives (Hewstone,
control measures, Furnham and Steele (1993) 1989). At the same time, clinical psycholo
noted the ongoing proliferations of measures gists, inspired by cognitive theories of de
that often correlate only modestly with one pression, developed and extended learned
another. They also raised the fundamental helplessness theory (Abramson et al., 1978),
question of whether new scales show in which suggested that people develop a spe
cremental validity beyond that provided by cific, pervasive, and highly consequential at
existing scales. Furthermore, many of the tributional or explanatory style for making
measures have not considered some of the sense of what happens to them. The theory
theoretical distinctions, such as locus versus asserted that these styles could be efficiently
controllability and positive versus negative measured and, more important, changed
outcomes, that have been demonstrated to through cognitive behavior therapy.
be important to people’s beliefs about con Attribution or explanatory style is con
trol. Researchers who have recognized the cerned with how people explain or attribute
importance of these issues have tended to the causes of success and failure, happiness
move toward attributional style instruments and unhappiness, and other positive and
based on the work of Seligman (Abramson negative experiences in a fairly consistent
et al., 1978) rather than the relatively simple way. Early researchers tended to distinguish
unidimensional structure found in Rotter’s two styles that may reflect opposite ends of
(1966) scale. the same dimension, namely optimistic ver
A second, more practical issue concerns sus pessimistic attribution styles (sometimes
whether locus-of-control beliefs can be al called healthy vs. unhealthy). The early liter
tered by educational or therapeutic inter ature (from about 1980 to 2000) was mainly
ventions. Many researchers assume that concerned with pessimistic attribution styles
the purpose of designing locus-of-control and the causes and consequences of attribut
scales is to identify people whose beliefs ing negative outcomes in a particular way.
are maladaptive so that they can be helped. Nearly all the original work was in clini
Yet few researchers have discussed the issue cal psychology and was concerned with the
of whether locus-of-control beliefs can be cause and alleviation of depression. Howev
targeted by different treatments or how ef er, since the turn of the century, the interest
fective such interventions are. An extensive has swung to positive emotions, happiness,
literature in cognitive therapy addresses this and well-being.
issue, but it appears to have been ignored by
those who have designed locus-of-control
T h e Concept o f Style
measures.
Whereas the locus-of-control enterprise This literature deals with attribution and
has been primarily psychometric, the attri- explanatory style. The world style has dif
butional-style literature has been primarily ferent implications from terms such as trait,
clinical. Locus-of-control researchers have tem peram ent, or ability, all of which imply
been more interested in measurement, and greater stability and perhaps biological
attributional-style researchers more con or genetic determinants. Furnham (2008)
cerned with cognitive change. Thus, whereas pointed out a number of unanswered prob
research interest on locus of control appears lems with the concept of style. First, the
to have peaked, the same is not true of at question arises as to whether styles are bio
tribution or explanatory style. logically based, the result of early learning,
neither, or both. Aetiology determines both
how, and how much, a style may be changed
A ttribution and E xplanatory Style and therefore developed. Second is the issue
of variance accounted for. Specifically, is the
The concepts of attribution style and ex amount of variance accounted for by style
planatory style are used interchangably in factors so small as to be trivial, or are these
the literature. During the 1980s a great deal factors strongly related to people’s emotions
of research inspired by attribution theory and actions? Do styles have incremental
focused on the processes that underlie how validity over ability, personality, and value
people explain the events (particularly suc measures? Third is the nature of style as a
280 IV. C O G N IT IV E D IS P O S IT IO N S
F IG U R E 18.1. E xam p les o f attribu tions b ased on the basic three dim ensions.
more clearly linked with some types of de mixed, however. Some researchers suggest
pression (hopelessness) rather than others that the effect does not always replicate be
(endogenous) (Joiner, 2001). Overall, the cause of sample differences, range restric
results clearly and consistently show that at tions, or the low reliability of dependent
tribution style is one of the vulnerability fac variables. Others have suggested that other
tors for depression. The tendency to make factors need to be taken into consideration.
internal, stable, and global attributions for Academic success is influenced by students’
negative events predisposes people to experi ability (intelligence), personality, and learn
ence the symptoms associated with depres ing style, as well as their attribution styles
sion, such as passivity, negative affect, psy (Furnham, 2008). Indeed, researchers have
chosomatic problems associated with sleep demonstrated clear and predictable rela
and eating, and low self-esteem. tionships between these variables (Cheng &
Furthermore, changing attribution style Furnham, 2000).
is a (short-term) successful treatment for de
pression. Indeed, the great popularity and
Work Settings
widespread influence of cognitive therapy
and cognitive-behavioral therapy could be A good example of applied psychological
seen as the direct application of attribution- interest involves attribution style in work
style therapy to a range of conditions. What settings. In a seminal and often-cited paper,
is not clear, however, is how easy it is to Seligman and Schulman (1986) studied 94
maintain an optimistic as opposed to a pes experienced life insurance sales agents who
simistic attribution style and how style inter as a result of their jobs repeatedly encounter
acts with other vulnerability factors. failure, rejection, and indifference from pro
Studies have shown genetic influence on spective clients. In testing the link between
attribution style that has been thought to explanatory style and work productivity
mediate genetic influences on depression and quitting, they found that, as predicted,
(Lau, Rijsdijk, & Eley, 20 0 6 ). Other studies agents who saw failure as caused by inter
have pointed to the instability of attribution nal, stable, and global factors initiated fewer
style. Indeed, Ball, M cGuffin, and Farmer sales attempts, were less persistent, produced
(2008) suggested that attribution style is re less, and quit more frequently than those
ally little more than a mood state that does with a more optimistic explanatory style.
not reflect a risk factor for depression. They The results also showed that agents who had
noted that “the way in which individuals at optimistic explanatory styles sold 37% more
tribute their experiences may be less of a risk insurance in their fi rst 2 years of service than
factor and more of a symptom of depression those with pessimistic styles. More striking
than previously thought. Past episodes of de ly, agents in the top decile sold 88% more
pression may produce long-lasting negative insurance than those in the bottom decile.
attributions relating to the self, in addition In a prospective 1-year study of 103 newly
to other pessimistic attributions that are hired agents, individuals who had optimistic
linked to both observed and self reported explanatory styles when hired remained in
current depression. It is therefore important their jobs twice as long and sold more in
to look out for and address these pessimistic surance than agents with more pessimistic
attributions in people with recurrent depres explanatory styles (Seligman & Shulman,
sion.” (p. 278). 1986). Corr and Gray (1996) replicated this
finding, and Furnham and colleagues (1991)
showed that attribution style was linked, as
Academic Performance
predicted, to job satisfaction and motiva
Peterson and Barrett (1987) published a tion. Furthermore, Furnham, Brewin, and
longitudinal study that demonstrated that O ’Kelly (1994) found that it also predicted
students’ academic explanatory styles (i.e., job involvement and organizational commit
why they performed as they did in college) ment. More recently, Ashforth and Fugate
measured 2 weeks into the academic year (2006) showed a link between attributional
predicted their grades at the end of the year. style and work adjustment. In light of such
Efforts to replicate this finding have been findings, organizational psychologists have
18. L o cu s o f C o n tro l and A ttrib u tio n Style 283
2. In the future, at work, will this cause again influence what happens?
Will never again influence what 6 5 4 3 2 1 Will always influence what happens
happens
3. Is the cause something that just affects problem solving or does it influence other areas of your
life?
8. To what extent do you think you could have foreseen the cause?
F I G U R E 18.2. A rating scale for each cau se given by a particip an t for an event.
some aspects of theoretical clarity and mea trol or attributional style. Indeed, both are
surement, clinical practitioners have em seen as learned individual differences that
braced the ideas and developed training and can and sometimes should be unlearned.
therapeutic interventions that build on these There is still a discomfort at the heart of the
concepts. On the whole, differential psychol locus-of-control and attribution-style litera
ogists interested in one of the two worlds tures about the conceptualization of style as
of ability and personality have shown less stable and sometimes resistant to change.
interest in the other concept. This is not to The psychometrician looking at this lit
imply that there have not been efforts to lo erature may be disturbed by scale prolifera
cate locus of control in “five factor space” or tion that has been fueled by the belief that
at Carroll’s (1993) level II of intelligence but domain-specific measures perform better
rather that trait psychologists have not been than general ones. Dozens of scales have ap
as interested in these concepts. There may peared and then disappeared after one or
be many reasons for this. For example, both two studies that were in some senses little
constructs originate from social learning more than pilot studies of a new question
theory, which, at least traditionally, did not naire. Although it is easy to demonstrate
favor ideas of stable individual differences. face and concurrent validity, the develop
In addition, there have been few attempts to ment of a reliable and robust measure needs
look at biological markers of locus of con to demonstrate predictive and incremental
18. L o cu s o f C o n tro l and A ttrib u tio n Style 285
Furnham, A., Sadka, V., & Brewin, C. (1991). The McAuley, E., Duncan, T. E., 8c Russell, D. W. (1992).
development of an occupational attributional style Measuring causal attributions: T he Revised Causal
questionnaire. J o u r n a l o f O r g a n is a tio n a l B e h a v Dimension Scale (CDSII). P erson ality a n d S o c ia l
iou r, 13, 2 7 - 3 9 . P sy ch o lo g y B u lletin , 18, 5 6 6 - 5 7 3 .
Furnham, A., 8c Steele, FI. (1993). Measuring locus McAuley, E., 8c Gross, J. (1983). Perceptions of causal
of control. B ritish J o u r n a l o f P sy ch olog y , 8 4, 4 4 3 - ity in sport. J o u r n a l o f S p o rt P sy ch o lo g y , 5, 7 2 - 7 6 .
479. Maier, S., 8c Seligman, M . (1976). Learned helpless
Goggin, K., Murray, T., Malcarne, V., Brown, S., 8c ness: Theory and evidence. J o u r n a l o f E x p e r im e n
Wallston, K. (200 7). Do religious and control cog tal P sy ch o lo g y , 105, 3 - 4 6 .
nitions predict risky behaviour?: I. Development Marshall, G., Collins, B., 8c Crooks, V. (1990). A
and validation of the Alcohol-Related God Locus- comparison of two multidimensional health locus
of-Control Scale for Adolescents (AGLOC-A). C o g of control instruments. J o u r n a l o f P erso n a lity A s
n itiv e R e se a rc h a n d T h era p y , 3 1, 1 1 1 - 1 2 2 . sessm en t, 5 4 , 1 8 1 - 1 9 0 .
Gordon, R. (2008). Attribution style and athletic per Messick, S. (1994). The matter of style: Manifestations
formance. P sy c h o lo g y o f S p o rt a n d E x ercis e, 9, of personality in cognition, learning and teaching.
336-350. E d u c a tio n a l P sy ch o lo g ist, 2 9 , 1 2 1 - 1 3 6 .
Gregory, W. (1978). Locus of control for positive and Mirels, H. (1970). Dimensions of internal vs. external
negative outcomes. J o u r n a l o f P ers o n a lity a n d S o control. J o u r n a l o f C o n su ltin g a n d C lin ica l P sy
c ia l P sy c h o lo g y , 3 6 , 8 4 0 - 8 4 9 . c h o lo g y , 3 4 , 2 2 6 - 2 2 8 .
Gurin, P., Gurin, G., Lao, R., 8c Beattie, M . (1969). O ’ Brien, G. (1981). Locus of control, work, and retire
Internal-external control in the motivation dynamics ment. In H. Lefcourt (Ed.), R e se a rc h w ith th e lo cu s
of Negro youth .J o u r n a l o f S o cia l Issu es, 2 5 , 2 9 - 5 3 . o f c o n t r o l c o n str u c t (Vol. 3, pp. 7 -71 ). London:
Hanrahan, S., Grove, J. R., 8c Hattie, J . A. (1989). Academic Press.
Development of a questionnaire measure of sport- Paulhus, D., 8c Christie, R. (1981). Spheres of control:
related attributional style. I n te r n a tio n a l J o u r n a l o f An interactionist approach to assessment of per
S p o rt P sy ch olog y , 2 0 , 1 4 4 - 1 3 4 . ceived control. In H. Lefcourt (Ed.), R e se a rch w ith
Hewstone, M . (1989). C a u s a l a ttr ib u tio n . Oxford, th e lo c u s o f c o n t r o l co n str u c t (Vol. 1, pp. 161-188).
UK: Blackwell. New York: Academic Press.
Flodgkinson, G. (1992). Research notes and communi Peterson, C., 8c Barrett, L. C. (1987). Explanatory
cations development and validation of the strategic style and academic performance among university
locus of control scale. M a n a g e m en t J o u r n a l, 13, freshmen. J o u r n a l o f P erson ality a n d S o c ia l P sy
31 1-3 1 7. c h o lo g y , 5 3, 6 0 3 - 6 0 7 .
Holt, C., Clark, E., Kreuter, M ., 8c Rubio, D. (2003). Peterson, C., Semmel, A., Von Baeyer, C., Abramson,
Spiritual health locus of control and breast cancer L., Metalsky, G., 8c Seligman, M. (1982). The At
beliefs among African American women. H ea lth tributional Style Questionnaire. C o g n itiv e T h era p y
P sy ch olog y , 2 2 , 2 9 4 - 2 9 9 . a n d R esea rch , 6 , 2 8 1 - 3 0 0 .
Joiner, T. (2001). Negative attribution style, hopeless Roberts, L., 8c Ho, R. (1996). Development of an Aus
ness, depression and endogenous depression. B e tralian health locus of control scale. P erson ality a n d
h a v io u r a l R e se a rc h a n d T h era p y , 3 9, 1 39 -1 4 9 . In d iv id u a l D iffe r e n c e s , 2 0 , 6 2 9 - 6 3 9 .
Jones, J . , 8c Wuebker, L. (1985). Development and Rothbaum, F., Weisz, J ., 8c Snyder, R. (1982). Chang
validation of the safety locus of control scale. P er ing the world and changing the self: A two process
c e p tu a l a n d M o to r S kills, 61, 15 1 -16 1. model of perceived control. J o u r n a l o f P erson ality
Kent, R. L., 8c M artinko, M . J. (1995). T he develop a n d S o c ia l P sy ch olog y , 4 2 , 5 - 3 7 .
ment and evaluation of a scale to measure organiza Rotter, J . (1966). Generalised expectancies for internal
tional attributional style. In M. J . M a rtin ko (Ed.), versus external control of reinforcement. P sy c h o
A ttrib u tio n th e o r y : An o r g a n iz a tio n a l p er s p e c tiv e lo g ic a l M o n o g ra p h s, 80(1), Whole No. 609.
(pp. 5 3 - 7 5 ) . Delray Beach, FL: St. Lucie Press. Rotter, J. (1975). Some problems and misconceptions
Kinderman, P., 8c Bentall, R. (1997). Causal attri related to the construct of internal versus external
butions in paranoia: Internal, personal, and situ control of reinforcement. J o u r n a l o f C o n su ltin g a n d
ational attributions for negative events. J o u r n a l o f C lin ic a l P sy ch olog y , 4 3 , 5 6 - 6 7 .
A b n o r m a l P sy ch olog y , 106, 3 4 1 - 3 4 5 . Rotter, J . (1990). Internal versus external control of re
Kneebone, I., 8c Dunmore, E. (2004). Attribution style inforcement: A case history of a variable. A m erica n
and symptoms of depression in persons with mul P sy ch o lo g ist, 4 5, 4 8 9 - 4 9 3 .
tiple sclerosis. I n te r n a tio n a l J o u r n a l o f B e h a v io u r a l Sanger, S., 8c Walker, H. (1972). Dimensions of
M ed icin e, 11, 1 1 0 - 1 1 5 . internal-extern al control and the women’s libera
Lau, J . , Rijsdijk, F., 8c Eley, T. ( 2 0 06 ). 1 think therefore tion movement. J o u r n a l o f S o c ia l Issu es, 2 8 , 115—
I am: A twin study of attributional style in adoles 129.
cents. J o u r n a l o f C h ild P sy ch o lo g y a n d P sychiatry, Schmidt, F., 8c Gifford, R. (1989). A dispositional ap
47, 6 9 6 - 7 0 3 . proach to hazard perception. J o u r n a l o f E n v ir o n
Lefcourt, H. (1991). Locus of control. In J. Robinson, m e n ta l P sy ch o lo g y , 9, 5 7 - 6 7 .
P. Shaver, 8c L. Wrightsman (Eds.), M easu res o f Schneider, J ., 8c Parsons, O. (1970). Categories on the
p e r s o n a lity a n d s o c ia l p s y c h o lo g ic a l a ttitu d es (Vol. locus of control scale and cross-cultural compari
1, pp. 4 1 3 - 4 9 9 ) . New York: Academic Press. sons in Denmark and the United States. J o u r n a l o f
Levenson, H. (1981). Differentiating among internal C ro ss-C u ltu ra l P sy ch olog y , 2 , 1 3 1 - 1 3 8 .
ly powerful others and chance. In H. M. Lefcourt Seligman, M. (1990). L e a r n e d o p tim is m . New York:
(Ed.), R e se a rc h w ith th e lo c u s o f c o n t r o l c o n str u c t Pocket Books.
(Vol. 1, pp. 1 5 - 6 3 ) . New York: Academic Press. Seligman, M., 8c Schulman, P. (1986). Exploratory
18. L o cu s o f C o n tro l and A ttrib u tio n Style 287
style as a predictor of productivity and quitting Wallston, K., Wallston, B., & De Vellis, R. (1978). De
among life insurance sales agents. J o u r n a l o f P er velopment of the multidimensional health locus of
so n a lity a n d S o c ia l P sy ch olog y , 5 0 , 8 3 2 - 8 3 0 . control (M H L C ) scales. H ea lth E d u c a tio n M o n o
Sternberg, R., & Grigorenko, E. (1997). Are cogni g ra p h s, 6, 1 6 0 - 1 6 9 .
tive styles still in style? A m erica n P sy ch o lo g ist, 5 2 , Wallston, K. A., Malcarne, V. L., Flores, L., Hansdot-
700-712. tir, I., Smith, C. A., Stein, M. J ., et al. (1999). Does
Toner, M ., & Heaven, P. (2005). Peer social attribu God determine your health?: The God Locus of
tional predictors of socio-economic adjustment in Health Control Scale. C o g n itiv e T h era p y a n d R e
early adolescence. P erso n a lity a n d In d iv id u a l D if sea rch , 2 3 , 1 3 1 - 1 4 2 .
fe r e n c e s , 3 8 , 5 7 9 - 5 9 0 . Weiner, B. (1980). H u m an m o tiv a tio n . New York:
Wagner, K., Berenson, A., Harding, O., & Joiner, T. Holt, Rinehart & Winston.
(1998). Attribution style and depression in pregnant Weisz, J ., Rothbaum, F., &c Blackburn, T. (1984).
teenagers. A m eric a n J o u r n a l o f P sy ch iatry , 155, Standing out and standing in: The psychology of
1227-1233. control in America and Japan. A m er ica n P s y c h o lo
Wallston, K., & Wallston, B. (1981). Health locus of gist, 3 9, 9 5 5 - 9 5 6 .
control scales. In H. Lefcourt (Ed.), R e se a rch w ith Xenikou, A., Furnham, A., & McCarrey, M. (1997).
th e lo c u s o f c o n tr o l c o n str u c t (Vol. l , p p . 189-2 41 ). Attribution style for negative events. B ritish J o u r n a l
New York: Academic Press. o f P sy ch olog y , 8 8 , 5 3 - 6 9 .
CHAPTER 19
■ •A . ♦ ■ * A ♦
C l a u d ia D a l b e r t
288
19. B e lie f in a Ju st W orld 289
only when they were certain that compensa Thus, the stronger the belief in a just world,
tion would be given that the injustice was the stronger the justice motive. Experimen
no longer assimilated. This innocent-victim tal just-world research typically does not
paradigm remains the most influential in assess individual differences, however, but
modern experimental just-world research; interprets experimental reactions in the light
it is only the type of innocent victim that of just-world reasoning. Such research thus
has changed (e.g., Correia, Vala, & Aguiar, addresses justice motivation, and not the jus
2007). tice motive as an individual-differences dis
position. Motivation can be defined as a per
son’s orientation toward a specific goal in a
T h e B e lie f in a fu s t World as a Disposition
specific situational state; thus justice motiva
A substantial amount of research on belief in tion means the orientation toward justice in
a just world has been experimental in nature a given situation. Justice motivation is trig
(for a review, see Hafer & Begue, 2005), fo gered by specific situational circumstances
cusing primarily on the maladaptive func in interaction with personal dispositions. In
tions of the belief in a just world, such as the case of justice motivation, that person
disdain for the victim. Since the 1970s, al disposition may be the justice motive or
however, another strand of research has ex other dispositions (e.g., Lind & van den Bos,
amined individual differences in the belief 2 0 0 2 ; Miller, 1999).
in a just world and found that it also serves
important adaptive functions (for a review,
Differentiation of the B elief
see Furnham, 2003). This research agenda
in a Just-W orld Disposition
was triggered by the introduction of the first
Belief in a Just World Scale by Rubin and Since the 1990s, more studies have inves
Peplau (1973, 1975), which assessed individ tigated the positive as well as the negative
ual differences in the belief that the world is social consequences of the belief in a just
generally a just place. This approach allowed world, and the focus of these investigations
the role of the belief in a just world to be in has been extended to cover the consequenc
vestigated within the framework of person es of holding a belief in a just world for the
ality dispositions, and positive associations believers. Based on suggestions originating
were found particularly with authoritarian from earlier research (Furnham & Procter,
ism and internal locus of control (for a re 1989; Lerner &c Miller, 1978), these studies
view, see Furnham & Procter, 1989). have shown that it is necessary to distinguish
the belief in a personal just world, in which
one is usually treated fairly, from the belief
justice Motive versus Justice Motivation
in a general just world, or the belief in a just
In the context of just-world research and world for others in which people in gen
theory, scholars often speak of the justice eral get what they deserve (Dalbert, 1999;
motive (e.g., Ross & M iller, 2002). The shift Lipkus, Dalbert, & Siegler, 1996). In line
from the experimental to the individual- with the self-serving bias in general (Taylor,
differences approach to the belief in a just Wright, Moghaddam, & Lalonde, 1990)
world made it necessary to differentiate and with fairness reasoning in particular
between a justice motive and justice moti (Messick, Bloom, Boldizar, & Samuelson,
vation. Motives are individual dispositions 1985), research evidenced that people tend
reflecting individual differences in the ten to endorse the personal more strongly than
dency to strive for a specific goal. A justice the general belief in a just world and that
motive is thus an individual disposition to the two constructs have a different meaning.
strive for justice as an end in itself. Accord The personal belief in a just world is a better
ing to Lerner (1977), the individual belief in predictor of adaptive outcomes (e.g., subjec
a just world can be interpreted as an indica tive well-being), and the belief in a just world
tor of such a justice motive. The belief in a for others or in general is a better predictor,
just world indicates a personal contract; the for example, of harsh social attitudes (e.g.,
more people want to rely on being treated Begue & Muller, 2006).
justly by others, the more obligated they O f course, other differentiations of the
should feel to behave justly themselves. just-world construct have also been pro
290 IV. C O G N IT IV E D IS P O S IT IO N S
posed. To give just two examples for the gen associated with the Rubin and Peplau scale
eral just-world belief: A general belief in im and has also been successfully implemented
manent justice has been distinguished from in several studies (e.g., Hafer, 20 0 0 ). All
a general belief in ultimate justice (Maes & three general just-world scales are positively
Kals, 2 0 0 2 ), and a general belief in distribu correlated with each other (Lipkus et al.,
tive justice has been distinguished from a 1996). Surprisingly, however, although there
general belief in procedural justice (Lucas, are at least two homogenous, short, and
Alexander, Firestone, & LeBreton, 2007). valid measures of general just-world belief,
Finally, the general belief in a just world has the 20-item Rubin and Peplau scale is still
been differentiated from the general belief in use (e.g., Edlund, Sagarin, & Johnson,
in an un\\ist world (Dalbert, Lipkus, Sallay, 2007). Finally, in line with the differentia
& Goch, 2 0 0 1 ; Loo, 2002). This research tion of the just-world construct, Lipkus and
showed that general belief in a just world colleagues (1996) and Dalbert (1999) intro
should not be seen as a bipolar construct duced reliable scales differentiating the be
but as a two-dimensional one. Because the lief in a just world for others or, in general,
differentiation between a more general and from the belief in a personal just world.
a more personal just-world belief thus far
seems to be the most widespread and well-
B e lie f in a Ju s t World
examined distinction, however, this sum
and O ther Personality Dispositions
mary focuses on research on general and
personal just-world beliefs. One of the first associations observed be
tween the belief in a just world and other
personality dispositions was the positive
M easures o f the B e lie f in a Ju s t World
correlation between general just-world belief
The starting point for individual-differences and religiosity (Dalbert & Katona-Sallay,
research on the belief in a just world was 1996; Rubin & Peplau, 1973). Research on
Rubin and Peplau’s (1975) 20-item Belief the differences between the two has con
in a Just World Scale (sample items: “Basi firmed that they are distinct dispositions
cally, the world is a just place”; “Men who (e.g., Hui, Chan, &C Chan, 1989), and cross-
keep in shape have little chance of suffering cultural research has found few differences
a heart attack”; “Good deeds often go unno in the just-world belief across cultures with
ticed and unrewarded”). This scale was later contrasting religious and political back
criticized as being heterogeneous in content grounds (e.g., Furnham, 1993). A positive
(e.g., Furnham & Procter, 1989): It includ and sometimes substantial association has
ed both general and domain-specific items, also been found between authoritarianism
as well as items on the belief in an unjust and general just-world belief (for a review,
world and items tapping other constructs, see Furnham & Procter, 1989). Analyses
such as authoritarianism (e.g., “When par of the common factor structure of the two
ents punish their children it is almost always constructs support the two-factor hypoth
for good reasons”). Consequently, some re esis and their differential meaning, with
searchers used a subsample of the general the belief in a just world providing a more
items to assess general just-world belief (e.g., positive outlook than authoritarianism (Dal
Steensma & van Dijke, 2 0 0 6 ). In the light bert, 1992; Lerner, 1978). The positive as
of these criticisms, two homogenous general sociations repeatedly observed between just-
just-world scales were developed. Dalbert, world belief and internal locus o f control
Montada, and Schmitt (1987) constructed a have prompted speculation about an overlap
homogenous six-item scale tapping general between these two constructs as well (for a
belief in a just world (sample item: “I think review, see Furnham &C Procter, 1989). The
people try to be fair when making important two constructs should be distinct from a the
decisions”), which shows convergent valid oretical perspective, however. The belief in
ity with the Rubin and Peplau scale, is in personal agency is consistent with the belief
dependent of social desirability (Loo, 2002), in a just world as long as the justice principle
and has been used in numerous studies (e.g., endorsed is the equity norm. Other ideas of
Allen, Ng, & Leiser, 2005). In addition, Lip justice (e.g., the equality or the need principle
kus (1991) constructed a seven-item Global of justice or the belief in a just God) are not
Belief in a Just World Scale that is positively consistent with the belief in internal control.
19. B e lie f in a Ju st W orld 291
and events in their own lives as being more are expected to be honest with one another,
just. For example, school students with a and people who have been deceived may con
strong belief in a personal just world have clude that they deserved it in some way. It
been found to be more likely to evaluate their can thus be hypothesized that people with a
school grades and their teachers’, peers’, and strong just-world belief prefer not to think
parents’ behavior toward them as just (Cor they have been deceived or taken advantage
reia & Dalbert, 2 0 0 7 ; Dalbert & Stoeber, of. Research has shown the expected positive
200 6 ). Similarly, prisoners with strong per association of just-world belief with general
sonal just-world beliefs are more likely to interpersonal trust (e.g., Begue, 2 0 0 2 ; Zuck
evaluate the justice of the legal proceedings erman & Gerbasi, 1977), trust in societal in
leading to their conviction, the treatment by stitutions (Correia & Vala, 20 0 4 ), and young
their prison officers, and decisions on prison adolescents’ trust in the justice of their future
affairs as more just (Dalbert & Filke, 2007; workplace (Sallay, 2004). This trust in future
O tto & Dalbert, 2005). justice has a number of implications.
The personal just-world belief is usually
seen as a personal disposition, but results in
R isk Perception
dicating a causal effect of justice experiences
on the belief in a just world qualify this as Individuals with a strong just-world belief
sumption. Research has shown that justice are convinced that good things happen to
experiences in the school and the family good people and that bad things happen to
modify the personal just-world belief (Dal bad people. Because individuals tend to think
bert & Stoeber, 2 0 0 6 ) and that factors such of themselves as good people (e.g., Brown,
as length of imprisonment (Otto & Dal 1986; Messick et al., 1985), the belief in a
bert, 2 0 0 5 ), monotony at work, and mob just world can be expected to give them an
bing experiences at work (Cubela Adoric optimistic outlook on the future. This buff
&C Kvartuc, 20 0 7 ; Dzuka &C Dalbert, 2007; ering effect is expected to be particularly
O tto & Schmidt, 2 0 0 7 ) are negatively relat evident when people are threatened by un
ed to the personal just-world belief. Thus the fairness. Lambert, Burroughs, and Nguyen
belief in a personal just world must be seen (1999) were the first to study the meaning of
as a partly experiential construct (Maes & the belief in a just world for risk perception
Schmitt, 200 4 ). Nevertheless, an unambigu and showed that the just-world belief seems
ous pattern of results clearly indicates that to enable fearful individuals (i.e., those
a strong personal just-world belief leads to high in authoritarianism) to be confident of
events being evaluated as just. Cubela Ador avoiding an unjust fate. It is particularly im
ic and Kvartuc (2007) have suggested that portant for individuals exposed to external
injustice experiences only affect the belief in risks (i.e., those perceived to be controlled
a just world when they reach a specific de by others or by fate; e.g., robbery) rather
gree of adversity. Further studies are needed than to internal risks (i.e., those that are
to determine under which conditions the under their personal control; e.g., suicide) to
just-world belief fosters the assimilation of be able to rely on the environment being fair.
injustice and under which conditions injus Indeed, Dalbert (2001) found that the buff
tice can no longer be assimilated but instead ering effect of the general just-world belief
undermines the belief in a just world. for fearful individuals held only for external
risks, not for internal risks. Finally, Hafer,
Bogaert, and M cMullen (2001) found that
B e lie f in a Ju s t World
individuals with strong general just-world
and the Trust in Justice
beliefs but low in interpersonal control seem
People with a strong belief in a just world to put themselves at greater risk, presumably
are thought to be confident in being treated as a consequence of a lowered risk percep
justly by others, and it is this trust in particu tion. In sum, the general just-world belief
lar that is hypothesized to give the just-world appears to function as a buffer against the
belief the character of a resource in everyday perception of external risk for those who
life. Assuming that people get what they de tend to need such a buffer, but this mecha
serve, they will be punished for deceiving nism may result in higher exposure to risks
others. Accordingly, in a just world, people in reality.
19. B e lie f in a Ju st W orld 293
just-world beliefs and subjective well-being. adults have no difficulty in identifying ran
Moreover, research has shown that the be dom events. Nevertheless, they sense that a
lief in a personal just world is more impor random fate is unjust, and when given the
tant than the general just-world belief in ex possibility to justify a random fate, they will
plaining well-being (Dalbert, 1999; Lipkus do so (e.g., Jose, 1990; Weisz, 1980). Chil
et al., 1996; O tto, Boos, Dalbert, Schops, & dren thus develop a belief in a just world—
Hoyer, 2 0 0 6 ; Sutton & Douglas, 2005) and which can be interpreted as a more mature
that this positive association between just- version of the belief in immanent justice—
world belief and well-being is true for non the belief that people generally deserve their
victims (e.g., Dzuka & Dalbert, 2 0 0 6 ; Otto fate accompanied by the cognitive ability to
& Schmidt, 20 0 7 ; Ritter, Benson, & Snyder, identify causality and randomness (Raman
1990) and for various groups of victims (e.g., & Winer, 2004).
Agrawal & Dalai, 1993; Bulman & Wort- During adolescence, personal and gen
man, 1977; O tto et al., 20 0 6 ). In addition, eral just-world beliefs emerge as two dis
Dzuka and Dalbert (2007) demonstrated tinct beliefs. The strength of both beliefs
that teachers’ well-being was positively as seems to decrease slightly during adoles
sociated with their beliefs in a personal just cence and young adulthood. Both of these
world and that this relationship held when developmental changes— differentiation and
exposure to student violence was controlled. decline— can be interpreted as consequences
This study is one of the few to have found of increasing cognitive maturity. Even after
evidence for a buffering effect of the just- its initial decline, the belief in a personal just
world belief: It was only among teachers world tends to be rather strong. The strength
with weak personal just-world beliefs that of just-world belief seems to increase again
exposure to violence was associated with slightly in late adulthood and old age (e.g.,
more negative affect; exposure to violence Dalbert, 20 0 1 ; Maes & Schmitt, 2004).
did not explain negative affect among those The meaning of the just-world belief also
with strong personal just-world beliefs. seems to differ systematically across the
A personal resource can be defined as a lifespan (Maes & Schmitt, 2004). In ado
personal disposition that helps people to lescence and young adulthood, especially,
cope with the events of their daily lives. The the just-world belief’s main function seems
stronger the resource, the better equipped to be to provide trust in the fairness of the
they are to cope. A personal resource thus world, thus enabling people to master chal
implies a main-effect hypothesis. A personal lenges in school and at the workplace and
buffer, in contrast, is usually seen as a re to invest in their personal goals. In old age,
source that takes effect only under specific when the remaining lifetime is shorter, the
adverse conditions. A buffer thus implies a just-world belief’s primary function seems
moderator hypothesis; the buffer moderates to be to provide a framework to help people
the association between strain and outcome. interpret the events of their lives in a mean
Overall, research findings are very much in ingful way. A strong just-world belief allows
line with the resource hypothesis and do not older adults to see themselves as having been
support the buffer hypothesis. The belief in less discriminated against during the course
a personal just world should thus be seen as of their lives, prevents them from ruminat
a personal resource helping to sustain the ing about the negative aspects of their lives,
well-being of people of all ages in diverse and instead enables them to find meaning in
situations, victims and nonvictims alike. them.
To explore the development of individual
differences in the just-world belief, studies
T h e D evelopm ental Trajectories o f B e lie f
have investigated the impact of parenting on
in a Ju s t World
the just-world belief. In adolescence, at least
Until the age of 7 or 8, children typically (cf. Schonpflug & Bilz, 20 0 4 ), there does
believe in immanent justice, and they are not seem to be direct transmission from par
convinced that wrongdoings are automati ent to child; however, parenting styles have
cally punished (Piaget, 1932/1997). As they proved to be positively associated with the
grow older, however, they slowly abandon children’s just-world belief (e.g., Dalbert &
this belief in immanent justice. As a result Radant, 2004). Nurture, as reflected by a
of cognitive development, older children and harmonious family climate with a low rate
19. B e lie f in a Ju st W orld 295
Dalbert, C. (1999). The world is more just for me than progress over the past decade. P erson ality a n d In d i
generally: About the Personal Belief in a Just World v id u al D iffe r e n c e s , 3 4 , 7 9 5 - 8 1 7 .
Scale’s validity. S o c ia l Ju s t ic e R e s e a r c h , 1 2, 7 9 - 9 8 . Furnham, A., & Procter, E. (1989). Belief in a just
Dalbert, C. (2001). T h e ju stic e m o tiv e a s a p e r s o n a l world: Review and critique of the individual differ
re so u r c e: D ea lin g w ith ch a llen g e s a n d c ritic a l life ence literature. B ritish J o u r n a l o f S o c ia l P sy c h o l
ev en ts. New York: Kluwer Academic/Plenum. ogy, 2 8 , 3 6 5 - 3 8 4 .
Dalbert, C., & Dzuka, J . ( 2 0 0 4 ). Belief in a just Hafer, C. L. ( 2 00 0). Investment in long-term goals and
world, personality, and well-being of adolescents. commitment to just means drive the need to believe
In C. Dalbert & H. Sallay (Eds.), T h e ju s tic e m o in a just world. P erso n a lity a n d S o c ia l P sy ch o lo g y
tiv e in a d o le s c e n c e a n d y o u n g a d u lt h o o d : O rig in s B u lletin , 2 6 , 1 0 5 9 - 1 0 7 3 .
a n d c o n s e q u e n c e s (pp. 1 0 1 - 1 1 6 ). London: Rout- Hafer, C. L., & Begue, L. (20 05 ). Experimental re
ledge. search on just-world theory: Problems, develop
Dalbert, C., & Filke, E. ( 2 00 7). Belief in a just world, ment, and future challenges. P s y c h o lo g ic a l B u lletin ,
justice judgments, and their functions for prisoners. 131, 1 2 8-1 6 7 .
C r im in a l J u s t ic e a n d B eh a v io r, 3 4 , 1 5 1 6 -1 5 2 7 . Hafer, C. L., Bogaert, A. F., & McMullen, S. L. (2001).
Dalbert, C., & Katona-Sallay, H. (1996). The “belief Belief in a just world and condom use in a sample
in a just world” construct in Hungary. J o u r n a l o f of gay and bisexual men. J o u r n a l o f A p p lie d S o c ia l
C ro ss-C u ltu ra l P sy ch olog y , 27, 2 9 3 - 3 1 4 . P sy ch olog y , 3 1, 1 8 9 2 - 1 9 1 0 .
Dalbert, C., Lipkus, I. M ., Sallay, H., & G och, I. Hafer, C. L., & Correy, B. L. (1999). Mediators of
(2001). A just and an unjust world: Structure and the relation of beliefs in a just world and emotional
validity of different world beliefs. P erson ality a n d responses to negative outcomes. S o c ia l J u s t ic e R e
In d iv id u a l D iffe r e n c e s , 3 0 , 5 6 1 - 5 7 7 . sea rch , 1 2, 1 8 9 - 2 0 4 .
Dalbert, C., Montada, L., & Schmitt, M . (1987). Hui, C. H., Chan, I. S. Y., & Chan, J . (1989). Death
Glaube an eine gerechte Welt als Motiv: Vali- cognition among Chinese teenagers: Beliefs about
dierungskorrelate zweier Skalen [The belief in a just consequences of death. J o u r n a l o f R e se a rc h in P er
world as a motive: Validity correlates of two scales], so n a lity , 2 3 , 99 - 1 1 7 .
P s y c h o lo g is c h e B eitrd g e, 2 9, 5 9 6 - 6 1 5 . Jose, P. E. (1990). Just world reasoning in children’s
Dalbert, C., & Radant, M. (200 4). Parenting and immanent justice judgements. C h ild D ev e lo p m en t,
young adolescents’ belief in a just world. In C. Dal 61, 1 0 2 4 - 1 0 3 3 .
bert 8c H. Sallay (Eds.), T h e ju stice m o tiv e in a d o Jo st, J. T„ Banaji, M . R., & Nosek, B. A. (20 04 ). A
le s c e n c e a n d y ou n g a d u lt h o o d : O rigin s a n d c o n s e decade of system justification theory: Accumulated
q u e n c e s (pp. 1 1 -25 ). London: Routledge. evidence of conscious and unconscious bolstering of
Dalbert, C., & Stoeber, J. (200 5). The belief in a just the status quo. P o litica l P sy c h o lo g y , 2 5 , 8 8 1 - 9 1 9 .
world and distress at school. S o c ia l P sy ch o lo g y o f Kiecolt-Glaser, J. K., & Williams, D. A. (1987). Self
E d u c a tio n , 8, 1 2 3 - 1 3 5 . blame, compliance, and distress among burn pa
Dalbert, C., & Stoeber, J . ( 2 0 06 ). The personal belief tients. J o u r n a l o f P erson ality a n d S o c ia l P sy c h o l
in a just world and domain-specific beliefs about og y , 5 3 , 1 8 7 - 1 9 3 .
justice at school and in the family: A longitudinal Lambert, A. J., Burroughs, T., & Nguyen, T. (1999).
study with adolescents. In te r n a tio n a l J o u r n a l o f Perceptions of risk and the buffering hypothesis:
B e h a v io r a l D e v e lo p m e n t, 3 0 , 2 0 0 - 2 0 7 . The role of just world beliefs and right-wing au
DePalma, M ., Madey, S. F., Tillm an, T. C., & Wheeler, thoritarianism. P erson ality a n d S o c ia l P sy ch o lo g y
J. (1999). Perceived patient responsibility and belief B u lletin , 2 5 , 6 4 3 - 6 5 6 .
in a just world affect helping. B a sic a n d A p p lie d S o Lerner, M . J . (1965). Evaluation of performance as a
c ia l P sy ch olog y , 2 1 , 131-137. function of performer’s reward and attractiveness.
Dette, D., Stober, J . , & Dalbert, C. (20 04 ). Belief in J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 1,
a just world and adolescents’ vocational and social 355-360.
goals. In C. Dalbert & H. Sallay (Eds.), T h e ju s Lerner, M . J . (1977). T he justice motive: Some hypoth
tic e m o tiv e in a d o le s c e n c e a n d y o u n g a d u lt h o o d : eses as to its origins and forms. J o u r n a l o f P e r s o n a l
O rig in s a n d c o n s e q u e n c e s (pp. 11—25) . London: ity, 4 5 , 1 - 5 2 .
Routledge. Lerner, M. J. (1978). “Belief in a just world” versus the
Dzuka, J., & Dalbert, C. ( 2 0 0 6). The belief in a just “authoritarianism” syndrome . . . but nobody liked
world’s impact on subjective well-being in old age. the Indians. E th n icity , 5, 2 2 9 - 2 3 7 .
A g in g a n d M en ta l H ea lth , 10, 4 3 9 - 4 4 4 . Lerner, M . J. (1980). T h e b e l i e f in a ju st w o r ld : A fu n
Dzuka, J ., & Dalbert, C. ( 2 0 07 ). Student violence d a m e n t a l d elu s io n . New York: Plenum Press.
against teachers: Teachers’ well-being and the be Lerner, M . J ., & Goldberg, J . H. (1999). When do de
lief in a just world. E u r o p ea n P sy ch olog ist, 12, cent people blame victims?: The differing effects of
253-260. the explicit/rational and implicit/experiential cogni
Edlund, J . E., Sagarin, B. J ., & Joh nson, B. S. (200 7). tive systems. In S. Chaiken & Y. Trope (Eds.), D u a l
Reciprocity and the belief in a just world. P ers o n a l p r o c e s s th e o r ie s in s o c ia l p s y c h o lo g y (pp. 6 2 7 - 6 4 0 ) .
ity a n d In d iv id u a l D iffer en c es , 4 3 , 5 8 9 - 5 9 6 . New York: Guilford Press.
Fetchenhauer, D., Jaco bs, G., & Belschak, F. (2005). Lerner, M. J ., & Matthews, J . (1967). Reactions to suf
Belief in a just world, causal attributions, and ad fering of others under conditions of indirect respon
justment to sexual violence. S o c ia l Ju s t ic e R esea rc h , sibility. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l
18, 2 5 - 4 2 . og y , 5, 3 1 9 - 3 2 5 .
Furnham, A. (1993). Just world beliefs in twelve societ Lerner, M. J ., & Miller, D. T. (1978). Just world re
ies. J o u r n a l o f S o c ia l P sy ch olog y , 133, 3 1 7 - 3 2 9 . search and the attribution process: Looking back
Furnham, A. (2 003). Belief in a just world: Research and ahead. P sy c h o lo g ic a l B u lletin , 8 5, 1 0 3 0 - 1 0 5 1 .
19. B e lie f in a Ju s t W orld 297
Lerner, M. J., 8c Simmons, C. H. (1966). T he observ Raman, L., 8c Winer, G. A. (20 0 4). Evidence of more
er’s reaction to the “ innocent victim”: Compassion immanent justice responding in adults than chil
or rejection? J o u r n a l o f P erson ality a n d S o c ia l P sy dren: A challenge to traditional developmental theo
c h o lo g y , 4, 2 0 3 - 2 1 0 . ries. B ritish J o u r n a l o f D e v e lo p m e n t a l P sy ch olog y ,
Lind, E. A., 8c van den Bos, K. (2002). When fairness 22, 255-274.
works: Toward a general theory of uncertainty man Ritter, C., Benson, D. E., &c Snyder, C. (1990). Belief in
agement. R e se a rc h in O r g a n iz a tio n a l B eh a v io r, 2 4, a just world and depression. S o c io lo g ic a l P ersp ec-
181-223. tw es, 3 3 ,2 3 5 - 2 5 2 .
Lipkus, 1. (1991). The construction and preliminary Ross, M ., 8c Miller, D. T. (Eds.). (2 002). T h e ju stice
validation of a Global Belief in a Just World Scale m o tiv e in e v e ry d a y life. Cambridge, UK: Cambridge
and the exploratory analysis of the Multidimen University Press.
sional Belief in a Just World Scale. P erso n a lity a n d Rubin, Z ., 8c Peplau, L. A. (1973). Belief in a just world
In d iv id u a l D iffe r e n c e s , 12, 117 1-11 78 . and reaction to another’s lot: A study of participants
l.ipikus, I. M., Dalbert, C., 8c Siegler, 1. C. (1996). The in the national draft lottery. J o u r n a l o f S o c ia l Is
importance of distinguishing the belief in a just su es, 29(4), 7 3 - 9 3 .
world for self versus for others: Implications for Rubin, Z ., 8c Peplau, L. A. (1975). W ho believes in a
psychological well-being. P erson ality a n d S o c ia l just world? J o u r n a l o f S o c ia l Issu es, 31(3), 6 5 - 8 9 .
P sy c h o lo g y B u lletin , 2 2 , 6 6 6 - 6 7 7 . Sallay, H. (20 0 4). Entering the job market: Belief in
Lipkus, I. M., 8c Siegler, I. C. (1993). The belief in a a just world, fairness and well-being of graduating
just world and perceptions of discrimination. J o u r students. In C. Dalbert 8c H. Sallay (Eds.), T h e ju s
n a l o f P sy ch olog y , 127, 4 6 5 - 4 7 4 . tice m o tiv e in a d o le s c e n c e a n d y ou n g a d u lt h o o d :
Loo, R. (2 002). Belief in a just world: Support for O rigin s a n d c o n s e q u e n c e s (pp. 2 1 5 - 2 3 0 ) . London:
independent just-world and unjust-world dimen Routledge.
sions. P erso n a lity a n d In d iv id u a l D iffe r e n c e s , 3 3, Schonpflug, U., 8c Bilz, L. (200 4 ). Transmission of the
7 0 3 - 7 1 1. belief in a just world in the family. In C. Dalbert 8c
Lucas, T., Alexander, S., Firestone, I., 8c LeBreton, H. Sallay (Eds.), T h e ju stice m o tiv e in a d o le s c e n c e
J . ( 2 0 07 ). Development and initial validation of a a n d y o u n g a d u lt h o o d : O rigin s a n d c o n s e q u e n c e s
procedural and distributive just world measure. P er (pp. 15 3 -1 7 1 ). London: Routledge.
so n a lity a n d In d iv id u a l D iffer en c es , 4 3 , 7 1 - 8 2 . Steensma, H., 8c van Dijke, R. (20 06 ). Attributional
Maes, J ., tk Kals, F2. (2002). Justice beliefs in school: styles, self-esteem, and just world belief of victims
Distinguishing ultimate and immanent justice. S o of bullying in Dutch organizations. In te rn a tio n a l
c ia l Ju s t ic e R e se a rc h , 15, 2 2 7 - 2 4 4 . Q u a rter ly o f C o m m u n ity H ea lth E d u c a tio n , 2 5 ,
Maes, J . , 8c Schmitt, M . (20 04 ). Belief in a just world 381-392.
and its correlates in different age groups. In C. Dal Strelan, P. (20 07 ). T he prosocial, adaptive qualities of
bert 8c H. Sallay (Eds.), T h e ju stice m o tiv e in a d o just world beliefs: Implications for the relationship
le s c e n c e a n d y o u n g a d u lt h o o d : O rig in s a n d c o n s e between justice and forgiveness. P erson ality a n d In
q u e n c e s (pp. 11-25 ). London: Routledge. d iv id u a l D iffer en c es , 4 3 , 8 8 1 - 8 9 0 .
McClelland, D. C., Koestner, R., 8c Weinberger, J. Sutton, R. M., 8c Douglas, K. M. (20 0 5). Justice for all,
(1989). How do self-attributed and implicit motives or just for me?: More evidence of the importance of
differ? P s y c h o lo g ic a l R ev iew , 9 6 , 6 8 0 - 7 0 2 . the self-other distinction in just-world beliefs. P er
Messick, D. M., Bloom, S., Boldizar, J. P., 8c Samuel- so n a lity a n d In d iv id u a l D iffer en c es , 3 9, 6 3 7 - 6 4 5 .
son, C. I). (1985). Why are we fairer than others? Sutton, R. M ., 8c Winnard, E. J. (20 07 ). Looking
J o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch olog y , 2 1, ahead through lenses of justice: The relevance of
480-500. just-world beliefs to intentions and confidence in
Miller, D. T. (1999). The norm of self-interest. A m e r i the future. B ritish J o u r n a l o f S o c ia l P sy ch o lo g y , 4 6 ,
can P sy c h o lo g ist, 5 4 , 1 0 5 3 - 1 0 6 0 . 649-666.
Montada, I.., Schmitt, M., 8c Dalbert, C. (1986). Taylor, D. M ., Wright, G. C., Moghaddam, F. M .,
Thinking about justice and dealing with one’s 8c Lalonde, R. N. (1990). The personal/group
own privileges: A study of existential guilt. In H. discrimination discrepancy: Perceiving my group,
W. Bierhoff, R. L. Cohen, 8c J. Greenberg (Eds.), but not myself, to be a target for discrimination.
J u s t ic e in s o c ia l rela tio n s (pp. 1 2 5 - 1 4 3 ). New York: P erso n a lity a n d S o c ia l P sy ch o lo g y B u lletin , 16,
Plenum Press. 2 5 4 -2 6 2 .'
Ot to , K., Boos, A., Dalbert, C., Schops, D., 8c Hoyer, Tom ak a, J ., 8c Blascovich, J. (1994). Effects of justice
J. (2006). Posttraumatic symptoms, depression, and beliefs on cognitive, psychological, and behavioral
anxiety of flood victims: T he impact of the belief responses to potential stress .J o u r n a l o f P erson ality
in a just world. P erson ality a n d In d iv id u a l D iffe r a n d S o c ia l P sy ch olog y , 67, 7 3 2 - 7 4 0 .
en c e s, 4 0 , 1 0 7 5 - 1 0 8 4 . Weisz, J . R. (1980). Developmental change in perceived
Ot to , K., & Dalbert, C. (20 0 5). Belief in a just world control: Recognizing noncontingency in the labora
and its functions for young prisoners. J o u r n a l o f tory and perceiving it in the world. D e v e lo p m en ta l
R e se a rc h in P erson ality , 3 9, 5 5 9 - 5 7 3 . P sy ch olog y , 16, 3 8 5 - 3 9 0 .
Ot to , K., 8c Schmidt, S. (20 0 7). Dealing with stress in Zuckerman, M. (1975). Belief in a just world and al
the workplace: Compensatory effects of belief in a truistic behavior. J o u r n a l o f P erso n a lity a n d S o c ia l
just world. E u r o p ea n P sy ch o lo g ist, 12, 2 5 3 - 2 6 0 . P sy ch olog y , 3 1, 9 7 2 - 9 7 6 .
Piaget, J. (1997). T h e m o r a l ju d g m en t o f th e ch ild . Zuckerman, M., 8c Gerbasi, K. C. (19 77 ). Belief in a
Glencoe, IL: Free Press. (Original work published just world and trust. J o u r n a l o f R e se a rch in P erso n
1932) ality, 11, 3 0 6 - 3 1 7 .
CHAPTER 20
• . <s m a ♦
J o h n D u c k it t
298
20. A u th o ritarian ism and D o g m atism 299
and 1960s form a second stage, beginning culminating with a detailed outline and dis
with the publication of Adorno, Frenkel- cussion of new theories and directions that
Brunswik, Levinson, and Sanford’s The have emerged during the past decade.
Authoritarian Personality (1950), which
presented the first systematic theory of the
authoritarian personality and the first at First Stage:
tempt to measure it psychometrically by Em ergence o f the C onstruct
means of their F Scale. This period was
dominated by controversies over the meth The rise of fascism and viru lent anti-Semitism
odological and measurement problems of in Europe during the 1930s stimulated at
their research and problems with alternative tempts to understand the psychological ap
conceptualizations and measures, such as peal of these ideologies. These speculative
Rokeach’s (1954) dogmatism, that emerged theories were strongly influenced by psy
in response to critiques of their work. The choanalysis and M arxism , as well as by the
failure of these alternatives to resolve these culture and personality approach that was
problems led to a general loss of interest and widely held in the social sciences at the time.
confidence in the validity and utility of the Thus they typically argued that the family
construct. structures characteristic of Western, capital
A third stage opened in 1981, with the ist societies would produce a particular kind
publication of Altemeyer’s research on his personality that would make people emo
Right-Wing Authoritarianism (RWA) Scale, tionally and cognitively susceptible to right-
which for the first time seemed to provide a wing, fascist, nationalist ideologies and to
reliable and unidimensional measure of au hostility and aggression against vulnerable,
thoritarianism. This was followed by a great culturally deviant minorities, particularly
deal of research validating the construct and under stressful social circumstances.
exploring its correlates, with most of the Reich (1975), for example, argued that
research reflecting Altemeyer’s conceptual capitalist social structures generated author
ization of authoritarianism as a personality itarian families, which used childrearing
dimension, and validating his RWA Scale as practices involving extensive sexual repres
a measure of this authoritarian personality sion to create authoritarian personalities,
dimension. The period continued with the who would be unlikely to rebel against ex
discovery of the concept of social dominance ploitative social conditions. This authori
orientation (SDO), viewed by Altemeyer tarian character structure was described as
(1998) as a second authoritarian personality conservative, afraid of freedom, submissive
dimension, and research delineating the dif to authority, obedient, yet with “natural
ferences between these two “authoritarian aggression distorted into brutal sadism”
personality” dimensions. (p. 66). Maslow (1943) and Fromm (1941)
Today, this third stage still continues, but, produced very similar descriptions of an au
together with it, a new, fourth stage emerged thoritarian personality.
during the past decade. This new stage is An interesting question is why these
characterized by a very different conceptu theorists and those who followed them ig
alization of authoritarianism that no longer nored situational factors and focused so
views it, or SDO, as personality dimensions completely on personality-based explana
but as social attitude or value dimensions. tions. Historically, the rise of German fas
This perspective has raised a number of re cism was clearly linked to social factors
search and theoretical questions that were such as the social, economic, and political
previously obscured or neglected, most no consequences of Germany’s defeat in World
tably the role of situational factors in influ War I, the Great Depression, and profound
encing authoritarian attitudes and in moder political instability. Some early theorists,
ating and mediating their impact. notably Fromm (1941), did discuss the role
This chapter reviews the way in which the that social factors, and particularly social
constructs of authoritarianism and dogma threat, might play in causing people to sup
tism have been conceptualized and measured port fascist movements and adopt antidem
over these four stages. It shows how these ocratic and prejudiced attitudes. However,
shifts have influenced research and theory, theorists largely ignored the role of social
300 IV. C O G N IT IV E D IS P O S IT IO N S
factors in favor of personality-based expla ian personality dimension (the F Scale), were
nations. There seem to be several reasons based on extensive empirical research. Their
for this. initial research demonstrated that individu
First, research had begun to show stable als’ prejudiced attitudes were generalized
individual differences in peoples’ social at over outgroups, with some people being
titudes, with some persons being generally generally prejudiced and others generally
prejudiced, ethnocentric, nationalistic, con tolerant. Moreover, generalized prejudice
servative, and inclined to agree with profas seemed to be strongly associated with other
cist sentiments, whereas others were gener social, political, and economic attitudes and
ally tolerant, liberal, and opposed to fascism. beliefs, that is, broadly ideological beliefs.
These findings suggested that these attitudes People who were generally prejudiced were
were caused by some stable characteristic, also characterized by nationalism, ethno-
and personality seemed a likely possibility. centrism, social and economic conserva
Second, the liberal and radical political sym tism, antiegalitarianism, and proauthority
pathies of these early theorists led them to see attitudes. Adorno and colleagues followed
racism, prejudice, and fascism as profoundly Maslow, Reich, and Fromm in assuming
irrational and unjustified attitudes and ide that these covarying social and ideological
ologies, so explanations in terms of underly attitudes must be an expression of needs in
ing problems within the personality seemed the personality, that is, an authoritarian per
plausible. And third, M arxist ideas, such as sonality dimension.
that of false consciousness, which were very Their F Scale was developed to measure
influential at the time, rejected the view that this authoritarian personality dimension.
ordinary working people could accept ideol The F Scale was always controversial as a
ogies such as fascism out of any realistic ap measure of authoritarianism and ultimately
praisal of their social circumstances. These did not survive. However, it was very impor
considerations made an explanation in terms tant because it powerfully influenced later
of authoritarian personality dynamics gen measures of authoritarianism, such as the
erated by the socialization practices typical most widely used measure today, the RWA
of authoritarian families and social systems Scale, by providing the broad pool of items
seem particularly plausible. from which later measures were developed.
These early speculative approaches were Thus it delineated and defined the initial
not based on systematic empirical research bounds of the construct.
and did not develop empirically based mea Where, then, did these items come from?
sures of the construct. As a result, they had They did not come from any explicit con
relatively little influence on social scientists struct definition or any clearly defined con
and generated little research. However, their struct domain. They were collected and writ
core theoretical idea that social and ideolog ten more or less ad hoc, with their primary
ical beliefs are a direct expression of basic sources being fascist writings, the speeches
needs in the personality was enormously of anti-Semitic agitators, and persistent
influential and dominated the way in which themes in the interview protocols and T he
authoritarianism has been conceptualized matic Apperception Test (TAT) stories of the
up to the end of the 20th century. participants classified high in ethnocentrism
in Adorno and colleagues’ research (Brown,
1965). All of the items were statements of
Second Stage: opinion that seemed to the researchers to
First Theories and Measures express attitudes and beliefs that were im
plicitly antidemocratic and involved a poten
A dorno and C olleagues’ A uthoritarian
tiality for fascism. They were explicitly se
Personality and the F Scale
lected not to express antiminority attitudes
The publication of Adorno and colleagues’ and political or economic conservatism, be
classic volume in 1950 introduced the con cause these attitudes were assessed in their
struct of an authoritarian personality to research by the Ethnocentrism and Politico-
mainstream social science. Unlike its specu Economic Conservatism scales. The items of
lative predecessors, Adorno et al.’s approach, the F Scale, therefore, covered a broad range
along with their measure of this authoritar of relatively indirect expressions of socially
2 0 . A u th o ritarian ism and D o g m a tism 301
conservative, antiliberal, and antiegalitarian of interest and enthusiasm. In the first de
attitudes and beliefs of a broadly ideologi cade after the publication of their book, the
cal nature. None of its items measured per F Scale was used in hundreds of studies.
sonality in the sense of being self-descriptive However, their theory and research, and the
statements indicating behavioral consisten F Scale itself, soon generated a great deal of
cies or behavioral traits. criticism and controversy (see Christie & Ja-
Adorno and colleagues (1950) organized hoda, 1954). First, research did not support
their F Scale items into the following nine the psychodynamic propositions of their
content categories: theory, and much of their research validating
their F Scale, which compared participants
• Conventionalism (rigid adherence to con high and low in prejudice, seemed seriously
ventional middle-class values). compromised because of their failure to use
• Authoritarian submission (a submissive, blind ratings and to control for sociodemo
uncritical attitude toward authorities). graphic and other group differences. Second,
• Authoritarian aggression (tendency to the construct of authoritarianism seemed to
condemn, reject, and punish people who explain only authoritarianism of the politi
violate conventional values). cal right and ignored authoritarianism of the
• Anti-intraception (opposition to the sub left. And, third, the F Scale became a par
jective, imaginative, and tender-minded). ticular focus of controversy because all of its
• Superstition and stereotypy (belief in mys items were positively worded. As a result, ac
tical determinants of the individual’s fate; quiescence might have inflated the internal
disposition to think in rigid categories). consistency reliability of the F Scale, and its
• Pow er and toughness (preoccupation with apparent unidimensionality might therefore
the dominance-submission, strong-weak, be spurious.
leader-follower dimension; identification A great deal of research, therefore, focused
with power, strength, toughness). on trying to develop a balanced F Scale with
• D estructiveness and cynicism (general equal numbers of protrait and contrait items
ized hostility; vilification of the human). to control acquiescence. None of these at
• Projectivity (disposition to believe that tempts succeeded, however, with balanced
wild and dangerous things go on in the F Scales typically having very low internal
world; the projection outward of uncon consistency. At the time it was thought that
scious emotional impulses). the nature of the original F Scale items made
• Sex (an exaggerated concern with sexual it difficult to psychologically reverse their
“goings-on”). meaning. The alternative possibility that the
F Scale might simply cover a range of item
Although these content categories were content that was not unidimensional was
defined in largely attitudinal terms, they not seriously investigated until much later
came to be seen by the researchers as a set when Altemeyer (1981) showed that this was
of traits that covaried to form the authori the case.
tarian personality syndrome. Their theory The many criticisms of Adorno and col
of this personality suggested that these sur leagues’ (1950) theory and their F Scale re
face traits arose from underlying psychody sulted in a number of attempts to develop al
namic conflicts originating in harsh, puni ternative conceptualizations and measures of
tive parental socialization in childhood. the construct, with the most prominent being
This parental style was presumed to create those of Allport (1954), Rokeach (1954), and
underlying feelings of resentment and anger Wilson (1973). All three shared the basic as
toward parental authority, later generalized sumption that the social attitudes being mea
to all authority, which were repressed and sured either by the F Scale or, in the case of
replaced by deference and idealization of au Rokeach and Wilson, by new measures, were
thority. The underlying repressed anger and expressions of an underlying personality di
aggression were then displaced in the form of mension. All three, however, refined and sim
hostility toward deviant persons, outgroups, plified the conceptualization of this dimension
and minorities. by narrowing the core meaning of the con
Initially, Adorno and colleagues’ (1950) struct and discarding Adorno and colleagues’
theory and the F Scale elicited a great deal complex psychodynamic explanation.
302 IV. C O G N IT IV E D IS P O S IT IO N S
lar to Allport’s (1954) idea that the core of shared two important, and more problem
the authoritarian personality was fearful atic, assumptions.
ness and insecurity. Like Adorno and col One assumption, which went unchal
leagues (1950), Wilson viewed this personal lenged at the time, was that the relatively
ity dimension as being directly expressed in stable individual differences in people’s
conservative social attitudes. The Conserva social and ideological attitudes seemingly
tism, or C, Scale that he developed therefore shown by early empirical research were di
consisted entirely of social attitude items, rect expressions of a particular personality
which covered a wide range of content but dimension. Thus measures consisting en
had as their core themes dislike for and resis tirely of these social attitudes were seen as
tance to change, novelty, and diversity and directly measuring and fully determined by
a preference for order, structure, the tradi that underlying personality dimension.
tional, and the established. The items, there A further, even more problematic assump
fore, covered a similar range of content to tion was that these social and ideological
the F and D Scales, with which it correlated attitudes used to measure the authoritarian
very highly, suggesting that it measured es personality dimension were unidimensional.
sentially the same construct. There seem to be several reasons for this.
As Altemeyer (1981) noted, the C Scale First, if these social attitudes were direct ex
was a major improvement in one important pressions of a particular personality dispo
respect: By using items on which a high con sition, then they should be unidimensional.
servatism score required both agreement Second, early research such as that by Ador
and disagreement, it controlled for acqui no and colleagues (1950) suggested that so
escent responding. Unfortunately, it also cial and ideological attitudes were strongly
shared its predecessors’ weaknesses of hav correlated. However, much of this research
ing extremely low internal consistency. The used scales that were not balanced so that
mean interitem correlation obtained in Alte acquiescence could have spuriously inflat
meyer’s own research with the C Scale was ed these correlations. And third, although
a fairly typical r = .05, indicating that, like theorists had focused on conceptualizing
the F and D Scales, the C Scale did not mea the presumed underlying authoritarian (or
sure a unidimensional construct. Moreover, dogmatic, or conservative) personality di
factor-analytic studies of the C Scale also mension, they generally did not try to clarify
did not reveal a meaningful factor structure the core meaning of the social attitudes they
that enabled the extraction of a unidimen were using to measure this personality and
sional core of items. W ilson’s theory and his so left its boundaries and domain broad,
C Scale had very little impact subsequent to vague, and unclear.
the publication of his edited volume on the
topic in 1973. This was partly due to the
weaknesses of the C Scale and partly due to Third Stage: RW A and SDO
disenchantment with individual-difference
A ltem ey er’s R W A Scale
approaches to explaining ideological beliefs
and H is A uthoritarian Personality
and prejudice that derived from the failure
of a succession of approaches to the issue. Interest in authoritarianism revived in the
1980s with the development and validation
of a new and psychometrically sound mea
Conclusion: A Failure o f M easurem ent
sure by Altemeyer (1981, 1988, 1996)— the
The theories of authoritarian personality Right-Wing Authoritarianism (RWA) Scale.
that emerged in the 1950s and 1960s differed The development of the RWA Scale did not
markedly in their nomenclature, but less so proceed from any theoretical conceptualiza
in how they conceptualized this personal tion or construct definition but rather was
ity construct, with Allport (1954), Rokeach purely inductive. Altemeyer began with a
(1954), and Wilson (1973) refining the com large pool of items comprising those used to
plex psychodynamics initially proposed by develop the F Scale, items from other previ
Adorno and colleagues (1950) in a similar ous authoritarianism scales, and items he had
manner. These theories and measures also written. This item pool was then subjected to
304 IV. C O G N IT IV E D IS P O S IT IO N S
repeated item analyses to produce a core set and discussed later, but for virtually all the
of items that intercorrelated highly enough studies reviewed, the “validity” indices are
to form a unidimensional scale that was bal actually attitudinal self-ratings that overlap
anced against acquiescence. This produced with RWA item content rather than indepen
the initial version of the RWA Scale, which dent behavioral indices of the three “traits.”
showed high internal consistency reliability.
Its unidimensionality was confirmed by fac
A ltem ey er’s “ Validational” R W A Research
tor analyses that typically produced only
on A uthoritarian Subm ission
two factors— one comprising most protrait
items and the other most contratrait items Altemeyer (1996) cited a number of research
(Altemeyer, 1981, 1988, 1996). In addition, findings as supporting the validity of the
research showed high levels of stability over RWA Scale as a measure of a “trait” of au
time, with test-retest correlations being .85 thoritarian submission. Thus RWA scores
over 6 months, .75 over 4 years, .62 over correlated powerfully (i.e., >.50) with rat
12 years, and .59 over 18 years (Altemeyer, ings of how justified or unjustified illegal,
1996). All of the items included in the RWA unfair acts by government officials were
Scale were statements of social attitude and perceived to be (i.e., vignette descriptions
belief of a broadly ideological nature, as had of illegal wiretaps, illegal searches, denial
been the case for other authoritarianism of right to protest, and use of agent provo
measures. Altemeyer (1981) followed ear cateurs), particularly when these targeted
lier authoritarianism theorists in assuming unconventional groups (Altemeyer, 1981).
that these social attitudes and beliefs were RWA scores also correlated moderately to
expressions of an authoritarian personal strongly with American students’ reports of
ity dimension and that his RWA Scale was how long they had continued to believe in
therefore a measure of personality. Richard N ixon’s innocence during the W a
Altemeyer’s (1981) inspection of the tergate crisis (Altemeyer, 1981) and with
content of the RWA Scale suggested that the way Canadian and U.S. students rated
it expressed only three of Adorno and col bogus letters attacking the Canadian Char
leagues’ (1950) nine content categories— ter of Rights and Freedoms and the Ameri
authoritarian submission, authoritarian can Bill of Rights, respectively (Altemeyer,
aggression, and conventionalism. He inter 1988, 1996). Moghaddam and Vuksanovic
preted these as covarying traits and defined (1990) also found that the RWA Scale cor
their covariation as constituting the core related moderately to strongly with lower
of the authoritarian personality dimension support for human rights, and McFarland,
(e.g., Altemeyer, 1981, p. 19). He suggested Ageyev, and Abalakina (1990) found that
that his RWA Scale was unidimensional RWA correlated powerfully with lower sup
precisely because his successive item analy port for democracy in the U .S.S.R. in the
ses had stripped away all items and facets late 1980s.
included in the F Scale and other earlier These findings clearly show an association
measures that had been peripheral to these between RWA and positive attitudes toward
core traits of the authoritarian personality and greater support for established authori
dimension. ties and less positive attitudes and lower sup
The development of the RWA Scale stim port for individual rights and liberties. They
ulated a great deal of research and interest do not, however, as Altemeyer (1996) sug
in the construct. M ost of this research was gested, support a link between RWA and a
guided by Altemeyer’s theoretical assump personality trait, or behaviors, indicative of
tion that authoritarianism was a personal authoritarian submission.
ity dimension comprising three covarying
traits. Indeed, in comprehensively reviewing
A ltem ey er’s “ Validational” R W A Research
RWA research, Altemeyer (1996) treated
on A uthoritarian A ggression
much of it as directly validating RWA as a
measure of the three traits of authoritarian Altemeyer’s (1996) review also suggested
submission, authoritarian aggression, and that research findings supported the validity
conventionalism. This “validational” RWA of the RWA Scale as a measure of a “trait”
research on these three traits is described of authoritarian aggression. RWA scores
20. A u th o ritarian ism and D o g m atism 305
correlated strongly with length of sentences ing attitudes to a number of minorities and
recommended for lawbreakers (Altemeyer, outgroups in Canada and the United States
1981, 1988; Wylie & Forest, 1992). How (Altemeyer, 1996). In a series of “posse”
ever, the correlation was not found in cases studies, Altemeyer (1988, 1996) also found
in which the lawbreakers were government that RWA scores were associated with lower
officials and was reversed if the victims of opposition to the government’s proscribing
the government officials who broke the law and persecuting a variety of radical or devi
were unconventional or deviant, with higher ant groups, even when they were right-wing
RWA then associated with more lenient sen groups. RWA scores were also associated
tences (Altemeyer, 1981, 1988). In addition, with being less condemning of gay bashing
RWA was correlated with selecting more (Altemeyer, 1996).
severe shock levels to punish a learner for These findings replicate earlier research
mistakes on a task. However, although these using measures such as the F Scale (Meloen,
studies seem to deal with a kind of behav 1983) in indicating a relationship between
ioral aggression, the interpretation that they RWA and less favorable attitudes toward
support a direct link between RWA and au outgroups, minorities, and deviant social
thoritarian aggression seems questionable. groups and, in some cases, less opposition
In the sentencing studies, for example, the to measures harming such groups. How
general levels of sentences recommended ever, the inference that these attitudes in
were typically low, so the responses seem dicate a trait of authoritarian aggression
unlikely to reflect aggression as such and seems doubtful. In all the research reviewed,
seem more parsimoniously interpreted as re only one finding involved actual behavior,
flecting greater favorability to the authorities and that study found a very weak relation
and lower favorability to persons challeng ship between RWA and sexual aggression
ing or violating authority. In the learning among men toward women (Walker, Rowe,
task, the levels of shocks selected were also & Quinsey, 1993). The finding was not rep
extremely weak, so the finding may simply licated in other, larger scale research (Alte
reflect an association between RWA and a meyer, 1996).
belief in the efficacy of punishment in learn
ing situations rather than indicating a trait
A ltem ey er’s “ Validational” R W A Research
of authoritarian aggression. These interpre
on Conventionalism
tations seem to be supported by Altemeyer’s
(1988, pp. 1 8 6 -1 8 7 ) finding that general lev Altemeyer’s (1996) review cited a number
els of aggression irrespective of target were of findings to support the empirical validity
uncorrelated with RWA. of the RWA Scale as a measure of a “trait”
A great deal of research has shown moder of conventionalism. First, RWA scores cor
ate to strong correlations between RWA and related powerfully with religiosity, particu
prejudice, assessed as less favorable attitudes larly fundamentalist religiosity, whether
toward a variety of ethnic minority or de Christian, Jewish, or Muslim (Altemeyer,
viant social groups. Targets have included 1996). Indeed, the correlations were so pow
gay people (Altemeyer, 1988; Whitley & erful that Altemeyer (1996) concluded that
Lee, 2 0 0 0 ), Jews and ethnic minorities in fundamentalism was simply the “religious
the U .S.S.R . (McFarland et al., 1990), AIDS manifestation of right-wing authoritarian
victims in the United States (Peterson, Doty, ism” (p. 166). Second, RWA correlated con
&c Winter, 1993), blacks among South Afri sistently and generally powerfully with sup
can whites (Duckitt, 1992), homeless people port for traditional sex roles in a variety of
(Peterson et al., 1993), atheists, drug users, samples and cultures (Altemeyer, 1996; Leak
and welfare recipients (Leak & Randall, & Randall, 1995; McFarland et al., 1990).
1995). Correlations have also been shown Third, RWA correlated with agreement with
with indices of generalized prejudice (i.e., traditional norms of distributive justice in
aggregated attitudes to blacks, women, ho different societies, correlating negatively
mosexuals, and patriotism) in the United with belief in equality in the United States
States and Canada (Altemeyer, 1998; M c and positively in the Soviet Union. Fourth,
Farland, 1998; McFarland & Adelson, the RWA Scale correlated very strongly with
1996) and with ethnocentrism scales target measures of conservatism, traditionalism,
306 IV. C O G N IT IV E D IS P O S IT IO N S
and acceptance of the rules and norms of so and traditional social practices, and the ef
ciety (McFarland et al., 1990; Tarr & Lorr, ficacy and appropriateness of punishment in
1991; Trappnell, 1992, cited in Altemeyer, various situations. This research therefore
1996). And fifth, the RWA Scale showed seems more reasonably interpreted as show
consistent, though only weak to moderate, ing concurrent validity for the RWA scale as
correlations with support for right-wing assessing a broad social attitudinal dimen
political parties in a number of countries, sion without really elucidating the crucial
with the effects stronger for persons who issue of what lies at the core of these atti
were more interested in politics (Altemeyer, tudes and gives them their coherence.
1996).
Altemeyer (1996) argued that these rela
Research on the O rigins o f R W A
tionships supported the validity of the RWA
Scale as measuring a “trait” of convention Research on the origins or determinants of
alism, which he defined as “a high degree of RWA by Altemeyer (1996) and others dur
adherence to conventions that are perceived ing this period focused primarily on social
to be endorsed by society and established learning, parental attitudes, family envi
authorities” (Altemeyer, 1981, p. 148). ronment, and personal experiences. Inter
However, virtually all the research reported estingly, given their assumption that RWA
pertained to conventionalism, conservatism, measured a personality dimension, they
and traditionalism in attitudes, beliefs, or showed little interest in genetic or biologi
values, rather than to the expressions of any cal inheritance as a possible causal factor.
kind of behavioral trait. However, during this period other research
Although Altemeyer (1996) viewed certain ers reported important findings bearing on
other correlates of the RWA Scale as not per the role of both genetic and environmental
taining to the three presumed “traits,” they factors in authoritarianism.
do seem broadly classifiable as indicative of
conventionalism or traditionalism in social
G enetic Influences
attitudes. Thus RWA scores had weak to
moderate negative correlations with proen- During the past few decades two major sets
vironmental attitudes (Peterson et al., 1993; of studies, the Minnesota and Jena twin
Schultz & Stone, 1994). They also correlated studies, have reported on the correlation
with unfavorable attitudes toward drug use between monozygotic and dizygotic twins
(Peterson et al., 1993), though there was no reared apart or together for RWA and close
correlation with actual drug use (Cormer, ly related measures such as traditionalism,
1993, cited in Altemeyer, 1996), and were conservatism, and religiosity (M cCourt,
moderately correlated with antiabortion at Bouchard, Lykken, Tellegen, & Keyes,
titudes (Altemeyer, 1996; Moghaddam & 1999; StoEell, Kampfe, & Riemann, 2006).
Vuksanovic, 1990). These findings have shown consistently
powerful correlations between RWA scores
of monozygotic twins reared apart that have
Conclusions fro m A ltem ey er’s Review
been much higher than the correlations for
o f “ Validational” R W A Research
dizygotic twins reared apart. Thus findings
Virtually all Altemeyer’s research validat from both sets of studies have concurred in
ing RWA as a measure of three authoritar finding strong genetic effects (accounting for
ian personality traits used attitudinal rather 4 0 - 6 0 % of phenotypic variance) on these
than behavioral indictors. M ost of these in social-attitude measures similar in magni
dicators also involved content overlap with tude to those found for standard personality
items included in the RWA scale that ex measures, such as the Big Five. These studies
press attitudes and beliefs about homosexu also found strong environmental effects ac
als, fundamentalist religiosity, governmental counting for approximately 50% of variance
and other authorities, deviant people, people in RWA that were almost entirely due to un
challenging authority, right- and left-wing shared environmental sources (i.e., unique
political issues, traditional gender roles, the individual experiences as opposed to shared
importance of conforming to group norms family environmental influences).
20. A u th o ritarian ism and D o g m a tism 307
Finally, experimental research by Alte to social threat, and parental roles— cause
meyer (1988) and experimental and longi major changes in RWA that can occur at any
tudinal research by others (Doty, Peterson, stage in the lifespan, with adolescence and
& Winter, 1991; Duckitt & Fisher, 20 0 3 ; early adulthood seemingly particularly criti
M cCann, 1999; Sales, 1973; Sales & Friend, cal, seem to militate against the personality
1973) has shown that social threats— such interpretation.
as crime, political and economic crises, and
insecurity— increased RWA and other in
R W A and Cognition
dices of authoritarianism or conservatism.
These effects were sometimes substantial, Adorno and colleagues (1950) and other
with Altemeyer’s crisis scenarios produc early theorists assumed that authoritarian
ing increases in RWA of approximately personalities are characterized by particular
two-thirds of a standard deviation. Con cognitive styles, ways of processing informa
sistent with this, many other studies have tion, and making judgments. Allport (1954),
shown moderate to strong correlations of for example, suggested that the authoritar
RWA with the degree to which people per ian person’s weak ego and general fearful
ceive their social environments as dangerous ness would express itself in dichotomization
and threatening (Altemeyer, 1988; Duckitt, (engaging in simplistic categorical think
2001), and longitudinal research has shown ing) and a need for definiteness (being in
that these perceptions have causal effects on tolerant of ambiguity, preferring structure,
RWA (Sibley, W ilson, & Duckitt, 20 0 7 ). and avoiding and disliking uncertainty).
Overall, although much of the research For Rokeach (1954), pervasive rigidity and
on environmental influences on RWA is in closed-mindedness of thought and belief
conclusive or unclear, the findings of strong were direct expressions of dogmatism.
effects for liberal educational experiences, Altemeyer’s (1996) research has provided
becoming a parent, and exposure to social the most extensive effort to demonstrate an
threat support Altemeyer’s contention that association between RWA and a variety of
personal experiences are important in the deficits and limitations in reasoning, pro
formation and change of RWA. These find cessing information, and making decisions.
ings also seem consistent with his idea that He found that after reading essays on so
RWA is not formed in early childhood, as cialized medicine or corporal punishment,
Adorno and colleagues (1950) had thought, higher RWA was associated with remember
but that adolescence and early adulthood ing less about the material and being less
may be particularly important periods able to recognize false inferences about it.
and that important changes might occur High RWA was associated with greater in
throughout the lifespan. Interestingly, these consistency in responses on social-attitude
conclusions seem more consistent with the questionnaires and a greater tendency to
idea that RWA is an attitude or value dimen agree with contradictory statements on such
sion than a personality dimension. issues. People high in RWA were uncritical
about religious messages and made more
fundamental attribution errors about pro-
G eneral Conclusions
or antihomosexual essays, but particularly
fro m O rigins o f R W A Research
when the essay was antihomosexual (i.e.,
Although much is still unclear about RWA when they themselves agreed with it). Peo
and its origins, certain findings seem well ple high in RWA were also more ready to
established. First, there seems to be a pow believe a political message that they liked,
erful genetic influence. This, together with even though the source seemed likely to be
the high degree of unidimensionality for insincere. Those high in RWA showed great
the relatively diverse content of the RWA er double standards, first, by supporting the
Scale and its high level of stability over ex right of a majority to impose their religion
tended periods, could be consistent with the in public schools when it was their religion
idea that the attitudinal items of the RWA but opposing it when it was a different reli
Scale measure a personality trait dimension. gion and, second, by opposing the right to
However, the findings that certain experi secession for nationalities they disliked but
ences— such as liberal education, exposure supporting it for nationalities they liked.
20. A u th o ritarian ism and D o g m atism 309
People high in RWA were also less ready to may well be wrong”). Preliminary research
want further information about a test they showed that this DO G Scale was highly reli
believed they had done badly on and were able and the mean interitem correlation high
less receptive to negative information about enough to suggest unidimensionality (.28 for
themselves. a student sample, .30 for a parent sample;
M artin (2001), however, argued that virtu Altemeyer, 1996), comprehensively outper
ally all of this research involved issues and val forming the Rokeach (1960) D Scale in these
ues that are ego involving for persons higher respects.
in RWA (e.g., religion, national identity and However, validational research on the
cohesion, intergroup attitudes, homosexual DO G Scale has been limited. Altemeyer
ity) and that ego involvement, rather than any (1996) reported that the DO G Scale sig
general cognitive differences, might account nificantly predicted positive changes in ho
for the effects obtained. This possibility was mophobic attitudes and readiness to shift
supported by an important program of re pro- or antireligious beliefs. The DO G Scale
search that investigated liberal-conservative was also significantly correlated with denial
differences in dispositional versus situational of contradictions and inconsistencies in the
attributions across a range of situations (Skit- Bible (Altemeyer, 2002) and with lower flex
ka, Mullen, Griffin, Hutchinson, & Cham ibility and openness in religious attitudes,
berlin, 2002). Skitka and colleagues (2002) whereas the older Rokeach D Scale was not
found that liberals and conservatives made (Altemeyer, 1996).
the same kinds of attributions in social-issue The D O G Scale correlated strongly with
situations but then adjusted these attributions RWA (.52 for students, .53 for parents), but
to fit with their liberal or conservative values. these correlations were below the reliabili
This finding suggests that the biased and de ties of the two scales and so compatible with
ficient reasoning, judgment, and decisions their being empirically distinct constructs
associated with RWA in Altemeyer’s (1996) (Altemeyer, 1996). This suggests that RWA
research might have motivational rather than is strongly associated with generally dog
cognitive bases. matic inflexibility and rigidity in holding
beliefs. However, there is a problem with
this interpretation, which bears on a major
D ogm atism and A uthoritarianism
unresolved question about the validity of
Interest in dogmatism had languished since the DO G Scale. The DOG Scale correlates
Rokeach’s research during the 1960s for two highly with Religious Fundamentalism (.57
primary reasons. First, although Rokeach’s students, .60 parents) (Altemeyer, 1996),
(I960) conceptualization of dogmatism as and all the studies supporting the validity of
closed-mindedness seemed conceptually the DOG Scale thus far pertain to religious
distinguishable from authoritarianism, his beliefs or beliefs closely related to religion
D Scale correlated so powerfully with the F (e.g., attitudes to homosexuals). This raises
Scale and other measures of authoritarian the possibility that the items of the DOG
ism that it seemed to be measuring essential Scale assess religious dogmatism specifically
ly the same dimension. Second, the D Scale, and not dogmatism in other spheres of be
like the earlier measures of authoritarian lief. Thus the correlation between the DOG
ism, proved to be a psychometrically very Scale and RWA might be due to the religios
poor measure. ity of people high in RWA rather than to any
After his success in developing the RWA tendency for them to be generally rigid or
Scale, Altemeyer (1996) turned his attention inflexible in their beliefs. Indeed, with reli
to dogmatism and its measurement. He began gious fundamentalism controlled, the strong
by clarifying Rokeach’s idea of dogmatism and significant correlations between RWA
by defining it as a “relatively unchangeable, and DO G in Altemeyer’s research seem like
unjustified certainty” (p. 201) in one’s be ly to largely disappear.
liefs and produced a balanced, 20-item scale In conclusion, Altemeyer’s (1996) new
with the items having an impressively high Dogmatism Scale seems a promising devel
level of face validity (“My opinions are right opment but still needs validation. A par
and will stand the test of tim e” and “Flex ticular issue that needs to be clarified is its
ibility is a real virtue in thinking, since you capacity to assess and detect dogmatism in
310 IV. C O G N IT IV E D IS P O S IT IO N S
nonreligious belief domains or only in the in which the strong win and the weak lose,
way religious beliefs are held. whereas RWA is not.
Third, the correlations between the RWA
and SDO scales suggest that they are sub
Social D om inance O rientation:
stantially independent dimensions. Although
A Second A uthoritarian Personality?
some studies, notably in Western European
During the 1990s an important new countries, have reported strong positive cor
individual-difference construct and mea relations (e.g., Duriez & Van Hiel, 2 0 0 2 ;
sure, social dominance orientation (SDO), Van Hiel & Mervielde, 2002), most re
was proposed (Pratto, Sidanius, Stallworth, search, particularly in North America, has
&c M alle, 1994; Sidanius & Pratto, 1999). found weak or nonsignificant correlations
The SDO Scale taps a “general attitudinal (see reviews and meta-analyses by Duckitt,
orientation toward intergroup relations, re 2001, and Roccato & Ricolfi, 2005). Some
flecting whether one generally prefers such studies, notably in East European countries,
relations to be equal, versus hierarchical” have even found negative correlations be
(Pratto et al., 1994, p. 742). Research has tween RWA and SDO (Krauss, 2 0 0 2 ; Van
shown that the SDO Scale predicts a range Hiel, Duriez, & Kossowska, 2006).
of “authoritarian” sociopolitical and inter These findings indicate that although SDO
group phenomena similar to those predicted and RWA both predict “authoritarian” phe
by the RWA Scale, such as generalized preju nomena such as prejudice, intolerance, na
dice, intolerance, right-wing political party tionalism, punitive attitudes, and right-wing
preference, nationalism, patriotism, m ilita politics, they seem to be independent dimen
rism, support for capital punishment, and sions or syndromes. Altemeyer (1998) noted
generally punitive attitudes (Pratto, Sidanius, that the RWA and SDO scales relate to dif
8c Levin, 2 0 0 6 ; Sidanius & Pratto, 1999). ferent sets of the original nine “trait” clus
However, a great deal of evidence indicates ters listed by Adorno and colleagues (1950).
that the SDO and RWA scales are different He therefore concluded that these scales
and relatively independent dimensions (Alte measure two different kinds of authoritari
meyer, 1998; Duckitt, 2001). an personality dimensions: the “submissive”
First, the item content of the two scales is and the “dominant.”
clearly different. RWA items express beliefs The idea that there are two authoritarian
in coercive social control, obedience and re dimensions helps to explain the checkered
spect for existing authorities, and conform history of the authoritarian personality and
ing to traditional moral and religious norms the difficulties of the early theorists. It seems
and values. SDO items, on the other hand, that Adorno and colleagues’ (1950) original
pertain to beliefs in social and economic in conceptualization of the authoritarian per
equality as opposed to equality and the right sonality and their F Scale combined these
of powerful groups to dominate weaker two dimensions and syndromes, resulting in
ones. a lack of unidimensionality. Allport (1954),
Second, the RWA and SDO scales corre Rokeach (1960), and Wilson (1973) had at
late differently with other variables (Alte tempted to simplify the conceptualization of
meyer, 1998; Duckitt, 2 0 0 1 ; Ekehammar & this personality by focusing on the “submis
Akrami, 2 0 0 3 ; McFarland, 2 0 0 6 ; Van Hiel sive” authoritarian, but they failed to nar
& Mervielde, 2002). RWA is powerfully as row their measures correspondingly, and
sociated with religiosity and valuing order, they remained multidimensional. The suc
structure, conformity, and tradition, where cess of Altemeyer’s (1981) RWA Scale thus
as SDO is not. SDO, on the other hand, is seems largely due to its having stripped off
strongly associated with valuing power, those items that tapped the factorially differ
achievement, and hedonism and with being ent “authoritarian dominance” syndrome in
male, whereas RWA is not. RWA is influ his item development studies.
enced by social threat and correlated with Altemeyer (1998, 2004) also proposed
a view of the social world as dangerous and two new hypotheses suggesting different
threatening, whereas SDO is not. SDO is kinds of interaction between the presumed
correlated with a social Darwinist view of personality dimensions of RWA and SDO.
the world as a ruthlessly competitive jungle First, a “double high” hypothesis suggests
20. A u th o ritarian ism and D o g m a tism 311
that persons who are high in both RWA and mension that was measured reliably and
SDO may “combine the worst elements of was relatively stable over long periods of
each kind of personality” (Altemeyer, 2 0 0 4 , time. Not surprisingly, therefore, Altemeyer
p. 421) and so be particularly high in ethno- (1988, 1996) followed earlier authoritarian
centrism, prejudice, and right-wing political ism theorists in viewing this as a personal
orientation. And, second, Altemeyer (1998) ity dimension, and this assumption power
suggested that the combination of high-SDO fully influenced his and others’ research and
leaders and high-RWA followers would form thinking about the construct.
a “lethal union” that would be particularly When the SDO Scale was shown to be as
conducive to groups making and engaging in powerful a predictor of sociopolitical and
seriously unethical decisions and actions. intergroup phenomena as RWA but distinct
In apparent support of the double-high from it, Altemeyer (1998) extended this con
hypothesis, Altemeyer (2004) found that ceptual framework by seeing SDO and RWA
people with double highs had markedly as two independent but complementary au
higher levels of prejudice than persons who thoritarian personalities— the dominant and
were high on just SDO or just RWA. How the submissive authoritarians, respectively.
ever, it was subsequently pointed out that However, during the 1990s this assumption
Altemeyer had not tested for interactions was questioned, and a new view of RWA and
between RWA and SDO and that his find SDO as social-attitude dimensions express
ings might be due to the additive effects of ing motivationally based values emerged,
each on prejudice. This was confirmed by a opening up questions that had been neglect
meta-analysis of findings from 16 separate ed and leading to the development of new
samples, which showed strong additive ef theories.
fects for RWA and SDO on prejudice but no
significant interactive effects (Sibley, Robert
son, &c Wilson, 2006). Fourth Stage: New Perspectives
A study of the “lethal union” that investi on A uthoritarianism
gated the combination of followers with high
Reassessing the View o f A uthoritarianism
RWA and leaders with high SDO found more
as Personality
unethical decisions than for either on their
own, but the research design was not able There are a number of reasons why the view
to test for interactions, so the findings could of authoritarianism as a personality di
simply have reflected the well-documented mension has began to be questioned. First,
additive effects of RWA and SDO (Son the items on authoritarianism measures—
Hing, Bobocel, Zanna, & McBride, 2007). whether the F, D, C, RWA, or SDO scales—
Overall, therefore, although independent, are statements of beliefs and attitudes of a
additive effects of RWA and SDO on preju broadly ideological nature and do not de
dice and a range of social, political, and in scribe behavioral dispositions or traits, as the
tergroup phenomena have been extensively items of personality inventories typically do
documented, no evidence has yet supported (Duckitt, 1989, 20 0 1 ; Feldman & Stenner,
the hypotheses of interaction between them, 1997; Rosier & Willig, 2 0 0 2 ; Saucier, 2 0 0 0 ;
and the idea of a lethal union still needs to Stone, Lederer, & Christie, 1993, p. 232).
be adequately tested. Indeed, Pratto and colleagues (1994; see also
Pratto et al., 2006) have generally described
their SDO Scale as a measure of enduring
Conclusions:
beliefs rather than of personality. The as
A uthoritarianism in Its T h ird Stage
sumption that these social-attitude and be
Altemeyer’s (1981) development of the RWA lief items measure personality rather than
Scale demonstrated that authoritarianism social attitudes or values has never been em
was a viable individual-difference construct pirically tested or verified. For example, as
and revitalized interest and research on au noted already, the research reviewed by Al
thoritarianism. The RWA Scale integrated temeyer (1996) to show that the RWA Scale
a wide range of social, political, and inter measured three covarying personality traits
group attitudes and beliefs along a single actually investigated social attitudes and be
tightly organized individual-difference di liefs broadly covered in the item content of
312 IV. C O G N IT IV E D IS P O S IT IO N S
the RWA Scale. And although the strong ge these two social-attitude dimensions seem to
netic influence demonstrated on RWA would be two distinct sets of motivationally based
be consistent with RWA being a personality higher order sociocultural values, with RWA
dimension, it would be equally compatible correlated with conservation or conserva
with one or more personality dimensions tism values (such as tradition, social confor
influencing RWA but not being isomorphic mity, cohesion, social harmony) and SDO
with it. correlated with valuing power, dominance,
Strong empirical evidence also suggests hierarchy, and inequality in society. Stan-
that authoritarianism measures, and the gor and Leary (2006) have also interpreted
RWA Scale in particular, seem better viewed these dimensions as expressing the two core
as measuring a dimension of social attitudes values of conservatism (versus liberalism,
and values that might be influenced by per freedom, or openness) and egalitarianism
sonality but is not itself a dimension of per (versus power and hierarchy).
sonality. First, the research on twin studies
already reviewed indicated that RWA was
N ew Theories and N ew Research Issues
not influenced by early childhood fam il
ial environments that would be shared by These considerations have increasingly led
twin siblings. Instead, as Altemeyer (1996) researchers to adopt a new view of authori
concluded, it seems to form mainly in late tarianism, with RWA and SDO seen not as
adolescence and early adulthood. These en personality but as social-attitude dimen
vironmental influences seem more compat sions expressing basic social values (Duckitt,
ible with RWA as a social attitude or value 1989, 2 0 0 1 ; Duriez & Van Hiel, 2 0 0 2 ; Feld
dimension than as personality. Second, man, 2 0 0 3 ; Kreindler, 2005). This view has
RWA, SDO, and similar measures have also opened up new issues that had been ob
been shown to be highly reactive to priming, scured and neglected because of the person
situational manipulations, or sociopolitical ality assumption. One that had been com
changes (Duckitt & Fisher, 2 0 0 3 ; Guimond, pletely neglected in the case of RWA (though
Dambrun, Michinov, & Duarte, 2 0 0 3 ; less so for SDO) was that of identifying the
Huang & Liu, 2 0 0 5 ; Sales, 1973; Sales & central, core values that integrate and give
Friend, 1973; Schmitt, Branscombe, &c Kap- the diverse attitudinal content of the dimen
pen, 2003). sion its coherence. Second was the issue of
And third, investigations of the structure understanding the psychological and social
of sociopolitical attitudes and sociocultural bases of these dimensions; what personal
values have typically revealed two roughly ity or social environmental influences shape
orthogonal dimensions, with one corre them and how? And third was that of why
sponding closely to RWA and the other to and how dimensions such as RWA and SDO
SDO (Duckitt, 2 0 0 1 , Table 3). Investigators affect prejudice, political preferences, and
have usually labeled the RWA-like dimen other outcomes.
sion as social conservatism, traditionalism, An important feature of these new theo
or collectivism versus personal freedom, ries that contrasts with the earlier person
openness, or individualism and the SDO- ality approaches has involved giving greater
like dimension as economic conservatism emphasis to social or group factors, both as
and belief in inequality, or power distance underlying the motivational values thought
versus social welfare, egalitarianism, or hu- to be expressed in authoritarian attitudes
manitarianism. Moreover, the social conser and in shaping their effects, such as preju
vatism dimension of social attitudes, when dice. Four of these new theories— that is,
reliably measured, has correlated powerfully the group-cohesion model (Duckitt, 1989),
with the RWA Scale and scaled with it as a Feldman and Stenner’s (1997) interac-
single general factor or dimension (Forsyth, tionist model, Jost and colleagues’ (2003)
1980; Raden, 1999; Saucier, 2 0 0 0 ). For ex motivated-cognition model, and Kreindler’s
ample, Saucier (2000), in a large-scale study (2005) dual-group-process model— tend to
of social attitudes, obtained a correlation of be partial theories focusing only on either
.77 between the RWA Scale and attitudinal RWA or SDO or only on either their causes
measures of social conservatism. Central to or effects. A fifth approach, the dual-process
20. A u th o ritarian ism and D o g m atism 313
motivational model (Duckitt, 2001), inte hierarchy and causes dislike and derogation
grates most of the factors specified by these of lower-status groups. This model has also
four theories to provide a more comprehen not been subject to much empirical testing,
sive approach to explaining how both situ but several studies have shown that experi
ational and dispositional factors underlie the mental manipulations and real-world events
motivational values expressed in both RWA that make peoples’ membership in high-
and SDO and how these generate their ef status groups salient increase SDO (e.g.,
fects on prejudice and politics. Each of these Huang &c Liu, 2 0 0 5 ; Schmitt et al., 2003).
theories is briefly described. There is also research indicating that mak
The group-cohesion model (Duckitt, ing particular group identities salient can
1989) was the first clearly systematized so influence the relationship between RWA and
cial or group approach to authoritarianism outgroup attitudes in a manner broadly con
and focused only on RWA. This model sug sistent with the model (e.g., Verkuyten &
gested that the core idea being expressed Hagendoorn, 1998).
by the items of the RWA Scale was that of Both the preceding models explain the
attitudes to the subordination of individual causes and effects (at least in terms of preju
autonomy to group authority. These authori dice) of authoritarianism largely or entirely
tarian attitudes were seen as direct expres in terms of social situational factors. A
sions of the need for and the value of group third approach, Jost and colleagues’ (2003)
cohesion. This implied that authoritarian motivated-cognition theory of political con
ism was a group phenomenon and could be servatism, focuses only on causes, which it
characteristic of any social group, though sees as both dispositional and situational.
as typically studied using measures such as This theory sees RWA (and closely related
the F Scale or the RWA Scale, the salient so constructs or measures such as the F, C, or
cial group would be the societal or national D scales) expressing attitudinal resistance
group. This need for societal group cohe to change and SDO expressing attitudinal
sion was seen as being a joint product of the support for inequality as two components of
degree to which people identified with their political conservatism. Conservatism, and
societal groups and the degree to which peo therefore both RWA and SDO, expresses
ple perceived threats to the cohesion of that motives to manage and reduce threat and
group. Outgroup dislike or prejudice would uncertainty, which arise from social situ
therefore be caused by perceiving outgroups ational factors likely to activate threat and
as threatening ingroup cohesion or security uncertainty and dispositional factors that
in some way. Evidence supporting this model index the strength of personal needs to avoid
has been reported by Stellmacher and Pet- uncertainty and threat. A meta-analysis
zel (2005), but in general it has not yet been showed that indicators of these factors did
subject to systematic empirical testing. correlate as expected with RWA, SDO, and
Kreindler’s (2005) dual-group-processes other indices of political conservatism (Jost
model (DGPM) also sees authoritarianism et al., 2003). However, it has been pointed
as a group phenomenon and makes similar out that these findings were less convincing
predictions about the causes of RWA. Like for SDO than for RWA. RWA and related
the group-cohesion model, the DGPM sees constructs supplied the bulk of the indices
the primary causal determinants of RWA as used, and their correlations were typically
a joint product of group identification and much stronger than those obtained for SDO.
the perception of threat to the group, though Thus the effects obtained for SDO might
it suggests that the threats that really matter have been spurious and due to the positive
are threats to the norms of the group rather correlation between RWA and SDO (Duckitt
than threats to the group’s cohesion. Out & Sibley, 2009).
group prejudice therefore arises when per A fourth theory, Feldman and Stenner’s
sons high in RWA are hostile to persons seen (1997; Feldman, 2 0 0 3 ; Stenner, 2005) inter-
as threatening group norms. actionist model, focuses on RWA and its ef
The DGPM sees SDO resulting from iden fects. It sees RWA as expressing the value of
tification with high-status groups, which is social conformity. Social-situational factors
thereby expressed in valuing inequality and that threaten social conformity will activate
314 IV. C O G N IT IV E D IS P O S IT IO N S
this value or disposition for persons high in tender-mindedness (i.e., low agreeableness
RWA and generate authoritarian reactions. in terms of the Big Five). Tough-minded
This interactive hypothesis has been empiri personalities view the world as a ruthlessly
cally supported in research on authoritarian competitive jungle in which the strong win
reactions such as outgroup prejudice and and the weak lose. This worldview is also
support for hard-line right-wing social and influenced by exposure to and socialization
political policies (Feldman, 2 0 0 3 ; Feldman in social environments characterized by in
& Stenner, 1997; Rikert, 1998; Stenner, equality, group dominance, and competition
2005). Evidence also suggests that the kind over power, status, and resources. Being
of behavioral rigidity and biased judgments tough-minded and holding this competitive-
or decision making that Altemeyer (1996) jungle worldview makes chronically salient
tried to demonstrate in persons high in the motivational goals and values of power,
RWA seems to occur only under conditions dominance, and superiority over others,
of threat (Lavine, Lodge, & Freitas, 20 0 5 ; which are expressed in the social attitudes
Schultz & Searleman, 2002). This approach of SDO. This account of the origins of RWA
and these findings suggest that social- and SDO has been supported by correlation
environmental threat might not only cause al findings using structural equation model
RWA but also elicit authoritarian reactions ing, by longitudinal research showing the
in interaction with authoritarian attitudes expected causal effects for these personality
and values. and social worldview variables on RWA and
Finally, a dual-process motivational SDO, and by experimental research manipu
(DPM) model provides a broad approach to lating or making salient particular social or
explaining both RWA and SDO and their group environments and showing the ex
effects, incorporating most of the mecha pected effects on worldview beliefs, RWA,
nisms proposed by these new theories. and SDO (see Duckitt & Sibley, 2 0 0 9 , for a
Like the Jost and colleagues (2003) model, review; see also Sibley & Duckitt, 2008).
it sees RWA and SDO caused by both Like Feldman (2003) and Stenner’s (2005)
personality-dispositional factors and by interactionist model for RWA and Kreindler’s
social-environmental factors, but with dif dual-group-process model for both RWA
ferent factors causing RWA and SDO. The and SDO, the DPM model sees the effects of
DPM model proposes that RWA and SDO RWA and SDO as activated and directed by
represent two basic dimensions of social or social-environmental influences. For exam
ideological attitudes, with each expressing ple, persons high in RWA value collective se
motivational goals or values made chroni curity. Therefore, outgroup prejudice will be
cally salient for individuals by their world activated in persons high in RWA by the per
views and personalities (Duckitt, 2001). ception of threats to collective security from
High RWA expresses the motivational goal particular outgroups or minorities. Persons
and value of establishing or maintaining high in SDO value power, dominance, and
collective security, that is, social order, co superiority over others. Consequently, such
hesion, and stability. This motivational persons derogate outgroups low in power
goal or value is made chronically salient for or status (in order to justify their relatively
the individual by the belief that the social superiority) or outgroups competing with
world is inherently dangerous and threaten their own groups over relative power, status,
ing, a belief that is influenced by exposure and resources. Studies have supported these
to and socialization in social environments hypotheses by showing that RWA and SDO
that are threatening and dangerous. The correlate with the different kinds of outgroup
predisposing personality dimension is social prejudice expected from the model (RWA
conformity (which, in terms of the Big Five, with prejudice against dangerous outgroups,
comprises low openness and high conscien SDO with prejudice against low-status out
tiousness), which leads individuals to value groups) and that the effect of RWA on out
order, stability, and security, as well as influ group dislike is mediated by perceived out
encing their beliefs about how dangerous or group threat, whereas the effect of SDO on
threatening their social world may be. outgroup dislike is mediated by competitive
In contrast, SDO stems from the underly ness over relative group status and superior
ing personality dimension of tough- versus ity (Duckitt, 2 0 0 6 ; Duckitt Sc Sibley, 2007).
2 0. A u th o ritarian ism and D o g m a tism 315
Duckitt, J ., & Sibley, C. G. (2 009). In J. Jost, A. Kay, twins reared apart and together. P erson ality a n d In
6C H. Thorisdottir (Eds.), S o c ia l a n d p s y c h o lo g ic a l d iv id u a l D iffe r e n c e s , 27, 9 8 5 - 1 0 1 4 .
b a s e s o f id e o lo g y a n d sy stem ju s tific a tio n (pp. 2 4 2 - McFarland, S. (19 9 8, July). T o w a rd a ty p o lo g y o f
313). Oxford, UK: Oxford University Press. p r e ju d ic e d p er so n s . Paper presented at the annual
Duriez, B., Soenens, B., & Vansteenkiste, M . (2007). meeting of the International Society of Political Psy
In search of the antecedents of adolescent authori chology, Mo ntreal, Quebec, Canada.
tarianism: The relative contribution of parental goal McFarland, S. (20 06 ). P reju d ice d p e o p le : In d iv id u a l
promotion and parenting style dimensions. E u r o p e d iffe r e n c e s in e x p lic it p r e ju d ic e. Unpublished man
an J o u r n a l o f P erson ality , 2 1 , 5 0 7 - 5 2 7 . uscript.
Duriez, B., & Van Hiel, A. (2002). The march of mod McFarland, S., & Adelson, S. (1 9 9 6 , July). A n o m n ib u s
ern fascism: A comparison of social dominance stu dy o f p e r so n a lity , v alu es, a n d p r e ju d ic e. Paper
orientation and authoritarianism. P erson ality a n d presented at the annual meeting of the International
I n d iv id u a l D iffe r e n c e s , 3 2 , 1 1 9 9 - 1 2 1 3 . Society of Political Psychology, Vancouver, British
Ekehammar, B., & Akrami, N. (20 03). The relation Columbia, Canada.
between personality and prejudice: A variable and McFarland, S., Ageyev, V., & Abalakina, M . (1990).
person-centered approach. E u r o p ea n J o u r n a l o f Authoritarianism in the former Soviet Union.
P erson ality , 17, 4 4 9 - 4 6 4 . J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 6 3 ,
Farnen, R., & Meloen, J . ( 2 0 00 ). D em o cr a cy , a u t h o r i 1004-1010.
ta ria n ism , a n d e d u c a tio n : A c r o s s -n a tio n a l em p ir i Meloen, J . (1983). D e a u to r ita ir e r e a k tie in tijd en van
c a l su rvey. New York: St. M artin’s. iv elv aa rt en krisis [ T h e a u th o rita ria n re sp o n s e in
Feldman, S. (2003). Enforcing social conformity: A tim es o f p r o s p e r ity a n d crisis]. Unpublished doc
theory of authoritarianism. P o litic a l P sy ch olog y , toral dissertation, University of Amsterdam.
24, 41-74. Moghaddam, F., &. Vuksanovic, V. (1990). Attitudes
Feldman, S., & Stenner, K. (1997). Perceived threat and behavior towards human rights across different
and authoritarianism. P o litica l P sy ch o lo g y , 18, contexts: The role of right-wing authoritarianism,
741-770. political ideology, and religiosity. I n te r n a tio n a l
Forsyth, D. (1980). A taxonomy of ethical ideologies. J o u r n a l o f P sy ch olog y , 2 5 , 4 5 5 - 4 7 4 .
J o u r n a l o f P erson ality a n d S o c ia l P sy ch o lo g y , 39, Newcomb, T., Koenig, K., Flacks, R ., 8c Warwick, D.
175-184. ( 1967). P ersisten ce a n d ch a n g e: B en n in g ton C o lleg e
Fromm, E. (1941). E s c a p e fr o m fr e e d o m . New York: a n d its stu d en ts a f t e r 2 5 yea rs. New York: Wiley.
Rinehart. Peterson, B., Doty, R., Sc Winter, D. (1993). Authori
Guimond, S., Dambrun, M ., Michinov, N., & Duarte, tarianism and attitudes to contemporary social is
S. (2 003). Does social dominance generate preju sues. P erson ality a n d S o c ia l P sy ch o lo g y B u lletin ,
dice?: Integrating individual and contextual deter 19, 1 7 4 - 1 8 4 .
minants of intergroup cognitions. J o u r n a l o f P er Pratto, F., Sidanius, J ., & Levin, S. (20 06 ). Social dom
so n a lity a n d S o c ia l P sy ch olog y , 8 4 , 6 9 7 - 7 2 1 . inance theory and the dynamics of intergroup rela
Huang, L., & Liu, J . (200 5). Personality and social tions: Taking stock and looking forward. E u r o p ea n
structural implications of the situational priming of R e v iew o f S o c ia l P sy ch olog y , 17, 2 7 1 - 3 2 0 .
social dominance orientation. P erson ality a n d In d i Pratto, F., Sidanius, J ., Stallworth, L., &c Malle, B.
v id u a l D iffe r e n c e s , 3 8 , 2 6 7 - 2 7 6 . (1994). Social dominance orientation: A personal
Jost, J ., Glaser, J ., Kruglanski, A., & Sulloway, F. ity variable predicting social and political attitudes.
(2003). Political conservatism as motivated social J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 67,
cognition. P sy c h o lo g ic a l B u lletin , 129, 3 3 9 - 3 7 5 . 741-763.
Krauss, S. (2 002). Romanian authoritarianism 10 Raden, D. (1999). Is anti-Semitism currently part of an
years after communism. P erson ality a n d S o c ia l P sy authoritarian attitude syndrome? P o litica l P sy ch o l
c h o lo g y B u lletin , 2 8 , 1 2 5 5 - 1 2 6 4 . og y , 2 0 , 3 2 3 - 2 4 4 .
Kreindler, S. (20 0 5). A dual group processes model of Reich, W. (1975). T h e m ass p s y c h o lo g y o f fa scism .
individual differences in prejudice. P erso n a lity a n d Harmondsworth, UK: Penguin.
S o c ia l P sy c h o lo g y R e v iew , 9, 9 0 - 1 0 7 . Rikert, E. (1998). Authoritarianism and economic
Lavine, H., Lodge, M ., & Freitas, K. (20 05 ). Threat, threat: Implications for political behavior. P o litic a l
authoritarianism, and selective exposure to infor P sy ch olog y , 19, 7 0 7 - 7 2 0 .
mation. P o litic a l P sy ch olog y , 2 6 , 2 1 9 - 2 4 4 . Roccato, M ., & Ricolfi, L. (2 005). On the correlation
Leak, G., &c Randall, B. (1995). Clarification of the between right-wing authoritarianism and social
link between right-wing authoritarianism and reli dominance orientation. B a s ic a n d A p p lied S o c ia l
giousness: The role of religious maturity. J o u r n a l o f P sy ch olog y , 27, 1 8 7 - 2 0 0 .
th e S cien tific Study o f R elig io n , 3 4 , 2 4 5 - 2 5 2 . Rokeach, M. (1954). T he nature and meaning of dog
Martin, J . (2001).The authoritarian personality, 50 matism. P s y c h o lo g ic a l R e v iew , 61, 1 9 4 - 2 0 4 .
years later: W hat lessons are there for political psy Rokeach, M. (1960). T h e o p e n a n d th e c lo s e d m in d.
chology? P o litic a l P sy ch olog y , 2 2 , 1 - 2 6 . New York: Basic Books.
Maslow, A. (1943). The authoritarian character struc Rosier, M ., & Willig, C. (2002). The strange death of
ture. J o u r n a l o f S o c ia l P sy ch olog y , 18, 4 0 1 - 4 1 1 . the authoritarian personality: 5 0 years of psycho
M c C a nn , S. (1999). Threatening times and fluctua logical and political debate. H isto r y o f th e H u m an
tions in American church memberships. P erson ality S cien ces, 15, 7 1 - 9 6 .
a n d S o c ia l P sy c h o lo g y B u lletin , 2 5 , 3 2 5 - 3 3 6 . Sales, S. (1973). Threat as a factor in authoritarianism:
M c C ou rt, K., Bouchard, T. J., Lykken, D., Tellegen, An analysis of archival data. J o u r n a l o f P erson ality
A., & Keyes, M . (1999). Authoritarianism revisited: a n d S o c ia l P sy ch olog y , 2 8 , 4 4 - 5 7 .
Genetic and environmental influences examined in Sales, S., & Friend, K. E. (1973). Success and failure
20. A u th o ritarian ism and D o g m atism 317
R ic h a r d E . P e t t y
Pa b l o B r i n o l
C h r is L o e r s c h
M i c h a e l J . M c C a s l in
s conceptualized by Cacioppo and Petty themes in NC work, and we are not able to
A (1982), the need for cognition (NC) re
fers to the tendency for people to vary in the
cover all of the interesting studies that have
been conducted. Nevertheless, we aim to il
extent to which they engage in and enjoy lustrate the major conceptual findings. Most
effortful cognitive activities. Some indi important, the available evidence indicates
viduals have relatively little motivation for that as NC increases, people are more likely
cognitively effortful tasks, whereas other to think about a wide variety of things, in
individuals consistently engage in and enjoy cluding their own thoughts. This enhanced
cognitively challenging activities. O f course, thinking often produces more consequential
people can fall at any point in the distribu (e.g., enduring) judgments and can some
tion. For people high in N C, thinking sat times provide protection from common
isfies a desire and is enjoyable. For people judgmental biases. At other times, however,
low in N C, thinking can be a chore that is enhanced thinking can exacerbate a bias or
engaged in mostly when some incentive or even reverse it. We begin our review with a
reason is present. brief history of the NC concept and its mea
surement. Then we turn to the role of NC in
current dual-process and system theories of
Background and M easurem ent judgment. We conclude with a summary of
some of the key research areas in which the
Since its introduction, NC has been ex NC construct has proven useful.
amined in a large number of studies. In a The NC construct was originally con
comprehensive review over a decade ago ceptualized by Cohen, Stotland, and Wolfe
(Cacioppo, Petty, Feinstein, & Jarvis, 1996), (1955) as reflecting a need to make sense of
well over 100 studies examining NC were the world. Therefore, greater NC was asso
described. Since then, over 100 additional ciated with preference for structure and clar
publications have appeared. To date, over ity in one’s surroundings, making it appear
1,000 publications have either cited the closer to contemporary scales that measure
original article on NC (Cacioppo & Petty, need for structure (see Webster & Kruglan
1982) or the short version of the scale (Ca ski, 1994) than to the current definition. Be
cioppo, Petty, & Kao, 1984). Given the cause Cohen’s original NC measurement de
small amount of space allocated here, we vice was no longer available, Cacioppo and
can only begin to outline some of the major Petty (1982) developed a new scale to reflect
318
21. N eed for C o g n itio n 319
their new conceptualization but retained the theories of judgment were beginning to be
term need for cognition in acknowledge come popular in social psychology. In par
ment of the pioneering efforts of Cohen and ticular, the elaboration likelihood model
colleagues. (Petty & Cacioppo, 1981, 1986), the heuris
Cacioppo and Petty (1982) proposed that tic systematic model (Chaiken, 1987), and
NC is a stable individual difference in the still other dual-process theories (see Chaiken
tendency to engage in and enjoy cognitively & Trope, 1999) proposed that some judg
effortful activities across a wide range of do ments were thoughtfully based on a careful
mains. NC was conceptualized as reflecting consideration of the information presented,
a stable intrinsic motivation that developed whereas other judgments were based on a
over time rather than a need in the tradi more cursory analysis. Within the context
tional sense (i.e., a source of energy that mo of the dual-process theories, NC was used as
tivates behavior). In this conceptualization, a way to determine the mechanism by which
the emphasis is on cognitive processing rath individuals’ judgments would be formed or
er than particular cognitive outcomes. The changed. Considerable research has suggest
idea that NC taps into differences in motiva ed that individuals low in NC are, absent
tion rather than ability is supported by re some incentive to the contrary, more likely to
search showing that NC is only moderately rely on simple cues in a persuasion situation
related to measures of cognitive ability (e.g., (Haugtvedt, Petty, & Cacioppo, 1992) and
verbal intelligence) and continues to predict on stereotypes alone in judging other people
relevant outcomes after cognitive ability is (Carter, Hall, Carney, & Rosip, 2006) than
controlled (see Cacioppo et al., 1996). are those high in NC. Those high in NC are
Although the NC scale was originally de more likely to consider all of the pertinent
veloped as a 34-item inventory (Cacioppo information. Thus, as explained further
& Petty, 1982), the most commonly used later, if cues and stereotypes have any im
version contains 18 statements that people pact on individuals high in N C, it is more
rate on 5-point scales to reflect how char likely to be an indirect effect and to occur
acteristic the statement is of themselves by a mechanism that requires some cognitive
(Cacioppo et al., 1984). Some examples of effort (e.g., Wegener, Clark, & Petty, 2006).
scale items are “I prefer complex to simple Although the 1980s and 1990s were dom
tasks” and “Thinking is not my idea of fun” inated by dual-process models of judgment,
(reverse scored). The scale has high inter the most recent decade has brought forth
nal consistency (reflecting one factor) and various dual-system theories. One system
test-retest reliability. The scale also dem has been referred to as emotional, impul
onstrates good convergent and discriminant sive, intuitive, implicit, or slow learning and
validity. For instance, the scale correlates is contrasted with the other system, which
highly with a recent scale designed to assess is labeled as cognitive, reflective, rational,
elaborated forms of thinking and judgment explicit, or fast learning (Petty & Brinol,
(Eigenberger, Critchely, & Sealander, 2006) 20 0 6). The dual-system theories share with
but is uncorrelated with social desirability the dual-process models the idea that judg
(Fletcher, Danilovics, Fernandez, Peterson, ments are sometimes deliberative and some
& Reeder, 1986; for correlations with many times are not but also propose that high- and
other variables, see Cacioppo et al., 1996; low-thought judgments depend on different
Petty & Jarvis, 1996). Sometimes fewer mental systems that act independently and
than 18 items have been used to assess NC rely on distinct brain structures (e.g., Lie-
with success (e.g., Verplanken, 1991), and a berman, 2000). As was the case with some
two-item version of the scale was developed dual-process models, some dual-system ap
for and used in the 2 0 0 0 National Election proaches have explicitly incorporated the
Study (Bizer et al., 2002). NC construct. In particular, in his cognitive-
experiential self-theory, Epstein (2003) uses
a slightly modified NC scale to tap into the
N C and Theories o f Judgm ent rational system, whereas the Faith in Intu
ition Scale (e.g., “I am a very intuitive per
Cacioppo and Petty (1982) developed the son”) is used to tap the experiential system
NC construct at a time when dual-process (Epstein, Pacini, Denes-Raj, & Heier, 1996).
320 IV. C O G N IT IV E D IS P O S IT IO N S
The rational system is assumed to be logical, the relationship between individuals’ delib
verbal, and relatively affect free, whereas the erative (explicit) versus intuitive (implicit)
experiential system is assumed to be intui self-esteem. The key result was that indi
tive, based on images, and highly dependent viduals who were high in their faith in in
on affect. Because the NC scale is used to tuition showed a larger correlation between
tap the rational system, one might expect their implicit and explicit self-esteem scores
that those high in NC would not rely on in than those low in this trait. However, faith
tuition, images, or affect. However, empiri in intuition moderated the correlation most
cally, the NC and Faith in Intuition scales ly for people high rather than low in N C .1
are uncorrelated, suggesting that individuals This finding is consistent with other work on
both high and low in NC make use of their metacognition showing that confidence in
intuitions, images, and emotions in forming mental content is more important for indi
their judgments. Indeed, the evidence sug viduals high rather than low in N C. That is,
gests that individuals high and low in NC just as individuals high in NC rely on their
use their intuitions, images, and emotions in subjective experiences only to the extent that
different ways. they have confidence in them, so too do they
Specifically, research indicates that affect, rely on any salient mental contents primar
intuitions, and images, like any other men ily when perceived validity is high (see Petty,
tal content, can affect judgments in a variety Brinol, Tormala, & Wegener, 2007).
of relatively thoughtful or nonthoughtful Over the past 25 years, NC has been ex
ways. When a person is not thinking much, amined in a wide variety of areas. For ex
the input (whether emotion, intuition, or ample, in the domain of survey research, it
image) is used in a rather direct way, having has been shown that individuals high in NC
implications for judgment consistent with its provide more thoughtful survey responses
valence (e.g., positive images lead to posi and are less likely to satisfice in their an
tive judgments). However, when thinking is swers (Krosnick, 1991). People high in NC
higher, the impact on judgment is indirect not only engage in more thinking, but they
because the input serves in some other capac are also more aware of their thinking. Thus
ity (e.g., biasing the thoughts that are gener research shows that people high in NC
ated). Thus it may be confusing to think of are more likely to experience lucid dream
NC as assessing “rationality” (Epstein & Pa ing (Blagrove & H artnell, 2 0 0 0 ; Patrick &
cini, 1999) because one might expect purely Durndell, 20 0 4 ), which is the awareness that
rational outcomes from a rational system. one is dreaming. Although there are numer
However, individuals high in NC can be ous studies relating NC to many phenom
highly influenced by their intuitions, emo ena, we have selected four broad domains to
tions, and images, but in thoughtful ways. illustrate the utility of the NC construct: at
This point is not always appreciated, as it is titudes and persuasion, social cognition and
sometimes assumed that only people low in decision making, interpersonal relations,
NC are influenced by these factors. For ex and various more applied domains.
ample, M cM ath and Prentice-Dunn (2005)
suggested that individuals low in NC invari
ably respond more to images than to text. Attitudes and Persuasion
Rather, images can have an impact under
R eliance on Effortful Evaluation
both high and low thinking conditions, but
versus L o w -E ffo rt Processes
by different mechanisms (e.g., see Miniard,
Bhatla, Lord, Dickson, & Unnava, 1991). The psychology of persuasion focuses on
Thus it is preferable to refer to NC as tap which variables produce changes in indi
ping into the tendency to engage in extensive viduals’ beliefs and attitudes and the mecha
thinking. To the extent that this thinking is nisms by which they do so. Consistent with
influenced (biased) by irrational intuitions, the idea that NC is associated with effortful
emotions, or images, the outcome of the thinking, people high in NC tend to form
thinking need not be rational. attitudes on the basis of an effortful analy
In one study investigating the impact of sis of the quality of the relevant information
intuitions on those who vary in N C, Jordan, in a persuasive message (e.g., discriminat
Whitfield, and Zeigler-Hill (2007) examined ing between strong and weak arguments—
21. N eed for C o gn ition 321
Cacioppo, Petty, & Morris, 1983; discrimi behavior) (e.g., Haugtvedt & Petty, 1992;
nating between diagnostic and nondiagnostic Ruiter, Verplanken, De Cremer, & Kok,
information— Chang, 2007). In contrast, 2 0 0 4 ; see Petty, Haugtvedt, & Smith, 1995).
absent any incentive to the contrary, indi If individuals high in NC are told that they
viduals low in NC tend to treat variables as based their attitudes on simple cues rather
simple cues. These include factors such as than on a careful assessment of the message
the attractiveness (e.g., Haugtvedt, Petty, & arguments, they feel ambivalent about their
Cacioppo, 1992) or credibility (Priester & attitudes, which can undermine attitude
Petty, 1995) of the message source (see also strength (Tormala & DeSensi, 2008). Also,
Brinol, Petty, & Tormala, 2 0 0 4 ; Kaufman, because individuals high (vs. low) in NC
Stasson, & H art, 1999), the appearance engage in more thinking, they tend to form
and frame (e.g., positive vs. negative, gains stronger automatic associations among at
vs. losses) of the message (e.g., Chatterjee, titude objects (Brinol, Petty, & M cCaslin,
Heath, Milberg, & France, 2 0 0 0 ; Smith & 2009), and to generalize their changes to
Levin, 1996; Zhang & Buda, 1999), and other, related beliefs (e.g., Murphy, Holle-
their own emotional states (Brinol, Petty, & ran, Long, & Zeruth, 2005).
Barden, 20 0 7 ; Petty, Schumann, Richman,
&C Strathman, 1993).
M etacognition
However, individuals low in NC can be
motivated to scrutinize the available infor Individuals high in NC not only tend to gen
mation carefully and eschew reliance on cues erate more thoughts than those low in N C,
if situational circumstances are motivating— but they are also more likely to think about
such as when the message is of high personal their thoughts (i.e., engage in metacognition;
relevance (Axsom, Yates, & Chaiken, 1987), Petty et al., 2007). For example, following
when there is some uncertainty regarding thought generation, individuals high in NC
the communication (Priester & Petty, 1995; are more likely to evaluate their thoughts
Priester, Dholakia, & Fleming, 2 0 0 4 ; Smith for validity, a process called self-validation
& Petty, 1996; Ziegler, Diehl, & Ruther, (Petty, Brinol, & Tormala, 2002). The more
2002), when the medium through which valid thoughts are seen to be, the more likely
they receive the information is entertaining they are to be used in forming judgments.
or engaging (e.g., when it uses comic strips) Many variables have been shown to affect
(Bakker, 1999; Stephan & Brockner, 20 0 7 ), thought confidence and subsequent thought
when the message matches some aspect of reliance for individuals high but not low in
the recipient’s self-concept (e.g., Brannon NC, including whether people were nod
& M cCabe, 2 0 0 2 ; Evans & Petty, 2003), ding rather than shaking their heads during
and when the message includes emotional thought generation (Brinol & Petty, 2003)
contents (Vidrine, Simmons, & Brandon, or experiencing ease rather than difficulty in
2007; see also Haddock, M aio, Arnold, & thought generation (Tormala, Fakes, Brinol,
Huskinson, 2008). When strong arguments & Petty, 2007; Tormala, Petty, & Brinol,
are presented, increasing thinking enhances 2002). Thought confidence has also been in
persuasion, but when weak arguments are creased for individuals high (vs. low) in NC
presented, increasing thinking diminishes if following thought generation they learned
persuasion. It is important to note that the that the message source was of high versus
normally extensive thinking of individuals low credibility (Brinol, Petty, & Tormala,
high in NC can be undermined when a mes 2 0 0 4 ), were made to feel powerful rather
sage is framed as being for people who do than powerless (Brinol, Petty, Valle, Rucker,
not like to think (Wheeler, Petty, & Bizer, & Becerra, 2 0 0 7 ), or were led to believe that
2005) or when the thinking is demanded their thoughts were shared by similar oth
rather than spontaneous (Lassiter, Apple, & ers (Petty et al., 2002). Enhanced thought
Slaw, 1996; Leone & Ensley, 1986). confidence can increase persuasion when
Because individuals high (vs. low) in NC thoughts are favorable toward the proposal
typically engage in more thinking, they also but decrease it when thoughts are mostly un
tend to have stronger attitudes (e.g., more favorable.
accessible in memory, resistant to change, Not only do individuals high in NC think
and having more impact on subsequent about the thoughts that they have generated
322 IV . C O G N IT IV E D IS P O S IT IO N S
to a message, but they also think about the toward the pen when they were happy. How
process by which they either changed their ever, emotion worked differently for those
attitudes or resisted change. First, people high and low in N C. For individuals high in
high in NC are typically aware of the greater N C, emotion biased the thoughts that were
thought they put into their judgments and as generated (i.e., a happy state led to more fa
a result tend to have more confidence in their vorable thoughts being produced that medi
opinions than individuals low in NC (Barden ated attitude change). For individuals low
& Petty, 2008). Furthermore, when people in N C, a happy state produced more favor
high in NC change their attitudes, they be able attitudes without affecting thoughts
come more confident of their new opinions if (i.e., happiness served as a simple cue). In a
they believe that they have considered both similar vein, Priluck and Till (2004) found
sides of the issue rather than just one side that a deliberative aspect of conditioning—
(Rucker & Petty, 2 0 0 4 ; Rucker, Petty, & contingency awareness— mediated the clas
Brinol, 20 0 8 ). On the other hand, if people sical conditioning effect for individuals high
have resisted persuasion, they can become (but not low) in NC.
more confident in their original attitude if
they are impressed with their resistance
O ther A ttitudinal Effects
(Petty, Tormala, & Rucker, 2 0 0 4 ), such as
when they think they have resisted strong In other research, NC has been related to
arguments rather than weak ones (Tormala a number of well-established attitudinal
& Petty, 2004). phenomena, such as the m ere thought ef
Finally, as a result of their enhanced fect (Smith, Haugtvedt, & Petty, 1994) and
thinking and concern about validity, indi prim acy and recency effects (e.g., Petty,
viduals high (vs. low) in NC are more likely Tormala, Hawkins, & Wegener, 2 0 0 1 ; see
to correct their judgments for any perceived Brinol & Petty, 2 0 0 5 , for a review). Recent
judgmental biases that might be operating research has shown that individuals high
(e.g., DeSteno, Petty, Rucker, Wegener, & (vs. low) in NC are more susceptible to the
Braverman, 2 0 0 4 ; for a review, see Wegen sleeper effect. In this paradigm, individuals
er & Petty, 1997). For example, DeSteno, both high and low in NC initially discount a
Petty, Wegener, and Rucker (2000) found strong persuasive message due to its associa
that when an irrelevant source of emotion tion with a negative cue (e.g., low credibil
was made salient, people high in NC adjust ity source), but persons high in NC become
ed their judgments in a direction opposite to more influenced over time. The reason is
the perceived biasing impact of the emotion thought to be that individuals high but not
(see also Brinol, Rucker, Tormala, & Petty, low in NC had engaged in more processing
2004). of the strong message arguments, so the at
titudes from this emerged once the negative
cue was forgotten (Priester, Wegener, Petty,
M ultiple Roles f o r Variables
& Fabrigar, 1999).
D epend in g on N C
ment. Across a variety of studies, those low simple cue, the thoughts of participants high
in NC tend to show greater amounts of bias in NC were biased in a favorable direction
when this bias is created by a reliance on by the target’s attractiveness (as was the case
mental shortcuts. Alternatively, when the for happiness; see Petty et al., 1993).
bias is created through effortful thought,
individuals high in NC tend to be more
A nchoring
strongly affected. When a bias can come
about through either route, individuals both One well-studied judgmental bias is the
low and high in NC can show the effect, but anchoring effect— the tendency for an ac
it will be produced by different mechanisms. tivated irrelevant number to influence nu
We highlight various research findings that meric estimates (Tversky & Kahneman,
illustrate N C ’s role in producing judgmental 1974). In one study, Epley and Gilovich
bias. (2006) asked students questions that elic
ited self-generated anchors, such as “When
was George Washington elected president?”
False M em ories
(eliciting an anchor of 1776). The responses
One domain in which high thought leads to to these questions provided by individu
more bias is in the creation of false memo als low in NC were more influenced by the
ries. In a common paradigm, participants are starting anchors. Because individuals high
first asked to memorize lists of related words in NC engage in greater levels of thought,
(e.g., table, sit , legs). After this task, recogni they tend to entertain a greater range of pos
tion memory is tested by having participants sible values and subsequently provided esti
go through a larger list that contains both mates further from the initial anchor value.
studied and nonstudied items. The critical Importantly, although this specific process
items in this task are nonstudied words that renders individuals low in NC more suscep
are semantically related to those contained tible to a starting anchor, other anchoring
in the studied list (e.g., chair). Individuals mechanisms tend to emerge more strongly
high in NC are more likely to show false when one thinks extensively about the judg
memory for these lures (Graham, 20 0 7 ). Be ment and when one’s thoughts are biased by
cause individuals high in NC elaborate each the anchor (e.g., see Mussweiler & Strack,
list item and have stronger interconnections 2 0 0 1 , on selective accessibility). When this
in memory, they are more likely to think is the case, those high in NC can show equal
about and access the semantically related or greater judgmental bias from the anchor
(but nonpresented) items and therefore show (Blankenship, Wegener, Petty, Detweiler-
greater false memory for them. Bedell, & Macy, 2008).
in other situations (e.g., reaching consensus Brinol, P., Horcajo, J ., Diaz, D., Valle, C., Becerra, A.,
& De Miguel, J . (200 7 ). El efecto de la formacion
in large-group discussions). Finally, we have
sobre la influencia interpersonal [The effect of
seen how NC relates not only to some classic training on interpersonal influence]. P s ic o th e m a ,
topics in psychology (e.g., the sleeper effect, 19, 4 0 1 - 4 0 5 .
halo effects, priming, group influence) but Brinol, P., 8c Petty, R. E (2 003). Overt head move
also to more recent phenomena (e.g., dual ments and persuasion: A self-validation analysis.
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 8 4,
system models, metacognition). Although 1123-1139.
our review of the literature has been illustra Brinol, P., 8c Petty, R. E. (2005). Individual differences
tive rather than exhaustive, it provides a rea in persuasion. In D. Albarracfn, B. T. Johnson, 8c M.
sonably coherent picture of the proclivities P. Zanna (Eds.), T h e h a n d b o o k o f a ttitu d es a n d a t t i
tu d e ch a n g e (pp. 5 7 5 - 6 1 6 ) . Hillsdale, NJ: Erlbaum.
of those who vary in NC and the utility of
Brinol, P., Petty, R. E., & Barden, J. ( 2 0 07 ). Happi
this construct in a wide variety of basic and ness versus sadness as determinants of thought
applied domains. confidence in persuasion: A self-validation analysis.
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 93,
711-727.
Brinol, P., Petty, R. E., &c McCaslin, M . J . (2009).
Note Changing attitudes on implicit versus explicit mea
sures: W hat is the difference? In R. E. Petty, R. H.
1. T h e m od erat ional im pact o f N C was not shown Fazio, 8c P. Brinol (Eds.), A ttitu d es: In sights fr o m
in a second study that used a subs tantially sm all the n ew im p licit m ea su res (pp. 2 8 5 - 3 2 6 ) . New
er sample and a tru nca ted N C scale. York: Psychology Press.
Brinol, P., Petty, R. E., & Tormala, Z. L. (200 4). The
self-validation of cognitive responses to advertise
References ments. J o u r n a l o f C o n s u m er R e se a rc h , 3 0 , 5 5 9 -
573.
Ahlering, R. F., & Parker, L. D. (1989). Need for cog Brinol, P., Petty, R. E., Valle, C., Rucker, D. D., 8c Bec
nition as a moderator o f the primacy effect. Jo u r n a l erra, A. (20 07 ). The effects of message recipients’
o f R esea rc h in P erso n a lity , 2 3 , 3 1 3 - 3 1 7 . power before and after persuasion: A self-validation
Axsom, D., Yates, S. M ., 8c Chaiken, S. (1987). Au analysis. J o u r n a l o f P erson ality a n d S o c ia l P sy c h o l
dience response as a heuristic cue in persuasion. og y , 93, 1 0 4 0 - 1 0 5 3 .
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 53, Brinol, P., Rucker, D., Tormala, Z. L., 8c Petty, R. E.
30-40. (200 4). Individual differences in resistance to per
Bakker, A. B. (1999). Persuasive communication about suasion: The role of beliefs and meta-beliefs. In E.
AIDS prevention: Need for cognition determinates S. Knowles &c J. A. Linn (Eds.), R e sis ta n c e a n d p e r
the impact of message format. A ID S E d u c a tio n a n d su a s io n (pp. 8 3 - 1 0 4 ) . Mahwah, NJ: Erlbaum.
P rev e n tio n , 11, 1 5 0 - 1 6 2 . Cacioppo, J. T., & Petty, R. E. (1982). The need for
Barden, J., & Petty, R. E. (200 8 ). The mere perception cognition. J o u r n a l o f P erson ality a n d S o c ia l P sy
of elaboration creates attitude certainty: Exploring c h o lo g y , 4 2 , 1 1 6 - 1 3 1 .
the thoughtfulness heuristic. J o u r n a l o f P erson ality Cacioppo, J. T., Petty, R. E., Feinstein, J. A., 8c Jarvis,
a n d S o c ia l P sy ch olog y , 95, 4 8 9 - 5 0 9 . W. B. G. (1996). Dispositional differences in co g
Bizer, G. Y., Krosnick, J. A., Holbrook, A. L., Petty, nitive motivation: The life and times of individuals
R. E., Rucker, D. D . , '& Wheeler, S. C. ( 2 0 0 2 , Sep varying in need for cognition. P s y c h o lo g ic a l B u lle
tember). T h e im p a c t o f p e r s o n a lity on p o lit ic a l b e tin, 119, 1 9 7 - 2 5 3 .
liefs a n d b e h a v io r : N e e d f o r co g n itio n a n d n e e d to Cacioppo, J. T., Petty, R. E., & Kao, C. F. (1984). The
e v a lu a te. Paper presented at the annual meeting of efficient assessment of “need for cognition.” J o u r n a l
the American Political Science Association, Boston. o f P erson ality A s sessm en t, 4 8 , 3 0 6 - 3 0 7 .
Blagrove, M., 8c Hartnell, S. J . ( 2 0 0 0 ). Lucid dream Cacioppo, J. T., Petty, R. E., &c Morris, K. (1983). Ef
ing: Associations with internal locus of control, fects of need for cognition on message evaluation,
need for cognition and creativity. P erson ality a n d argument recall, and persuasion. J o u r n a l o f P ers o n
In d iv id u a l D iffe r e n c e s , 2 8 , 4 1 - 4 7 . a lity a n d S o c ia l P sy ch olog y , 4 5, 8 0 5 - 8 1 8 .
Blankenship, K. L., Wegener, D. T., Petty, R. E., Carter, J. D., Hall, J. A., Carney, D. R., & Rosip, J.
Detweiler-Bedell, B., & Macy, C. L. (200 8). Ela bo C. (20 0 6). Individual differences in the acceptance
ration and consequences of anchored estimates: An of stereotyping. J o u r n a l o f R e se a rc h in P erson ality ,
attitudinal perspective on numerical anchoring. 4 0 , 1 1 0 3 -1 1 1 8 .
J o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch olog y , 4 4 , Chaiken, S. (1987). The heuristic model of persua
1465-1476. sion. In M . P. Zanna, J. M . Olson, & C. P. Herman
Brannon, L. A., & McCabe, A. E. (2 002). Schema- (Eds.), S o c ia l in flu en ce: T h e O n ta rio S y m p osiu m
derived persuasion and perception of AIDS risk. (Vol. 5, pp. 3 - 3 9 ) . Hillsdale, NJ: Erlbaum.
H ea lth M arketin g Q u arterly , 2 0 , 3 1 - 4 8 . Chaiken, S., & Trope, Y. (Eds.). (1999). D u a l-p r o c e s s
Brinol, P., Becerra, A., Diaz, D., Horcajo, J ., Valle, t h e o r ie s in s o c ia l p sy c h o lo g y . New York: Guilford
C., 8c Gallardo, I. (2005). El efecto de la necesidad Press.
de cognicion sobre la influencia interpersonal (The Chang, C. ( 2 0 07 ). Diagnostic advertising content and
impact of need for cognition on interpersonal influ individual differences. J o u r n a l o f A d v ertisin g , 3 6 ,
ence], P sic o th em a , 17, 6 6 6 - 6 7 1 . 75-84.
21. N eed for C o gn ition 327
Chatterjee, S., Heath, T. B., Milberg, S. J ., & France, Haugtvedt, C. P., & Petty, R. E. (1992). Personal
K. R. (20 00 ). The differential processing of price in ity and persuasion: Need for cognition moderates
gains and losses: T he effects of frame and need for the persistence and resistance of attitude changes.
co gnition . J o u r n a l o f B eh a v io r a l D ecisio n M akin g , lo u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 6 3,
13, 6 1 - 7 5 . 308-319.
Cohen, A. R., Stotland, E., & Wolfe, D. M . (1955). Haugtvedt, C. P., Petty, R. E., & Cacioppo, J. T. (1992).
An experimental investigation of need for cogni Need for cognition and advertising: Understanding
tion. Jo u r n a l o f A b n o r m a l a n d S o c ia l P sy ch olog y , the role of personality variables in consumer behav
S I, 2 9 1 - 2 9 4 . ior. J o u r n a l o f C o n s u m er P sy ch olog y , 1, 2 3 9 - 2 6 0 .
Crawford, M. T., & Skowronski, J . (1998). When m o Henningsen, D. D., & Henningsen, M. L. M. (2004).
tivated thought leads to heightened bias: High need The effect of individual difference variables on
for cognition can enhance the impact of stereotypes information sharing in decision-making groups.
on memory. P erson ality a n d S o c ia l P sy ch o lo g y B u l H u m an C o m m u n ic a tio n R esea rch , 3 0 , 5 4 0 - 5 5 5 .
letin , 2 4 , 1 0 7 5 - 1 0 8 8 . Higgins, E. T„ Rholes, W. S., & Jones, C. R. (1977).
DeSteno, D., Petty, R. E., Rucker, D. D., Wegener, D. Category accessibility and impression formation.
T., & Braverman, J . (20 0 4). Discrete emotions and J o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch olog y , 13(2),
persuasion: The role of emotion-induced expe ctan 1 41-154.
cies. J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , Hittner, J. B. (2 004). Alcohol use among American
86, 43-56. college students in relation to need for cognition
DeSteno, D., Petty, R. E., Wegener, D. T., & Rucker, and expectations of alcohol’s effects on cognition.
D. D. (20 0 0). Beyond valence in the perception of C u rren t P sy ch o lo g y : D ev e lo p m e n ta l, L ea rn in g ,
likelihood: The role of emotion specificity. J o u r n a l P erson ality , S o cia l, 2 3 , 173 -1 87 .
o f P erson ality a n d S o c ia l P sy ch olog y , 78(3), 3 9 7 - Jordan, C. H „ Whitfield, M „ & Zeigler-Hill, V.
416. (200 7). Intuition and the correspondence between
Eigenberger, M. E., Critchely, C., & Sealander, K. A. implicit and explicit self-esteem. J o u r n a l o f P erso n
( 2 0 06 ). Individual differences in epistemic style: A ality a n d S o c ia l P sy ch olog y , 93, 1 0 6 7 - 1 0 7 9 .
dual-process perspective. J o u r n a l o f R e se a rch in Kaufman, D. Q., Stasson, M . F., & Hart, J. W. (1999).
P erson ality , 41, 3 - 2 4 . Are the tabloids always wrong or is that just what
Epley, N., & Gilovich, T. (200 6 ). T he anchoring-and- we think?: Need for cognition and perceptions of
adjustment heuristic: Why the adjustments are in articles in print media. J o u r n a l o f A p p lie d S o cia l
sufficient. P s y c h o lo g ic a l S cien ce, 17(4), 3 1 1 - 3 1 8 . P sy ch olog y , 2 9 , 1 9 8 4 - 1 9 9 7 .
Epstein, S. (2 003). Cognitive-experiential self-theory Krosnick, J. A. (1991). Response strategies for coping
of personality. In T. Millon & M . J. Lerner (Eds.), with the cognitive demands of attitude measures
H a n d b o o k o f p s y c h o lo g y : Vol. S. P erson ality a n d in surveys. A p p lied C o g n itiv e P sy ch olog y , 5, 2 1 3 -
s o c ia l p s y c h o lo g y (pp. 1 5 9 - 1 8 4 ). Hoboken, NJ: 236.
Wiley. Lassiter, G. D., Apple, K. J., & Slaw, R. D. (1996).
Epstein, S., & Pacini, R. (1999). Some basic issues Need for cognition and thought-induced attitude
regarding the dual-process theories from the per polarization: Another look. J o u r n a l o f S o c ia l B e
spective of cognitive-experiential self-theory. In S. h a v io r a n d P erson ality , 11, 6 4 7 - 6 6 5 .
Chaiken & Y. Trope (Eds.), D u a l-p r o c e s s th e o r ie s Leippe, M . R., Eisenstadt, D., Rauch, S. M ., & Seib,
in s o c ia l p s y c h o lo g y (pp. 4 6 2 - 4 8 2 ) . New York: H. M . (20 04 ). Timing of eyewitness expert testi
Guilford Press. mony, jurors’ need for cognition, and case strength
Epstein, S., Pacini, R., Denes-Raj, V., & Heier, as determinants of trial verdicts .J o u r n a l o f A p p lie d
H. (1996). Individual differences in intuitive- P sy ch olog y , 8 9, 5 2 4 - 5 4 1 .
experiential and analytical-rational thinking styles. Leone, C., & Ensley, E. (1986). Self-generated attitude
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 71, change: A person by situation analysis of attitude
390-405. polarization and attenuation. J o u r n a l o f R esea rch
Evans, L., & Petty, R. E. (2003). Self-guide framing in P erson ality , 2 0 , 4 3 4 - 4 4 6 .
and persuasion: Responsibly increasing message Levin, 1. P., Huneke, M. E., & Jasper, J. D. (20 00 ). In
processing to ideal levels. P erson ality a n d S o c ia l formation processing at successive stages of decision
P sy c h o lo g y B u lletin , 2 9 , 3 1 3 - 3 2 4 . making: Need for cognition and inclusion-exclusion
Feingold, A. (1992). Good-looking people are not what effects. O rg a n iz a tio n a l B e h a v io r a n d H u m an D e c i
we think. P sy c h o lo g ic a l B u lletin , 111(2), 3 0 4 - 3 4 1 . sio n P ro cesses, 8 2 , 1 71-193.
Fletcher, G. J. O., Danilovics, P., Fernandez, G., Pe I.ieberman, M . D. (20 00 ). Intuition: A social cogni
terson, D., &c Reeder, G. D. (1986). Attributional tive neuroscience approach. P s y c h o lo g ic a l B u lletin ,
complexity: An individual difference measure. 126, 1 0 9 - 1 3 7 .
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 51, M artin , L. L., Seta, J. J., & Crelia, R. A. (1990). As
875-884. similation and contrast as a function of people’s
Graham , L. M. ( 2 00 7). Need for cognition and false willingness and ability to expend effort in forming
memory in the Deese-R oed ig er-M cD e rm o tt para an impression. J o u r n a l o f P erson ality a n d S o c ia l
digm. P erson ality a n d In d iv id u a l D iffer en ces , P sy ch olog y , 59(1), 27 -3 7 .
42(3), 4 0 9 - 4 1 8 . M c M a th , B. F., & Prentice-Dunn, S. (2 005). Protec
Haddock, G., Maio, G., Arnold, K., & Huskinson, T. tion motivation theory and skin cancer risk: The
(2008). Should persuasion be affective or cognitive: role of individual differences in responses to persua
The moderating effects of need for affect and need sive appeals. J o u r n a l o f A p p lie d S o c ia l P sy ch olog y ,
for cognition. P erso n a lity a n d S o c ia l P sy ch o lo g y 35, 6 2 1 - 6 4 3 .
B u lletin , 3 4 , 7 6 9 - 7 7 8 . Miniard, P., Bhatla, S., Lord, K. R., Dickson, P. R.,
328 IV . C O G N IT IV E D IS P O S IT IO N S
& Unnava, H. R. (1991). Picture-based persuasion D. A. Prentice (Eds.), P ersp ectiv ism in s o c ia l p s y
processes and the moderating role of involvement. c h o lo g y : T h e yin a n d y an g o f scie n tific p ro g re ss
J o u r n a l o f C o n s u m er R e se a rch , 18, 9 2 - 1 0 7 . (pp. 3 7 - 5 1 ). Washington, DC : American Psycho
Murphy, P. K., Holleran, T. A., Long, J. F., & Zeruth, logical Association.
J. A. (200 5 ). Examining the complex roles of moti Priester, J . , Wegener, D., Petty, R. E., & Fabrigar, L.
vation and text medium in the persuasion process. (1999). Examining the psychological processes un
C o n te m p o r a r y E d u c a tio n a l P sy ch olog y , 3 0 , 4 1 8 — derlying the sleeper effect: The elaboration likeli
43 8 . hood model explanation. M ed ia P sy ch olog y , 1,
Mussweiler, T., & Strack, F. (2001). The semantics of 27-48.
anchoring. O r g a n iz a tio n a l B e h a v io r a n d H u m an Priester, J . R., Dholakia, U. M ., & Fleming, M. A.
D ec is io n P ro c esses, 8 6(2), 2 3 4 - 2 5 5 . (2004). When and why the background contrast
Patrick, A., & Durndell, A. (20 04 ). Lucid dreaming effect emerges: Thought engenders meaning by in
and personality: A replication. D rea m in g , 14, 2 3 4 - fluencing the perception of applicability. J o u r n a l o f
239. C o n s u m er R e se a rc h , 3 1, 4 9 1 - 5 0 1 .
Perlini, A. H., & Hansen, S. D. (2001). Moderating Priester, J. R., & Petty, R. E. (1995). Source attribu
effects of need for cognition on attractiveness ste tions and persuasion: Perceived honesty as a deter
reotyping. S o c ia l B e h a v io r a n d P erson ality , 29, minant of message scrutiny. P erson ality a n d S o c ia l
313-321. P sy ch o lo g y B u lletin , 2 1 , 6 3 7 - 6 5 4 .
Petty, R. E., & Brinol, P. ( 2 00 6). Understanding social Priluck, R., & Till, B. D. (20 04 ). The role of contin
judgment: Multiple systems and processes. P sy ch o gency awareness, involvement, and need for co gni
lo g ic a l In q u iry , 17, 2 1 7 - 2 2 3 . tion in attitude formation. J o u r n a l o f th e A c a d e m y
Petty, R. E ., Brinol, P., & Tormala, Z. L. (2002). o f M a rk etin g S cien c e, 3 2 , 3 2 9 - 3 4 4 .
Thought confidence as a determinant of persuasion: Rucker, D. D., &c Petty, R. E. (20 0 4). When resistance
The self-validation hypothesis. J o u r n a l o f P ers o n a l is futile: Consequences of failed counterarguing for
ity a n d S o c ia l P sy ch olog y , 8 2 , 7 2 2 - 7 4 1 . attitude certainty. J o u r n a l o f P erson ality a n d S o c ia l
Petty, R. E., Brinol, P., Tormala, Z. L., &c Wegener, D. P sy ch olog y , 8 6 , 2 1 9 - 2 3 5 .
T. (20 07 ). T he role of metacognition in social judg Rucker, D. D., Petty, R. E., & Brinol, P. (2008). W h a t ’s
ment. In A. W. Kruglanski & E. T. Higgins (Eds.), in a frame anyway?: A meta-cognitive analysis of
S o c ia l p sy c h o lo g y : H a n d b o o k o f b a s ic p rin cip les one- versus two-sided message framing. J o u r n a l o f
(2nd ed., pp. 2 5 4 - 2 8 4 ) . New York: Guilford Press. C o n s u m er P sy ch olog y , 18, 137 -1 49 .
Petty, R. E., & Cacioppo, J. T. (1981). A ttitu d es a n d Ruiter, R. A. C., Verplanken, B., De Cremer, D., &
p e r s u a s io n : C la ssic a n d c o n te m p o r a r y a p p r o a c h e s . Kok, G. (20 04 ). Danger and fear control in response
Dubuque, IA: Brown. to fear appeals: T he role of need for cognition. B asic
Petty, R. E., & Cacioppo, J. T. (1986). C o m m u n ic a a n d A p p lied S o c ia l P sy ch olog y , 2 6 , 1 3 - 2 4 .
tio n a n d p e r s u a s io n : C en tra l a n d p e r ip h e r a l ro u te s Sargent, M . (20 04 ). Less thought, more punishment:
t o a ttitu d e ch an g e. New York: Springer-Verlag. Need for cognition predicts support for punitive re
Petty, R. E., DeMarree, K. G., Brinol, P., Horcajo, J., sponses to crime. P erso n a lity a n d S o c ia l P sy ch o lo g y
& Strathman, A. J . (2008). Need for cognition can B u lletin , 3 0 , 1 4 8 5 - 1 4 9 3 .
magnify or attenuate priming effects in social judg Schei, V., Rognes, J . K., & Mykland, S. (200 6 ). T h in k
ment. P erso n a lity a n d S o c ia l P sy c h o lo g y B u lletin , ing deeply may sometimes help: Cognitive motiva
34, 900-912. tion and role effects in negotiation. A p p lie d P sy
Petty, R. E., Haugtvedt, C., & Smith, S. M . (1995). c h o lo g y : An I n te r n a tio n a l R e v iew , 55, 7 3 - 9 0 .
Elaboration as a determinant of attitude strength: Shestowsky, D., 8c Horowitz, L. M . (20 04 ). How the
Creating attitudes that are persistent, resistant, Need for Cognition Scale predicts behavior in mock
and predictive of behavior. In R. E. Petty & J. A. jury deliberations. L a w a n d H u m an B eh a v io r, 2 8 ,
Krosnick (Eds.), A ttitu d e stren g th : A n teced en ts 305-337.
a n d c o n s e q u e n c e s (pp. 9 3 - 1 3 0 ) . Mahwah, NJ: E r Shestowsky, D., Wegener, D. T., 8c Fabrigar, L. R.
lbaum. (1998). Need for cognition and interpersonal influ
Petty, R. E., &c Jarvis, B. G. (19 96). An individual dif ence: Individual differences in impact on dyadic de
ferences perspective on assessing cognitive process cisions. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l
es. In N. Schwarz & S. Sudman (Eds.), A n sw erin g og y , 74, 13 1 7 -1 3 2 8 .
q u e s tio n s : M e th o d o lo g y f o r d ete rm in in g co g n itiv e Smith, B. N., Kerr, N. A., Markus, M . J ., 8c Stasson,
a n d c o m m u n ic a tiv e p r o c e s s e s in su rv ey research M. F. (2001). Individual differences in social loaf
(pp. 2 2 1 - 2 5 7 ) . San Francisco: Jossey-Bass. ing: Need for cognition as a motivator in collective
Petty, R. E., Schumann, D. W., Richman, S. A., & performance. G r o u p D y n a m ics: T h eo r y , R esearch ,
Strathman, A. J . (1993). Positive mood and persua a n d P ractice, S, 1 5 0 - 1 5 8 .
sion: Different roles for affect under high- and low- Smith, S. M ., Haugtvedt, C. P., 8c Petty, R. E. (1994).
elaboration conditions. J o u r n a l o f P erso n a lity a n d Need for cognition and the effects of repeated ex
S o c ia l P sy ch olog y , 64(1), 5 - 2 0 . pression on attitude accessibility and extremity. A d
Petty, R. E., Tormala, Z ., Hawkins, C., 6C Wegener, v a n c es in C o n s u m er R e se a rc h , 2 1 , 2 3 4 - 2 3 7 .
D. T. (2001). Motivation to think and order effects Smith, S. M ., 8c Levin, I. P. (1996). Need for cognition
in persuasion: The moderating role of chunking. and choice framing effects. J o u r n a l o f B e h a v io r a l
P erso n a lity a n d S o c ia l P sy ch o lo g y B u lletin , 27, D ecisio n M a k in g , 9, 2 8 3 - 2 9 0 .
332-344. Smith, S. M ., 8c Petty, R. F^. (1996). Message fram
Petty, R. E ., Tormala, Z . L., & Rucker, D. D. (2004). ing and persuasion: A message processing analysis.
Resisting persuasion by counterarguing: An attitude P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , 2 2 ,
strength perspective. In J . T. Jost, M . R. Banaji, & 257-268.
21. N eed fo r C o gn ition 329
Sommers, S. R., Sc Kassin, S. M. (2001). On the many Wegener, D. T., Clark, Jj. K., Sc Petty, R. E. (2006).
impacts of inadmissible testimony: Selective co m Not all stereotyping is created equal: Differential
pliance, need for cognition, and the overcorrection consequences of thoughtful versus non-thoughtful
bias. P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , stereotyping. J o u r n a l o f P erson ality a n d S o c ia l P sy
27, 1 3 6 8 - 1 3 7 7 . c h o lo g y , 9 0 , 4 2 - 5 9 .
Stephan, J ., Sc Brockner, J. (2 007). Spaced out in cy Wegener, D. T., Kerr, N. L., Fleming, M. A., Sc Petty,
berspace?: Evaluations of computer-based infor R. E. ( 2 0 0 0). Flexible corrections of juror judg
mation. J o u r n a l o f A p p lied S o c ia l P sy ch olog y , 37, ments: Implications for jury instructions. P sy c h o l
210-226. ogy, P u b lic P olicy, a n d L a w , 6, 6 2 9 - 6 5 4 .
Tormala, Z. L., & DeSensi, V. L. (20 0 8). The per Wegener, D. T., Sc Petty, E. (1995). Flexible co rrec
ceived informational basis of attitudes: Implications tion processes in social judgment: The role of naive
for subjective ambivalence. P erso n a lity a n d S o c ia l theories in corrections for perceived bias. J o u r n a l o f
P sy c h o lo g y B u lletin , 3 4 , 2 7 5 - 2 8 7 . P erson ality a n d S o c ia l P sy ch olog y , 68(1), 3 6 - 5 1 .
Tormala, Z. L., Falces, C., Brinol, P., 8c Petty, R. E. Wegener, D. T., Sc Petty, R. E. (1997). T he flexible co r
(2 00 7 ). Ease of retrieval effects in social judgment: rection model: The role of naive theories in bias co r
The role of unrequested cognitions. J o u r n a l o f P er rection. In M. P. Zanna (Ed.), A d v a n c es in e x p e r i
so n a lity a n d S o c ia l P sy ch olog y , 93, 14 3-15 7. m en ta l s o c ia l p s y c h o lo g y (Vol. 29, pp. 141-2 08 ).
Tormala, Z . L., Sc Petty, R. E. (200 4). Resistance to San Diego, CA: Academic Press.
persuasion and attitude certainty: The moderating Wheeler, S. C., Petty, R. E., & Bizer, G. Y. (2 005). Self
role of elaboration. P erson ality a n d S o c ia l P sy ch o l schema matching and attitude change: Situational
og y B u lletin , 3 0 , 1 4 4 6 - 1 4 5 7 . and dispositional determinants of message elabora
Tormala, Z. L., Petty, R. E., & Brinol, P. (2002). Ease tion. J o u r n a l o f C o n s u m e r R esea rch , 3 1 , 7 8 7 - 7 9 7 .
of retrieval effects in persuasion: A self-validation Williams-Piehota, P., Schneider, T. R., Pizarro, J.,
analysis. P erson ality a n d S o c ia l P sy ch o lo g y B u lle Mowad, L., 8c Salovey, P. (2003). Matching health
t i n ' s , 1700-1712. messages to information-processing styles: Need for
Tversky, A., & Kahneman, D. (1974). Judgment cognition and mammography utilization. H ea lth
under uncertainty: Heuristics and biases. S cien ce, C o m m u n ic a tio n , IS , 3 7 5 - 3 9 2 .
1 55( 4157), 1 1 2 4 - 1 1 3 1 . Yang, Y., Sc Lee, H. J. (1998). The effect of response
Verplanken, B. (1991). Persuasive communication of mode, prior knowledge, and need for cognition on
risk information: A test of cue versus message pro consumers’ information acquisition process. K o rea n
cessing effects in a field experiment. P erso n a lity a n d J o u r n a l o f In d u s tr ia l a n d O rg a n iz a tio n a l P sy c h o l
S o c ia l P sy c h o lo g y B u lletin , 17, 1 8 8 - 1 9 3 . ogy, 11, 8 5 - 1 0 3 .
Vidrine, J . E, Simmons, V. N., 8c Brandon, T. H. Zhang, Y., Sc Buda, R. (1999). Moderating effects of
(2 0 07 ). Construction of smoking-relevant risk per need for cognition on responses to positively versus
ceptions among college students: The influence of negatively framed advertising messages. J o u r n a l o f
need for cognition and message content. J o u r n a l o f A d v ertisin g , 2 8 , 1-15.
A p p lied S o c ia l P sy ch olog y , 37, 9 1 - 1 1 4 . Ziegler, R., Diehl, M., & Ruther, A. (2 002). Multiple
Webster, D. M ., Sc Kruglanski, A. W. (1994). Indi source characteristics and persuasion: Source in
vidual differences in need for cognitive closure. consistency as a determinant of message scrutiny.
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o lo g y , 67, P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , 2 8 ,
1049-1062. 496-508.
CHAPTER 22
Optimism
C h a r l e s S. C a r v e r
M i c h a e l F. S c h e i e r
330
22. O p tim ism 331
preferred measure is the Life Orientation One final issue that also bears mention is
Test— Revised, or LOT-R (Scheier, Carver, that there has been some controversy about
& Bridges, 1994). It consists of a series of whether the optimism construct should be
statements (e.g., “I ’m always optimistic viewed as one bipolar dimension or wheth
about my future,” “I rarely count on good er there are two separable dimensions, one
things happening to me” |reversed]) to which pertaining to the affirmation of optimism
people indicate the extent of their agreement and the other pertaining to affirmation of
or disagreement on a multipoint scale. Other pessimism. There have been cases in which
measures have also been created that simi separating those qualities has led to better
larly consist of statements about good and prediction of outcomes (Marshall, Wort-
bad outcomes, with respondents indicating man, Kusulas, Hervig, & Vickers, 1992;
their agreement or disagreement with the Robinson-Whelen, Kim, M acCallum , &
statement (e.g., Dember, M artin, Hummer, Kiecolt-GIaser, 1997), but this does not al
Howe, & Melton, 1989). Such generalized ways happen. A number of studies aimed at
expectancies, which pertain to the person’s settling the issue have arrived at different
entire life space, are what we mean by opti answers, with some concluding that a uni
mism and pessimism. dimensional view is accurate (Rauch, Sch-
Another approach to measuring optimism weizer, & Moosbrugger, 2007) and others
relies on the idea that people’s expectan concluding that there are two dimensions
cies for the future stem from their interpre (Herzberg, Glaesmer, & Hoyer, 2006). For
tations of the past (Peterson & Seligman, the sake of simplicity in presentation, in this
1984). If past failures are seen as reflecting chapter we treat optimism-pessimism as one
stable causes, expectancies will be pessimis dimension. Keep in mind, however, that in
tic, because the cause (which is relatively some situations what matters may be the
permanent) is likely to remain in force. If extent to which people endorse versus reject
past failures are seen as reflecting unstable a pessimistic outlook rather than endorse
causes, the outlook for the future may be versus reject an optimistic outlook, and vice
brighter, because the cause may no longer be versa.
there. Some define optimism and pessimism In this chapter we describe some ways in
in terms of patterns of attributions made which individual differences in optimism
about the causes of events (e.g., Peterson & versus pessimism, measured as expectations
Seligman, 1984; see also Furnham, Chapter for one’s future, relate to variations in other
18, this volume) and make the inference that important aspects of life (see also Seger-
the attributions result in expectancies. This strom, 2006a). M anifestations of optimism
view differs from ours in important ways, are grouped here into four sets: subjective
but both share the theme that expectations well-being, coping responses, physical well
for the future affect people’s actions and ex being, and social relationships.
periences.
Each of these measures gives a continuous
distribution of scores. Writers often refer O ptim ism and Subjective W ell-Being
to optimists and pessimists as though they
were distinct categories of people, but this is One straightforward influence of optimism
a verbal convenience. People actually range and pessimism is on how people feel when
from very optimistic to very pessimistic, facing problems. When people confront dif
with most falling somewhere between. An ficulty, the emotions they experience range
other issue that should be mentioned is that from excitement and eagerness to anger,
although trait optimism is thought of as a anxiety, and depression. The balance among
stable trait, moment-to-moment confidence such feelings relates to differences in opti
is subject to situational influences as well. mism. Optimists expect good outcomes,
For example, as people prepare themselves even when things are hard. This expectation
to encounter threats or undesired outcomes, yields a relatively positive mix of feelings. Pes
their momentary states of confidence may simists expect bad outcomes. This expecta
shift downward, whether they are basically tion yields more negative feelings— anxiety,
optimists or pessimists (Sweeny, Carroll, & anger, sadness, or even despair (Carver &
Shepperd, 2006). Scheier, 1998; Scheier & Carver, 1992).
332 IV. C O G N IT IV E D IS P O S IT IO N S
Relations between optimism and distress mism about life appeared to lead to a specific
have been examined in people facing a wide optimism about the surgery, and from there
range of difficulties, including students enter to satisfaction with life. A similar study by
ing college (Aspinwall & Taylor, 1992; Bris- Scheier and colleagues (1989) found that op
sette, Scheier, & Carver, 2002); survivors of timists retained higher quality of life even up
missile attacks (Zeidner & Hammer, 1992); to 5 years after the surgery.
people caring for cancer patients (Given et Optimism has also been studied in the
al., 1993) or Alzheimer’s patients (Hooker, context of other health crises. An example
M onahan, Shifren, & Hutchinson, 1992; is treatment for breast cancer (Carver et al.,
Shifren & Hooker, 1995); and people deal 1993). Women were interviewed at diagno
ing with childbirth (Carver & Gaines, 1987), sis, the day before surgery, a few days after
coronary artery bypass surgery (Fitzgerald, surgery, and 3, 6, and 12 months later. O p
Tennen, Affleck, & Pransky, 1993; Scheier et timism (at initial assessment) predicted less
al., 1989), aging (Giltay, Zitm an, & Krom- distress over time, controlling for effects of
hout, 2 0 0 6 ), failed attempts at in vitro fer medical variables and earlier distress. Thus
tilization (Litt, Tennen, Affleck, & Klock, optimism predicted not just lower initial
1992), bone marrow transplantation (Cur- distress but also resilience against distress
bow, Somerfield, Baker, Wingard, & Legro, during the following year. A study of head
1993), cancer (Carver et al., 1993; Friedman and neck cancer patients yielded similar re
et al., 1992), and the progression of AIDS sults (Allison, Guichard, & Gilain, 2000).
(Taylor et al., 1992). Patients were assessed before treatment and
The studies vary in complexity and thus in 3 months afterward. Optimists reported
what they can show. Researchers sometimes higher quality of life before treatment and
examine responses to an adverse event at one also posttreatment, controlling for initial
time point. Such studies consistently show ratings.
that greater pessimism relates to reports Another medical context in which opti
of experiencing more distress. W hat those mism has been studied is in vitro fertiliza
studies cannot show is whether pessimists tion, a procedure that lets people overcome
had more distress even before this particular fertility problems. This study focused on
adverse event. Other studies assess people at people whose attempts at in vitro fertiliza
multiple times. This gives a better picture of tion were unsuccessful (Litt et al., 1992).
how distress shifts over time and changing Eight weeks beforehand, the researchers
circumstances. It also allows researchers to measured optimism, expectancies for fertil
control for initial levels of distress. We focus ization success, distress, and the impact of
here on this sort of research. infertility on participants’ lives. Two weeks
A very early study of optimism and emo after notification of a negative pregnancy
tional well-being (Carver & Gaines, 1987) test, distress was measured again. None of
examined the development of depressed feel the initial variables predicted follow-up dis
ings after childbirth. Women completed the tress (controlling for Time 1 distress) except
original L O T and a depression scale in the optimism.
last third of their pregnancies. They com Yet another context in which effects of
pleted the depression measure again 3 weeks optimism have been examined is treatment
after delivery. Optimism was related to for ischemic heart disease. In this study
lower depression symptoms at initial assess (Shnek, Irvine, Stewart, & Abbey, 2001),
ment, and optimism predicted lower depres pessimism related to more symptoms of de
sion postpartum, controlling for the initial pression shortly after hospitalization for this
levels. Thus optimism conferred resistance disease. Furthermore, pessimism related to
to postpartum depressive symptoms. more symptoms of depression at a 1-year
Several projects have studied people deal follow-up, even when controlling for earlier
ing with coronary artery bypass. One as depression and a variety of other variables.
sessed people a month before surgery and 8 N ot only does optimism have a positive
months afterward (Fitzgerald et al., 1993). effect on the psychological well-being of
Optimists had less presurgical distress and people dealing with medical conditions, but
(controlling for presurgical life satisfaction) it also influences well-being among caregiv
more postsurgical life satisfaction. Opti ers. One project studied a group of cancer
22. O p tim ism 333
patients and their caregivers (Given et al., situational coping responses and general
1993). Caregivers’ optimism predicted less coping styles that differed from those of pes
depression and less impact of caregiving on simists (Scheier, Carver, & Bridges, 2001).
their physical health. Similar results were Optimism related to problem-focused cop
found in research on caregiver spouses of ing, especially in controllable situations.
Alzheimer’s patients (Hooker et al., 1992; Optimism also related to positive reframing
Shifren & Hooker, 1995): Optimism related and a tendency to accept the situation’s real
to lower depression and greater well-being. ity. Optimism related to less denial and less
Other studies have looked at events that of an attempt to distance oneself from the
are challenging, but less extreme. For ex problem. Thus optimists appear generally to
ample, the start of college is a stressful time, be approach copers, and pessimists appear
and studies have examined students adjust to be avoidant copers.
ing to their first semester of college (Aspin- Other projects have studied coping strate
wall & Taylor, 1992; Brissette et al., 2002). gies in specific contexts. Indeed, several stud
Optimism and other variables were assessed ies described earlier also looked at coping.
when the students first arrived on campus. In their study of coronary artery bypass sur
Measures of well-being were obtained at the gery, Scheier and colleagues (1989) assessed
end of the semester. Higher initial optimism attentional-cognitive strategies as ways of
predicted less distress at the end of the se dealing with the experience. Before surgery,
mester, along with greater development of optimists more than pessimists reported
friendship networks. making plans for their future and setting
Indeed, the simple process of aging con goals for recovery. Optimists also focused
fronts people with a variety of circumstances less on negative aspects of the experience—
that are difficult to adjust to. A Dutch study distress and symptoms. Once surgery was
of older men examined the role of personal past, optimists were more likely than pes
ity at the initial assessment as a predictor of simists to report seeking out information
depression across a 15-year follow-up (Gilt- about what the physician would require of
ay et al., 2 0 0 6 ). Optimism proved to predict them in the months ahead. Optimists also
significantly lower cumulative incidence of were less likely to say that they were sup
depression symptoms. pressing thoughts about their symptoms.
There was also evidence that the positive im
pact of optimism on quality of life 6 months
O ptim ism , Pessim ism , and Coping later occurred through the indirect effect of
these differences in coping.
If optimists experience less distress than pes The study of failed in vitro fertilization
simists when dealing with difficulties, is it described earlier (Litt et al., 1992) also ex
just because they are cheerful? Apparently amined coping. Pessimism related to escape
not, because the differences often remain as a coping response. Escape, in turn, led to
even when statistical controls are included more distress after the fertilization failure.
for prior distress. There have to be other Optimists were also more likely than pes
explanations. This section addresses one of simists to report feeling they had benefited
them: coping strategies. In many ways, this from the experience, for example, by becom
is just a more detailed depiction of the broad ing closer to their spouses.
behavioral tendencies discussed at the outset. Information on coping also comes from
That is, people who are confident about the the study of AIDS patients described earlier
future continue trying, even when it’s hard. (Taylor et al., 1992). Optimism predicted
People who are doubtful try to escape the positive attitudes and tendencies to plan for
adversity by wishful thinking, they employ recovery, seek information, and reframe bad
temporary distractions that do not help to situations more positively. Optimists used
solve the problem, and they sometimes even less fatalism, self-blame, and escapism, and
stop trying. they didn’t focus on negative aspects of the
Differences in coping that correspond situation or try to suppress thoughts about
to this divergence in behavior have been their symptoms. Optimists also appeared to
observed in several studies. Early projects accept unchangeable situations rather than
found that optimistic students reported both trying to escape them.
334 IV. C O G N I T I V E D I S P O S I T I O N S
Relations between optimism and cop that is no longer valid. Acceptance implies
ing also have been studied among cancer restructuring one’s perceptions to come to
patients. Stanton and Snider (1993) found grips with the situation. This does not mean
that pessimistic women used more cogni giving up. That response does not help. In
tive avoidance in coping with an upcoming fact, reacting to illness with resignation
biopsy than optimists. The avoidance ap may actually hasten death (Greer, M or
peared to mediate the relation of pessimism ris, Pettingale, & Haybittle, 1990; Reed,
to prebiopsy distress. Cognitive avoidance Kemeny, Taylor, Wang, & Visscher, 1994).
prebiopsy also predicted postbiopsy distress Acceptance of the diagnosis has very differ
among women with positive diagnoses. ent consequences. By accepting that life is
Another study of cancer patients, men compromised (but not over), people develop
tioned earlier, examined how women coped adaptive parameters within which to live
with treatment for breast cancer during the the time that is left. Acceptance may actu
first year (Carver et al., 1993). Both before ally serve the purpose of keeping the person
and after surgery, optimism was related to goal-engaged and, indeed, “life-engaged”
coping that involved accepting the reality of (Scheier 8c Carver, 2001).
the situation, placing as positive a light on
it as possible, and trying to relieve the situ
Prom oting W ell-B eing
ation with humor. Pessimism was related to
denial and tendencies to give up at each time Another aspect of coping is proactive or
point. The coping responses related to op preventive coping, processes that promote
timism and pessimism were also related to good health and well-being rather than just
distress. Further analyses revealed that the reacting to adversity. Perhaps optimists take
effect of optimism on distress was largely active steps to ensure positive outcomes in
indirect through coping, particularly post their future. This would resemble problem-
surgery. focused coping, except that it is intended to
Another study also examined the role of prevent a stressor from arising.
coping in women treated for breast cancer There are many ways in which this might
(Schou, Ekeberg, & Ruland, 2005). Two occur. An example is seeking knowledge per
coping strategies mediated the relationship taining to areas of potential risk. One study
between optimism and pessimism and qual investigated heart attack-related knowledge
ity of life 1 year after diagnosis. The greater in a group of middle-aged adults. It might
fighting spirit of optimists (assessed before be assumed that adults who are optimistic
diagnosis) predicted better quality of life at about their health would not make much ef
the 1-year follow-up. Hopelessness/helpless fort to learn about risks related to heart at
ness (reported by pessimists) predicted poor tacks. Those high in dispositional optimism,
er quality of life. however, actually knew more about the risk
In sum, it appears that optimists differ factors than those who were less optimistic
from pessimists both in stable coping ten (Radcliffe & Klein, 2002).
dencies and in the coping responses gener Proactive efforts in health promotion have
ated when confronting stressful situations also been examined among patients in a car
(for a detailed review, see Solberg Nes & diac rehabilitation program (Shepperd, Ma-
Segerstrom, 2 0 0 6 ). In general, optimists roto, & Pbert, 1996). Optimism was related
use more problem-focused coping strate to success in lowering levels of saturated fat,
gies than pessimists. When problem-focused body fat, and an index of overall coronary
coping is not a possibility, optimists turn to risk. Optimism was also related to increases
strategies such as acceptance, use of humor, in exercise. Another study of the lifestyles of
and positive reframing. Pessimists tend to coronary artery bypass patients 5 years after
cope through overt denial and by mentally surgery found optimists more likely than
and behaviorally disengaging from the goals pessimists to be taking vitamins and eating
with which the stressor is interfering. low-fat foods and to be enrolled in a cardiac
Particularly noteworthy is the contrast be rehabilitation program (Scheier & Carver,
tween acceptance and active denial. Denial 1992).
(refusing to accept the reality of the situa Another proactive health-related behavior
tion) means trying to maintain a worldview concerns H IV risk. By avoiding certain sex
22. O p tim ism 335
ual practices (e.g., sex with unknown part alcohol abuse were followed as they entered
ners), people reduce risk of infection. One an aftercare program. Pessimists were more
study of HIV-negative gay men found that likely to drop out of that program and to re
optimists reported fewer anonymous sexual turn to drinking than were optimists (Strack,
partners than pessimists (Taylor et al., 1992). Carver, & Blaney, 1987). Yet another study
This suggests that optimists were making ef (Park, M oore, Turner, & Adler, 1997) found
forts to reduce their risk, thus safeguarding that optimistic pregnant women were less
their health. likely to engage in substance abuse during
Optimists appear to take action to mini the course of their pregnancies.
mize health risks. They do not simply stick A more recent study examined a different
their heads in the sand and ignore threats indicator of giving up: the disruption of nor
to well-being. They attend to risks, but se mal social activities. Breast cancer patients
lectively. They focus on risks that are ap reported illness-related disruption of social
plicable to them and that relate to poten activities after treatment (Carver, Lehman,
tially serious health problems (Aspinwall & Antoni, 2003). At each assessment, pessi
& Brunhart, 1996). If the potential health mism predicted more disruption, along with
problem is minor, or if it is unlikely to bear emotional distress and fatigue. When con
on them, they are not especially vigilant. fronted with a health threat, pessimism led
Optimists appear to scan their surroundings to a withdrawal from the very social activi
for threats to well-being but save their be ties that are important to a normal life.
havioral responses for threats that are truly Giving up can be reflected in many ways.
meaningful. Alcohol dulls awareness of failures and
Is it paradoxical that people who expect problems. People can ignore problems by
good things to happen take active steps to distracting themselves with other activities.
make sure good things do happen? Maybe. Sometimes, though, giving up is more com
But years of experience presumably teach plete. Sometimes people give up not just on
people that their own efforts play an im specific goals, but on their lives, by suicide.
portant part in many kinds of outcomes in Some are more vulnerable to suicide than
life. Optimists may be more confident than others. It is commonly assumed that depres
pessimists that their efforts will be success sion is the best indicator of suicide risk. But
ful. For that reason, they are quicker to en at least one study found that pessimism was
gage in those efforts when there is a need actually a stronger predictor of this act, the
for them. ultimate disengagement from life (Beck,
Steer, Kovacs, & Garrison, 1985).
In sum, a sizeable body of evidence indi
Pessim ism and H ealth-D efeating Behaviors
cates that pessimism can lead people into
We have characterized optimists throughout self-defeating patterns. The result can be less
this chapter as being persistent in trying to persistence, more avoidance coping, health-
reach goals. Theory suggests that pessimists damaging behavior, and potentially even an
are less persistent and more likely to give up. impulse to escape from life altogether. W ith
There is, in fact, evidence of giving-up ten out confidence about the future, there may
dencies among pessimists, with bad conse be nothing to sustain life.
quences. For example, giving up may under
lie various forms of substance abuse, such as
D oes O ptim ism H ave a D ow n Side
excessive alcohol use, which is often seen as
in C oping and Goal P ursuit?
an escape from problems. This suggests that
pessimists should be more vulnerable than Although most evidence on coping and
optimists to such maladaptive behavior. Evi confronting difficulty links optimism to
dence supports this reasoning. adaptive, engaged coping, some have asked
One study of women with family histories whether the optimistic stance on life may
of alcoholism found that pessimists in that also have a down side. Confidence and per
group were more likely than optimists to re sistence are good, but they can potentially
port drinking problems (Ohannessian, Hes- lead to problems. Consider gambling, a form
selbrock, Tennen, & Affleck, 1994). In an of entertainment that can create major prob
other study, people who had been treated for lems for people who engage in it too much.
336 IV. C O G N I T I V E D I S P O S I T I O N S
Problem gambling can result in loss of large switch to. In effect, they gave up on a task
amounts of money, and it often leads to ad they could not master in order to turn to a
ditional problems in work and relationships. similar task that they could master. Indeed,
Gibson and Sanbonmatsu (2004) reasoned if they had been led to think that the other
that gambling is a context in which positive task measured a somewhat different skill,
expectancies and the resulting persistence they even outperformed the less optimistic
might be counterproductive. They found a people.
variety of worrisome tendencies among op Yet another set of studies deals with the
timists. They had more positive expectations question of whether optimism causes people
for gambling than did pessimists, and they to see only what they want to see and ignore
were also less likely to reduce their betting threats. The initial evidence suggested the
after poor outcomes. The people studied in opposite: that optimists pay closer attention
that research were not people with gambling to information about health threats than
problems. But this pattern suggests the pos pessimists, provided the threat is serious and
sibility that optimists may be more likely to relevant to them (Aspinwall 8c Brunhart,
develop such problems than pessimists. 1996). More recently, Luo and Isaacowitz
Another set of studies deals with the ques (2007) found the opposite pattern. Several
tion of whether the persistence of optimists other studies have found that optimism is as
results in problems because they are unable sociated with an attentional bias toward pos
to recognize what they cannot accomplish. itive rather than negative stimuli (Isaacowitz,
More simply, perhaps they don’t know when 2 0 0 5 ; Segerstrom, 2001). For example, opti
it is best to quit. Certainly there are circum mism has been linked to shorter times look
stances in which people must recognize that ing at pictures of skin cancers (Isaacowitz,
goals have been lost and that the adaptive 2005). Exactly how to interpret the aggre
course is to turn away from their pursuit gated information is not clear. It may be, for
(Wrosch, Scheier, Carver, 8c Schulz, 2003). example, that optimists prefer to attend to
Does the persistence that follows from opti positively valenced stimuli but are quick to
mism prevent that from occurring? encode threat-related information when that
One project relevant to this question was information is perceived as useful to them.
based on the reasoning that greater per In sum, there are cases in which optimism
sistence should lead to the development of has a down side. The question of when there
greater goal conflict, partly because com are costs as well as benefits to optimism is
mitment to many goals causes people to certain to receive additional scrutiny in fu
spread their resources thinner (Segerstrom ture work.
& Solberg Nes, 20 0 6 ). Two studies (one
of them prospective) found evidence that
optimism did relate to such an elevation in O ptim ism and Physical W ell-Being
goal conflict. However, this conflict did not
have adverse psychological consequences. The previous sections included frequent
Evidence from the second study suggested mention of stress and coping with medical
that optimistic people balanced expectancy, problems. As implied by that, much of the
value, and cost of goal pursuit more effec work on optimism has been conducted in
tively than did pessimistic people. They were the domain of health psychology. Some of
committed to more incompatible goals, but that research has gone on to examine opti
they were more efficient at managing the mism and physical well-being. The general
conflict. line of thinking underlying such research
Another project (Aspinwall 8c Richter, is that optimists may be less reactive than
1999) examined participants’ willingness pessimists to the general stresses of life; that
to disengage from tasks on which they were the lower physiological stress responses may
unable to succeed (the task actually was im (over many years) result in less physical wear
possible). In one condition, there was no al and tear on the body; and that the end result
ternative task to turn to; in other conditions, may be better physical health and potential
there was an alternative. Optimism related ly greater longevity.
to faster disengagement from the impossible In one study of physical well-being, mid-
task when there was an alternative task to dle-aged women were tested for carotid in-
22. O p tim ism 337
govern, G. J., Lefebvre, R. C., Abbott, R. A., et al. Solberg Nes, L., &c Segerstrom, S. C. ( 2 0 06 ). Dispo
(1989). Dispositional optimism and recovery from sitional optimism and coping: A meta-analytic re
coronary artery bypass surgery: The beneficial ef view. P erso n a lity a n d S o c ia l P sy c h o lo g y R e v iew ,
fects on physical and psychological well-being. 10, 2 3 5 - 2 5 1 .
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 57, Srivastava, S., M cGonigal, K. M ., Richards, J. M.,
1024-1040. Butler, E. A., 8c Gross, J . J . (200 6). Optimism in
Scheier, M. F., Matthews, K. A., Owens, J. F., Schulz, close relationships: How seeing things in a positive
R ., Bridges, M. W., Magovern, G. J ., Sr., et al. light makes them so. J o u r n a l o f P erson ality a n d S o
(1999). Optimism and rehospitalization following c ia l P sy c h o lo g y , 91, 1 4 3 - 1 5 3 .
coronary artery bypass graft surgery. A rch iv es o f Stanton, A. L., 8c Snider, P. R. (1993). Coping with
In te rn a l M ed icin e, 159, 8 2 9 - 8 3 5 . breast cancer diagnosis: A prospective study. H ea lth
Schou, I., Ekeberg, O., 8c Ruland, C. M. (2 005). The P sy ch olog y , 12, 1 6 - 2 3 .
mediating role of appraisal and coping in the rela Strack, S., Carver, C. S., 8c Blaney, P. H. (1987). Pre
tionship between optimism-pessimism and quality dicting successful completion of an aftercare pro
of life. P s y c h o -O n c o lo g y , 14, 7 1 8 - 7 2 7 . gram following treatment for alcoholism: T he role
Segerstrom, S C. (2001). Optimism and attentional of dispositional optimism. J o u r n a l o f P erson ality
bias for negative and positive stimuli. P erson ality a n d S o c ia l P sy ch olog y , 5 3 , 5 7 9 - 5 8 4 .
a n d S o c ia l P sy c h o lo g y B u lletin , 27, 1 3 3 4 - 1 3 4 3 . Sweeny, K., Carroll, P. J . , 8c Shepperd, J . A. (20 06 ). Is
Segerstrom, S. C. (2 005). Optimism and immunity: optimism always best? C u rren t D ire c tio n s in P sy
Do positive thoughts always lead to positive effects? c h o lo g ic a l S c ien ce , 15, 3 0 2 - 3 0 6 .
B rain , B eh a v io r , a n d Im m u n ity , 19, 1 9 5 - 2 0 0 . Szondy, M. (20 04 ). Optimism and immune func
Segerstrom, S. C. (2006a). B r ea k in g M u rp h y ’s la w : tions. M e n ta lh ig ien e es P s z ic h o s z o m a tik a , 5, 3 0 1 -
H o w o p tim is ts g et w h a t th ey w an t fr o m life — a n d 320.
p essim ists t o o . New York: Guilford Press. Taylor, S. E., Kemeny, M. E., Aspinwall, L. G.,
Segerstrom, S. C. (2 006b). How does optimism sup Schneider, S. G., Rodriguez, R., 8c Herbert, M.
press immunity?: Evaluation of three affective path (1992). Optimism, coping, psychological distress,
ways. H ea lth P sy ch olog y , 2 5 , 6 5 3 - 6 5 7 . and high-risk sexual behavior among men at risk
Segerstrom, S. C. ( 2 0 0 7 ). Optimism and resources: for acquired immunodeficiency syndrome (AIDS).
Effects on each other and on health over 10 years. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 6 3 ,
J o u r n a l o f R e s e a r c h in P ers o n a lity , 4 1 , 7 7 2 - 7 8 6 . 460-473.
Segerstrom, S. C., 8c Solberg Nes, E. ( 2 0 06 ). When Trunzo, J. J ., 8c Pinto, B. M. (2 003). Social support as
goals conflict but people prosper: The case of dispo a mediator of optimism and distress in breast cancer
sitional optimism. J o u r n a l o f R e se a rc h in P ers o n a l survivors. J o u r n a l o f C o n su ltin g a n d C lin ica l P sy
ity, 4 0 , 6 7 5 - 6 9 3 . ch o lo g y , 4, 8 0 5 - 8 1 1 .
Shepperd, J. A., M aroto, J . J ., 8c Pbert, L. A. (1996). Uchino, B. N. (200 4 ). S o c ia l s u p p o r t a n d p h y sic a l
Dispositional optimism as a predictor of health h e a lth : U n d erstan d in g th e h ea lth c o n s e q u e n c e s
changes among cardiac patients. J o u r n a l o f R e o f re la tio n sh ip s. New Haven, CT: Yale University
se a r c h in P erson ality , 3 0 , 5 1 7 - 5 3 4 . Press.
Shifren, K., 8c Hooker, K. (1995). Stability and change Wrosch, C., Scheier, M . F., Carver, C. S., 8c Schulz,
in optimism: A study among spouse caregivers. E x R. (2003). The importance of goal disengagement
p e r im e n t a l A gin g R esea rch , 2 1 , 5 9 - 7 6 . in adaptive self-regulation: When giving up is ben
Shnek, Z. M ., Irvine, J., Stewart, D., 8c Abbey, S. eficial. S e lf a n d Id en tity , 2 , 1 - 2 0 .
(20 01). Psychological factors and depressive symp Zeidner, M., 8c Hammer, A. L. (1992). Coping with
toms in ischemic heart disease. H ea lth P sy ch o lo g y , missile attack: Resources, strategies, and outcomes.
2 0 , 1 41-145. J o u r n a l o f P erson ality , 6 0 , 7 0 9 - 7 4 6 .
CHAPTER 23
A r ie W. K r u g la n sk i
S h i r a F is h m a n
343
344 IV. C O G N I T I V E D I S P O S I T I O N S
is assumed that the motivation toward clo formation processing in opposite directions.
sure varies along a continuum anchored at For example, when the need for closure is
one end with a strong NFC and at the other elevated, an individual may consider lim
end with a strong need to avoid closure. The ited information and rely on preconceived
N FC is elevated when the perceived benefits notions or stereotypes. When the need for
of possessing closure and/or the perceived closure is lower, however, one may be will
costs of lacking closure are high (Kruglanski ing to consider ample evidence before mak
& Webster, 1996; Webster & Kruglanski, ing up one’s mind. Such epistemic dynamics
1994). Likewise, the need to avoid closure is prompted by the NFC are not assumed to
elevated when the perceived benefits of lack be consciously accessible to the knower but,
ing closure and the perceived costs of pos rather, exert their effects implicitly and typi
sessing closure are high. These benefits and cally outside of awareness.
costs vary according to situational factors
and individual differences.
Individual Differences
als low in NFC (Mayseless & Kruglanski, search, despite the other student not having
1987). Thus, individuals high in N FC may had any choice in writing the essay, partici
seek less information about another person pants reported that the student’s actual at
before reaching a conclusion or forming a titude was similar to the perspective taken
definite impression about this person. in the essay. However, this effect was magni
fied when individuals were high rather than
low in N FC.
C u e Utilization
NFC also appears to heighten people’s reli
Stereotype Application
ance on initial information. The urgency ten
dency predisposes an individual to quickly Increased application of prevalent social
“seize” on early cues and utilize them toward stereotypes and prejudices to various so
the formation of initial judgments, whereas cial judgments may represent a particularly
the permanence tendency predisposes an in striking case of seizing and freezing under a
dividual to “freeze” or fixate on those par heightened N FC. Because culturally preva
ticular judgments. Research on a variety of lent stereotypes constitute knowledge struc
seemingly diverse social-psychological top tures that may readily come to mind, they
ics has lent support to these ideas. may be particularly likely to serve as bases
for judging stereotyped targets when the
perceiver is high (vs. low) in N FC. This pos
Primacy Effects in Impression Formation
sibility was supported in several studies and
A primacy effect refers to the tendency to with several different stereotypic contents,
base one’s social impressions on early in such as the stereotypes of Ashkenazi and
formation about that person, to the relative Sephardic Israelis (Kruglanski & Freund,
neglect of subsequent, potentially relevant 1983), of women in management (Jamieson
information. From the present perspective, & Zanna, 1989), and of soccer hooligans
primacy effects exemplify the seizing-and- and nurses (Dijksterhuis, van Knippenberg,
freezing tendencies that are assumed to be Kruglanski, & Schaper, 1996).
stronger for individuals who are high in
N FC. Indeed, when individuals are high in
Construct Accessibility
N FC , primacy effects are augmented (Web
ster & Kruglanski, 1994). In addition, the Stereotypes are more accessible to individ
higher the individual’s N FC , the stronger uals who are high in the need for closure,
the magnitude of the primacy effect. coming to mind more easily when they are
confronted with a judgmental target. In
deed, Ford and Kruglanski (1995) found
The Correspondence Bias
that, compared with individuals low in
The correspondence bias (Jones, 1979), or N FC , those high in N FC relied to a greater
the fundamental attribution error (Ross, extent on a previously primed concept when
1977), refers to the attributor’s tendency to judging an ambiguous target.
overascribe an actor’s behavior to her or his
unique attitudes or personality and to un
Recency Effects
derestimate the power of the situation. Just
as with primacy and anchoring effects, the In some circumstances, N FC may lead to
correspondence bias reflects the seizing- recency effects rather than primacy ef
and-freezing tendencies of individuals who fects, depending on construct accessibil
are high in N FC. Webster (1993b) asked ity. Specifically, timing of the impression-
participants to complete a typical attitude- formation goal should moderate the use of
attribution task in which they estimated a primacy versus recency heuristics. When the
target’s attitude after hearing her deliver a impression-formation goal exists from the
speech criticizing student exchange pro start, high NFC should predict an enhanced
grams with foreign universities. The speech primacy effect due to a seizing and freezing
was allegedly prepared under either high- or on the initial information. However, when
low-choice conditions. As in previous re the impression-formation goal is introduced
23. N eed for C o g n itiv e C losu re 347
following exposure to the stimulus materi in perspective taking and empathy emerged
als, participants should rely on their memo when the target and participant were similar
ries of the information, and high (vs. low) to each other.
NFC should predict a stronger recency ef
fect. Data from experimental studies con
Interpersonal Com m unication
firm these predictions (Richter & Kruglan
ski, 1999). When conveying messages to others, speak
Taken together, the research on intrap ers often take the audience’s perspective into
ersonal processes demonstrates that people account and make reference to the realities
who are high in NFC seek less information, that both parties share. However, under time
generate fewer hypotheses, and rely on early, pressure, speakers are less likely to reference
initial information when making judgments. common ground. As time pressure has been
Paradoxically, despite the reliance on less, one of the major ways in which N FC has
and perhaps incomplete, information, indi been operationalized (Kruglanski & Fre
viduals high in NFC display greater confi und, 1983; Shah, Kruglanski, & Thom p
dence in their decisions. son, 1998), a high-level NFC may reduce
the amount of effort communicators invest
in their search for common ground. As a
Interpersonal Phenom ena consequence, communications by individu
als high in NFC may be excessively biased
The processing of information about one’s in the direction of the communicator’s own
interaction partners and the formation of perspective, which might reduce their com
online judgments about their feelings, cog prehensibility to listeners. Richter and Krug
nitions, and probable actions is fundamen lanski (1999) investigated this hypothesis
tal to social relations and should be conse by asking participants to write descriptions
quential for interpersonal relationships. The of figures and then, on a subsequent visit,
seizing-and-freezing tendencies fostered by to match the descriptions to the pictures.
a heightened N FC should, therefore, exert Participants high (vs. low) in N FC used sig
important effects at the interpersonal level nificantly fewer words in their descriptions,
of analysis. and their descriptions were less likely to be
matched with the figure, as compared with
descriptions written bv individuals low in
Perspective Taking and Em pathy
N FC.
Taking another’s perspective often requires The way that individuals converse with
substantial cognitive effort, as one needs to each other should also be affected by N FC.
overlook one’s own perspective and focus on If NFC induces the tendency to seek per
the perspective of another. In this vein, if the manent knowledge and reduce ambiguity,
need for cognitive closure reduces individu then individuals high in N FC should prefer
als’ readiness to put effort into mental pro abstract descriptions and category labels to
cessing and predisposes them to seize and concrete, situational descriptions. Indeed,
freeze on early information, it may reduce evidence supports this idea (Boudreau,
perspective taking and empathic concerns Baron, & Oliver, 1992; Mikulincer, Yinon,
when individuals high in the need for closure & Kabili, 1991). In another study, individu
interact with dissimilar others. To examine als with high (vs. low) N FC preferred to ask
this possibility, Webster-Nelson, Klein, and more abstract interview questions (Rubini
Irvin (2003) had participants read descrip & Kruglanski, 1997). A subsequent study
tions of a person who was either similar or found that abstract questions from indi
dissimilar to themselves. Under a heightened viduals high in NFC elicited more abstract
NFC (experienced through mental fatigue), answers from respondents. The level of ab
the ability to take a different perspective straction was, in turn, related to liking, with
was reduced when the target was dissimi more abstract questions eliciting less liking
lar to the participant. Similarly, the ability from respondents. The latter decrease in lik
to show empathy was lower when the target ing occurred because the subject of abstract
was dissimilar. As expected, no differences questions is usually an object (e.g., because
348 IV. C O G N I T I V E D I S P O S I T I O N S
dogs are . . . ) rather than the self (e.g., be dividuals high in N FC are focused on their
cause I like dogs . . . ). own perspective, making it more difficult for
Individuals high in NFC prefer abstract others to understand their views and com
labels because they can be applied across munications. Individuals high in N FC prefer
a variety of situations, implying epistemic to use abstract labels, which can be applied
permanence. In different terms, abstract across various situations. Lastly, individuals
expressions are “multifinal” (Kruglanski, high in N FC are quick to apply significant-
Shah, Fishbach, et al., 2002) in that they other schemas to individuals who resemble
satisfy multiple goals (provide multiple clo them superficially, potentially producing
sures) with a single means and thus should substantial errors of person perception.
be preferred. A set of studies by Chun and
Kruglanski (2005) demonstrated that indi
viduals high in N FC preferred proverbs that Group Processes
espoused the multifinality idea (e.g., “kill
ing two birds with one stone”) as compared As a result of the tendency to seize and freeze
with proverbs that argued the opposite (e.g., on accessible information, individuals high
“if you run after two hares, you will catch in NFC exhibit a number of intriguing be
neither”). Furthermore, people who are high havioral tendencies within group settings.
in N FC pursue multifinality even at the ex
pense of quality or cost. Finally, when the
Task versus Socioemotional O rientation
number of goals was held constant, indi
viduals high in NFC chose fewer means to When assigned a group task, group mem
achieve the goal. bers may choose to focus on either task or
socioemotional goals. Because the task rep
resents the most accessible construct for de
Transference
fining the situation (because it is the obvious
Andersen and colleagues’ work (e.g., Ander reason for individuals finding themselves in
sen &c Berk, 1998; Andersen & Chen, 2 0 0 2 ; that situation), people who are high in NFC
Andersen & Cole, 1990) on the transference may be more task oriented than socioemo-
effect in social judgment demonstrates how tionally oriented. De Grada, Kruglanski,
a significant other’s schema stored in mem M annetti, and Pierro (1999) asked groups
ory can be applied to a new individual who of four students to role-play the managers
resembles the significant other in some way. of four corporate departments while nego
Information about a new individual who is tiating a monetary reward for a meritorious
similar to a significant other may activate the worker. Participants high in NFC produced
significant-other schema, which is then used a higher proportion of task-oriented re
to make (often inaccurate) inferences about sponses and a lower proportion of positive
the newly encountered individual. Such social-emotional acts than participants low
transference effects have been explained in in N FC.
terms of the high accessibility of the sig
nificant other’s representation in memory.
Consensus Striving
Given that N FC implies seizing and freezing
on accessible constructs, the transference ef W ithin a group, individuals strive toward
fect should be more pronounced under high homogeneity of opinions (Festinger, 1950).
NFC. Indeed, research by Pierro and Krug From an NFC perspective, such homogene
lanski (2008) found that the transference ef ity is essential for epistemic certainty; if so,
fect is more pronounced for individuals high individuals high in N FC should show a high
rather than low in NFC. er desire for consensus. Consistent with that
In summary, individual differences in prediction, De Grada and colleagues (1999)
N FC , as well as situational differences in found that during a negotiation session with
N FC , have important implications for social in a group, members of groups composed of
interaction. Individuals high in NFC (vs. low individuals high in NFC felt greater pressure
in NFC) have greater difficulty taking other toward uniformity as compared with mem
people’s perspectives and empathizing with bers of groups low in N FC. Blind coders,
them. While communicating with others, in unaware of differences in group composi
23. N eed for C o gn itive C losure 349
tion, confirmed this finding by rating social allows fewer opinions to be voiced dur
pressures higher in groups with high (vs. ing the course of the discussion. Indeed, a
low) N FC. In a different research paradigm, number of studies have found that groups
Kruglanski and colleagues (1993) found that composed of individuals high in N FC fos
when individuals high in N FC entered the ter the emergence of autocratic leadership to
situation with considerable certainty in their a greater extent than groups that are com
views, they refused to change their views, posed of individuals low in N FC (De Grada
even when others disagreed with them; et al., 1999; Pierro, M annetti, De Grada,
whereas when individuals high in NFC en Livi, & Kruglanski, 2003).
tered the situation with little confidence in Further exploring autocratic leadership,
their views, they showed a greater inclina Pierro and Kruglanski (2008) studied the
tion to change their own views toward their influence styles preferred by leaders high in
partner’s opinion. NFC and their subordinates. “Soft” power
Individuals can also seek consensus by re bases allow greater autonomy and are less
jecting members who deviate from the ma controlling than “hard” power bases (Raven,
jority opinion (Festinger, 1950). In a situa Schwarzwald, & Koslowsky, 1998). Results
tion in which groups were required to reach showed that subordinates high in NFC pre
consensus on an issue, individuals high (vs. fer “hard” social influence tactics, whereas
low) in NFC showed greater tendency to re subordinates low in NFC prefer “soft” tac
ject an opinion deviate (Kruglanski & Web tics. Similarly, supervisors who are high in
ster, 1991). Importantly, when the groups NFC tend to employ “hard” tactics, whereas
were able to use a majority rule to reach a those low in NFC tend to employ “soft” tac
conclusion, high NFC did not predict re tics. Finally, evidence suggests that organi
jection of a deviate. Therefore, only when zations are more effective to the extent that
collective closure is required (via consensus the types of tactics used by the supervisor
agreement) is there a tendency to derogate fit the preferences of the subordinates (Pierro
the deviate. & Kruglanski, 2008).
Consensus can also be built around shared The desire to have consensus might imply
information, and groups tend to focus their an unwillingness to embrace change. A
discussions around such shared information study by Livi (2002) found that, over time
(Stasser & Stewart, 1992; Stasser & Titus, and with turnover, the norms established at
1985, 1987). Webster (1993b) found that initial stages of group formation persisted
during group discussion, NFC was inversely more in groups whose members were high
correlated with the tendency to bring up in N FC than in groups whose members were
unique information (information that was low in N FC. Similarly, research conducted
possessed solely by some and not by other in organizational settings has shown that
group members). Thus individuals high in individuals high in NFC have trouble cop
NFC seem to focus on shared information, ing with organizational change (Kruglanski,
presumably in order to create consensus Pierro, Higgins, & Capozza, 2007). How
more quickly, which allows them to achieve ever, in a culture that is very supportive of
closure. In the absence of generating new such change, individuals high in NFC had
ideas, a group may actually become less cre an easier time coping with change. In other
ative, especially to the extent that it focuses words, although individuals high in NFC
on shared information. Indeed, Chirum- are generally averse to change, they are also
bolo, M annetti, Pierro, Areni, and Krug generally supportive of the “social reality”
lanski (2005) found that groups composed of their particular organization. Thus, when
of individuals high (vs. low) in N FC tend to the existing social reality supports change,
produce fewer ideas, to elaborate on those individuals high in NFC adjust better to
ideas less fully, and to be less creative in a changes in the workplace.
mock advertising task. Taken together, the research on group
Given the desire for consensus within the processes and NFC indicates that individu
high-NFC group, such groups should sup als with high NFC desire consensus and ho
port leaders who make quick and decisive mogeneity among group members. As such,
decisions. Thus groups high in NFC may they are willing to engage in activities per
prefer an autocratic leadership style, which ceived as likely to achieve and maintain sta
350 IV. C O G N I T I V E D I S P O S I T I O N S
bility, including focusing on the task at hand, in N FC are likely to want to maintain that
pressuring others to change their opinions, reality; this may be achieved by keeping (po
rejecting those who hold different opinions, tentially dissimilar) others out of the group.
sharing less information with others, and
supporting an autocratic leadership style.
T h e Linguistic Intergroup Bias
The linguistic intergroup bias (LIB) is the
Intergroup Processes
tendency for group members to describe the
To the individual, the ingroup represents positive characteristics of the ingroup and
an important provider of social knowledge the negative characteristics of the outgroup
concerning norms for acting and thinking. in abstract terms, thereby implying stable
Because of this, the ingroup can provide traits. In contrast, individuals describe the
closure to the individual. If the ingroup is negative characteristics of the ingroup and
valued partly because it constitutes a clo the positive characteristics of the outgroup
sure provider, it should be valued more by in concrete terms, implying that the charac
individuals high versus low in N FC. Shah, teristics are situationally specific rather than
Kruglanski, and Thompson (1998) inves fundamental (Maass & Arcuri, 1992). A
tigated this hypothesis in a study in which study by Webster, Kruglanski, and Pattison
participants believed they would be com (1997) found that individuals high (vs. low)
peting in a group of two members against in N FC used significantly more abstract
another similar group. After reading alleged terms when describing positive ingroup and
self-descriptions of their partners and com negative outgroup behaviors, thereby exhib
petitors, individuals high in N FC reported iting the LIB.
more liking for their own teammates and In summary, research on the dynam
less liking toward the members of the other ics of N FC in intergroup contexts suggests
teams than individuals low in N FC. that individuals high in NFC seek to protect
If the ingroup represents a stable social re and maintain their ingroups. Indeed, indi
ality, individuals high in N FC should prefer viduals high in NFC are biased toward their
groups that are homogeneous in their com own ingroups. They exhibit greater liking
position over heterogeneous groups. H om o for ingroup members and greater LIB. Fur
geneous (vs. heterogeneous) groups are likely thermore, individuals high in N FC prefer
to agree on their basic worldviews and may ingroups that are homogeneous as well as
thus come to consensus faster. If individu similar to themselves; once those groups are
als high in N FC attach particular value to established, they support attempts to main
consensus, they should prefer homogeneous tain the group and exclude others from the
groups, but this should be true only to the group.
extent that the views of a given homoge
neous group agree with the individual’s own
views. Consistent with this analysis, Krug Conclusions and Future D irections
lanski, Shah, Pierro, and M annetti (2002)
found that individuals high (vs. low) on N FC has been conceptualized as the desire
NFC had a greater preference for homoge for a definite answer, as opposed to uncer
neous (vs. heterogeneous) groups, but only tainty or ambiguity. Because N FC allows
when the group’s opinions were similar (vs. individuals to reach a decision in the process
dissimilar) to their own opinions. of knowledge formation, NFC has impor
Related experiments by Dechesne, tant implications for social interaction. Indi
Schultz, Kruglanski, Fishman, and Orehek vidual differences in NFC predict behavior
(2007) found that individuals high (vs. low) at the intrapersonal, interpersonal, group,
in NFC preferred groups with impermeable and intergroup levels of analysis. Individu
(vs. permeable) boundaries, but only when als high in NFC generate fewer hypotheses
the groups were perceived as homogeneous, and seek less information before making a
suggesting a greater likelihood of consen decision. Similarly, they are focused on their
sus. Thus, if a group represents a source of own perspectives and have difficulty taking
stable social reality, as is likely to be the case the perspectives of others. They seek closure
with homogenous groups, individuals high within groups, pressuring others or changing
23. N eed for C o g n itiv e C losu re 351
their own opinions in order to reach group cal (e.g., during war times) and economic
consensus. Finally, they prefer solid bound circumstances (e.g., during an economic de
aries to their own homogenous groups. In pression) might also induce a sense of pro
all, individuals high in NFC desire quick found and aversive uncertainty contributing
closure, seizing and freezing on information to one’s craving for assurance and predict
that is quickly and efficiently able to provide ability. In short, despite the considerable
such closure. work thus far on the need for cognitive clo
Though considerable knowledge exists sure, substantial further work is needed to
concerning the situational antecedents of the fully understand this fundamental aspect of
need for closure and its varied consequences, human behavior.
a substantial gap exists in understanding the
conditions and circumstances that prompt
the development of individual differences in References
the need for closure. Intriguing recent work
Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J.,
by Kossowska and her colleagues (Kossows-
8c Sanford, R. N. (1950). T h e a u th o r ita r ia n p e r s o n
ka, Orehek, & Kruglanski, in press; Legier- ality. New York: Harper.
ski & Kossowska, 2008) suggests that indi Altemeyer, B. (1981). R ig h t-w in g a u th o rita ria n is m .
viduals with low working memory capacity Winnipeg, M anitoba, Canada: University of M a n i
may develop a high need for closure. In this toba Press.
Andersen, S. M ., Sc Berk, M . S. (1998). The social-
vein, Kossowska and colleagues (in press) cognitive model of transference: Experiencing past
found that individual differences in working relationships in the present. C u rren t D ire c tio n s in
memory capacity are correlated with indi P s y c h o lo g ic a l S cien ce , 7, 1-7.
vidual differences in the need for closure. In Andersen, S. M., Sc Chen, S. (2 002). The rational self:
An interpersonal social-cognitive theory. P sy ch o
addition, individual differences in working
lo g ic a l R ev iew , 109, 6 1 9 - 6 4 5 .
memory capacity were found to mediate the Andersen, S. M ., Sc Cole, S. W. (1990). “D o I know
relation between individual differences in you?”: The role of significant others in general so
the need for closure, the type of information cial perception. J o u r n a l o f P erson ality a n d S o c ia l
sought in a judgmental task (i.e., simple vs. P sy ch olog y , 59, 3 8 4 - 3 9 9 .
Boudreau, L. A., Baron, R ., Sc Oliver, P. V. (1992). Ef
complex information), and the extent of in fects of expected communication target expertise
formation search exhibited by participants. and timing of set on trait use in person description.
These results are promising, yet their cor P erso n a lity a n d S o c ia l P sy ch o lo g y B u lletin , 18,
relational nature prevents firm conclusions 447-452.
Chirumbolo, A., Man net ti, I.., Pierro, A., Areni, A.,
as to the causal relations between working Sc Kruglanski, A. W. (2005). Motivated closed
memory capacity and need for closure. Fur mindedness and creativity in small groups. S m all
ther probing, perhaps exploring the brain G r o u p R e se a rch , 3 6 , 5 9 - 8 2 .
mechanisms involved in closed-mindedness, Chiu, C., Morris, M. W., Hong, Y., Sc Menon, T.
is needed to fully understand the nature of ( 2 0 00 ). Motivated cultural cognition: The impact
of implicit cultural theories on dispositional attribu
these relations. tion varies as a function of need for closure. J o u r n a l
In addition, possible developmental ante o f P erson ality a n d S o c ia l P sy ch olog y , 7 8, 2 4 7 - 2 5 9 .
cedents of individual differences in need for Chun, W. Y., Sc Kruglanski, A. W. (2005). Consump
closure merit further explorations. Though tion as a multiple goal pursuit without awareness. In
F. R. Kardes, P. M. Herr, 8c J. Nantel (Eds.), A p p ly
prior work on individual differences in
ing s o c ia l c o g n itio n to c o n s u m e r -fo c u s e d strateg y
closed- and open-mindedness (e.g., Adorno (pp. 2 5 - 4 3 ) . Mah wah, NJ: Erlbaum.
et al., 1950; Rokeach, 1960; Sorrentino & Cratylus. (1995). De Nederlandse Need for Closure
Short, 1986) emphasized psychosexual de Schaal (The Netherlands Need for Closure Scale).
velopment as a fundamental antecedent of N e d e r la n d s e T ijd s c h rift V oor d e P sy ch o lo g ie, SO,
231-232.
these differences, little empirical evidence Dechesne, M ., Schultz, J., Kruglanski, A. W., Fishman,
to date exists concerning these claims. Ad S., Sc Orehek, E. (2 007). P sy c h o lo g y o f b o u n d a r y
ditional aspects of the socialization context c o n d itio n s : N e e d fo r clo s u r e a n d th e allu re o f g ro u p
might also foster individual differences in im p er m ea b ility . Manuscript submitted for publica
tion.
closed- and open-mindedness. For instance,
De Grada, E ., Kruglanski, A. W., Man netti, L., 8c
disagreements between one’s parents dur Pierro, A. (1999). Motivated cognition and group
ing early childhood might give rise to an interaction: Need for closure affects the contents
aversive uncertainty, inducing a stable need and processes of collective negotiations. J o u r n a l o f
for closure. Growing up in unstable physi E x p e r im e n ta l S o c ia l P sy ch olog y , 3 5 , 3 4 6 - 3 6 5 .
352 IV. C O G N I T I V E D I S P O S I T I O N S
De Grada, E., Kruglanski, A. W., Mannetti, L., Pierro, formists at varying degrees of proximity to decision
A., & Webster, D. M. (1996). Un’analisi stutturale deadline and of environmental noise. J o u r n a l o f
comparative delle versioni USA e italiana della scala P erso n a lity a n d S o c ia l P sy ch olog y , 61, 2 1 2 - 2 2 5 .
di “ Bisogno di chiusura cognitive” di Webster and Kruglanski, A. W., Sc Webster, D. M. (1996). M o ti
Kruglanski [A comparative structural analysis of vated closing of the mind: “ Seizing” and “freezing.”
the U.S. and Italian versions of the “Need for Cog P sy c h o lo g ic a l R ev iew , 103, 2 6 3 - 2 8 3 .
nitive Closure” Scale of Webster and Kruglanski|. Kruglanski, A. W., Webster, D. M ., Sc Klem, A. (1993).
T esting, P sic o m etr ia , M e to d o lo g ia , 3 , 5 - 1 8 . Motivated resistance and openness to persuasion
Dijksterhuis, A., van Knippenberg, A., Kruglanski, A. in the presence or absence of prior information.
W., Sc Schaper, C. (1996). Motivated social cogni J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o lo g y , 6 5,
tion: Need for closure effects on memory and judg 861-876.
ment. J o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch olog y , Legierski, J ., Sc Kossowska, M. (2008). E p is tem ic m o
32, 254-270. tiv a tio n , w o r k in g m e m o r y a n d d ia g n o s tic in fo r m a
Festinger, L. (1950). Informal social communication. tio n sea rch . Unpublished manuscript, Jagielonski
P sy c h o lo g ic a l R e v iew , 57, 2 7 1 - 2 8 2 . Uniwersytet, Krakow, Poland.
Ford, T. E., & Kruglanski, A. W. (1995). Effects of Livi, S. (2002). II b is o g n a d i c h iu sc o r a co g n itiv a e la
epistemic motivations on the use of accessible con tra n s m is sio n e d e lle n o r m e n ei p ic c o li g r u p p i [T h e
structs in social judgment. P erso n a lity a n d S o cia l n e e d f o r co g n itiv e clo s u r e a n d n o rm -tra n s m is sio n
P sy c h o lo g y B u lletin , 2 1, 9 5 0 - 9 6 2 . in sm a ll g r o u p s ]. Unpublished doctoral dissertation,
Jamieson, D. W., Sc Z an na, M . P. (1989). Need for University of Rome La Sapienza.
structure in attitude formation and expression. In Maas s, A., Sc Arcuri, L. (1992). T he role of language
A. R. Pratkanis, S. J. Breckler, Sc A. G. Greenwald in the persistence of stereotypes. In G. Semin & K.
(Eds.), A ttitu d e stru ctu re a n d fu n c tio n (pp. 3 8 3 - Fiedler (Eds.), L a n g u a g e, in tera ctio n a n d s o c ia l
4 06 ). Hillsdale, NJ: Erlbaum. c o g n itio n (pp. 1 2 9 - 1 4 3 ). Newbury Park, CA: Sage.
Jones, E. E. (1979). The rocky road from act to disposi Mayseless, O., & Kruglanski, A. W. (1987). What
tion. A m er ic a n P sy ch o lo g ist, 3 4 , 107-117. makes you so sure?: Effects of epistemic motivations
Kossowska, M ., Orehek, E., & Kruglanski, A. W. (in on judgmental confidence. O rg a n iz a tio n a l B e h a v
press). Motivation towards closure and cognitive re io r a n d H u m an D ec is io n P ro cesses, 39, 1 6 2 - 1 8 3 .
sources: An individual differences approach. In A. Mikulincer, M ., Yinon, A., Sc Kabili, D. (1991).
Gruszka, G. Mathews, Sc B. Szymura (Eds.), H a n d Epistemic needs and learned helplessness. E u r o p ea n
b o o k o f in d iv id u a l d iffe r e n c e s in c o g n itio n : A tte n J o u r n a l o f P erson ality , 5, 2 4 9 - 2 5 8 .
tion , m e m o r y a n d ex e c u tiv e c o n tro l. New York: Pierro, A., & Kruglanski, A. W. (200 8). “Seizing and
Springer. freezing” on a significant-person schema: Need for
Kossowska, M ., Van Hiel, A., Chun, W. Y., Sc Krug closure and the transference effect in social judg
lanski, A. W. (2002). The Need for Closure scale: ment. P erson ality a n d S o c ia l P sy ch o lo g y B u lletin ,
Structure, cross-cultural invariance, and compari 34, 1492-1503.
son of mean ratings between European-American Pierro, A., Man netti, L., Converso, D., Garsia, V.,
and East Asian samples. P sy c h o lo g ic a B elg ica , 4 2 , Miglietta, A., Ravenna, M., et al. (1995). Caratter-
267-286. istiche strutturali della versione italiana della scale
Kruglanski, A. W. (1989). L a y e p is te m ic s a n d h u m an di bisogno di chiusura cognitiva (di Webster and
k n o w le d g e : C o g n itiv e a n d m o tiv a tio n a l b a se s. New Kruglanski) [Structural characteristics of the Ital
York: Plenum Press. ian version of the Need for Cognitive Closure Scale
Kruglanski, A. W. (20 04 ). T h e p s y c h o lo g y o f c lo s e d (of Webster and Kruglanski)]. T estin g, P sico m etr ia ,
m in d ed n es s. New York: Psychology Press. M e to d o lo g ia , 2 , 1 2 5 - 1 4 1 .
Kruglanski, A. W., Sc Freund, T. (1983). T he freezing Pierro, A., Mannetti, L., De Grada, E., Livi, S., Sc
and unfreezing of lay-inferences: Effects on impres- Kruglanski, A. W. (2 003). Autocracy bias in infor
sional primacy, ethnic stereotyping, and numerical mal groups under need for closure. P erson ality a n d
anchoring, j o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch o l S o c ia l P sy ch o lo g y B u lletin , 2 9 , 4 0 5 - 4 1 7 .
og y , 19, 4 4 8 - 4 6 8 . Raven, B. H., Schwarzwald, J ., Sc Koslowsky, M .
Kruglanski, A. W., Pierro, A., Higgins, E. T., Sc Ca- (1998). Conceptualizing and measuring a power/in
pozza, D. ( 2 0 07 ). “ On the move” or “staying put”: teraction model of interpersonal influence. J o u r n a l
Locomotion, need for closure and reactions to or o f A p p lie d S o c ia l P sy c h o lo g y , 2 8 , 3 0 7 - 3 3 2 .
ganizational change, jo u r n a l o f A p p lie d S o c ia l P sy Richter, I.., Sc Kruglanski, A. W. (1999). Motivated
c h o lo g y , 37, 1 3 0 5 - 1 3 4 0 . search for common ground: Need for closure effects
Kruglanski, A. W., Shah, J . Y., Fishbach, A., Fried on audience design in interpersonal communica
man, R., Chun, W., Sc Sleeth-Keppler, D. (2002). tion. P erson ality a n d S o c ia l P sy ch o lo g y B u lletin ,
A theory of goals systems. In M . P. Zanna (Ed.), 25(9), 11 01-1114.
A d v a n ces in e x p e r im e n ta l s o c ia l p s y c h o lo g y (Vol. Rokeach, M. (1960). T h e o p e n a n d c lo s e d m in d . New
3 4, pp. 3 3 1 - 3 7 8 ). New York: Academic Press. York: Basic Books.
Kruglanski, A. W., Shah, J . Y., Pierro, A., Sc Mannetti, Ross, L. (1977). The intuitive psychologist and his
L. (200 2). When similarity brings content: Need for shortcomings: Distortions in the attribution pro
closure and the allure of homogeneous and self- cess. In L. Berkowitz (Ed.), A d v a n ce s in e x p e r im e n
resembling groups. J o u r n a l o f P erson ality a n d S o ta l s o c ia l p s y c h o lo g y (Vol. 10, pp. 1 7 3 - 2 2 0 ) . New
c ia l P sy ch olog y , 8 3 , 6 4 8 - 6 6 2 . York: Academic Press.
Kruglanski, A. W., Sc Webster, D. M. (1991). Group Rubini, M ., Sc Kruglanski, A. W. (1997). Brief encoun
members’ reactions to opinion deviates and con ters ending in estrangement: Motivated language
23. N eed for C o g n itiv e C losure 353
use and interpersonal rapport in the question— Webster, D. M . (1993a). C r o u p s u n d er th e in flu en ce:
answer paradigm. J o u r n a l o f P erson ality a n d S o cia l N e e d f o r clo s u r e e ffe c ts o n in fo r m a tio n sh a rin g in
P sy c h o lo g y , 7 2 , 1 0 4 7 - 1 0 6 0 . d ec is io n m a k in g g ro u p s. Unpublished doctoral dis
Shah, J . Y., Kruglanski, A. W., & Thompson, E. P. sertation, University of Maryland.
(19 98). Membership has its (epistemic) rewards: Webster, D. M. (1993b). Motivated augmentation
Need for closure effects on ingroup bias. J o u r n a l o f and reduction of the overattribution bias. J o u r n a l
P erson ality a n d S o c ia l P sy ch olog y , 75, 3 8 3 - 3 9 3 . o f P erson ality a n d S o c ia l P sy ch o lo g y , 65(2), 2 6 1 —
Sorrentino, R. M., &. Short, J . C. (1986). Uncertainty 271.
orientation, motivation and cognition. In R. M. Webster, D. M., & Kruglanski, A. W. (1994). Indi
Sorrentino & E. T. Higgins (Eds.), H a n d b o o k o f vidual differences in need for cognitive closure.
m o tiv a tio n a n d c o g n itio n : Vol. I. F o u n d a tio n s o f J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 67,
s o c ia l b e h a v io r (pp. 1 8 9 - 2 0 6 ). New York: Guilford 1049-1062.
Press. Webster, D. M ., Kruglanski, A. W., Sc Pattison, D.
Stasser, G., &C Stewart, D. (1992). Discovery of hidden A. (1 997). Motivated language use in intergroup
profiles by decision-making groups: Solving a prob contexts: Need for closure effects on the linguistic
lem versus making a judgment. J o u r n a l o f P ers o n a l intergroup bias. J o u r n a l o f P erson ality a n d S o c ia l
ity a n d S o c ia l P sy ch olog y , 63(2), 4 2 6 - 4 3 4 . P sy ch olog y , 7 2, 1 1 2 2 - 1 1 3 1 .
Stasser, G., & Titus, W. (1985). Pooling of unshared Webster, D. M ., Richter, L., & Kruglanski, A. W.
information in group decision making: Biased infor (1996). On leaping to conclusions when feeling
mation sampling during discussion. J o u r n a l o f P er tired: Mental fatigue effects on impressional pri
s o n a lity a n d S o c ia l P sy ch olog y , 4 8 (6 ), 14 6 7 -1 4 7 8 . macy. J o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch olog y ,
Stasser, G., & Titus, W. (1987). Effects of informa 3 2 , 181-1 95 .
tion load and percentage of shared information on Webster-Nelson, D., Klein, C. F., Sc Irvin, J . E. (2003).
the dissemination of unshared information during Motivational antecedents of empathy: Inhibiting ef
group discussion. J o u r n a l o f P erso n a lity a n d S o cia l fects of fatigue. B asic a n d A p p lie d S o c ia l P sy ch o l
P sy ch olog y , 53(1), 8 3 - 9 3 . og y , 2 5 , 3 7 - 5 0 .
CHAPTER 24
Integrative Complexity
P e t e r Su e d f e l d
354
2 4. In teg rativ e C o m p lexity 355
ersch, & M cCaslin, Chapter 21, this vol chapter traces chronologically the major ver
ume) are closely related to each other; and sions of this approach— conceptual systems
other constructs, such as the theories of un theory and the theories of conceptual com
certainty orientation (Sorrentino, Roney, & plexity (subsuming interactive complexity)
Hanna, 1992) and telic dominance (Apter, and integrative complexity— and summa
1989), also have components that are very rizes a sampling of the research inspired by
similar to both need for closure and need for each.
cognition. The third construct— conceptual
complexity and its major offshoots— is the
focus of this chapter. Conceptual Systems T heory
The various conceptual and empirical re
Basic Concepts
lationships among formulations of cognitive
complexity raise the question of how many Conceptual complexity theory was formu
such traits really exist, what they are, and lated as a systems theory of personality de
the extent to which these theories are over velopment related to childrearing strategies
lapping or redundant. It is certainly the case (Harvey et al., 1961). The model was in
that their psychometric measures are corre spired by George Kelly’s (1955) psychology
lated, but no overarching correlation matrix of personal constructs to posit cognitive dif
incorporating all of these variables has been ferentiation as one of the basic components
published. For the sake of clarity, and to re of trait-like differences in thinking. Differ
duce clutter in the field (or at least to make entiation was defined as the perception of
sense of the clutter), such an analysis would clearly articulated parts within a situation,
be very valuable. whereas integration involves relating these
Another issue, which lies at the heart of parts to each other and to previously estab
this chapter, is the degree to which these lished constructs. In broad outline, differ
theories actually describe individual differ entiation and integration have remained the
ences, including cognitive processes, that hallmark variables of this school of thought
underlie behavior, imparting a relatively throughout the subsequent years and give
high level of stability that characterizes the every indication of remaining so. The inclu
individual’s responses across time, environ sion of integration also separates this theory
mental conditions, specific problems or is and its successors from most other cognitive
sues, and other dynamic variables. The ap personality models (e.g., Bieri, 1955; Her
proach with which this chapter is primarily mann, 1980).
concerned, conceptual/integrative complex According to systems theory, differentia
ity theory, recognized from the beginning tion and integration develop or fail to devel
that “concepts are jointly determined by the op differently in a series of personality types
totality of external (situational) and internal through four stages of conceptual develop
(dispositional) factors at the given time op ment. Progress from one stage to the next is
erating in mutual interdependence” (Harvey dependent on how family rules were gener
et al., 1961, p. 15). Note that the existence ated and applied. Briefly, the predisposition
of a trait-like (“dispositional”) characteristic (or ability; the distinction is not totally clear
is assumed, and this assumption underlies in the theory) for either concrete or abstract
most of the early research. Later emphasis thinking is based on how family rules are es
shifted to the consideration of how dynamic tablished and the reliability or unreliability
variables affected current (i.e., state) cogni of reward and punishment.
tion, the individual’s underlying trait predis If rules are laid down by the parents and
position being only inferred. consistently lead to reward for compliance
However, both early and late in the history and punishment for transgression (reliable
of this research tradition, researchers recog unilateral training), the child learns to trust
nized that although the relative importance and obey authority (System I conceptual
of trait and state factors varies depending on structure). If the outcomes of rule obser
a host of factors, the final cognitive process vance or violation are not consistent (unilat
is always the result of an interaction between eral unreliable training), counterdependence
these two large categories of influences. This and rebelliousness (System II) follow. Sys
356 IV. C O G N IT IV E D IS P O S IT IO N S
tems I and II are characterized as concrete avoidance of ambiguity, and the authoritar
modes of thought. If rules are developed ian personality, all of which have factors in
through interaction between the parents common with the various systems (and with
and the child, the result is abstract thinking. cognitive personality theories in general).
If, when the child breaks a rule, the parent Research specifically aimed at the systems
steps in to shield the child from adverse con model showed that: (1) people functioning
sequences (interdependent protective train at a System I level made more extreme judg
ing), the individual comes to expect help and ments about others than those at Systems II,
support from others (System III). However, III, and IV; (2) System II is associated with
if the parents allow natural consequences to heightened sensitivity to control by others
occur— except when these would be danger and with disengagement from feedback,
ous or seriously damaging— the child learns commitment, and responsibility; (3) System
to search for and process relevant informa III individuals are especially open to other
tion before making decisions (System IV). people and their reactions to oneself; and (4)
“Arrestation” at a particular stage can occur System IV functioning emphasizes the per
if childrearing methods prevent differentia son’s own standards and autonomy, as well
tions that would lead to more abstract inte as a high level of sensitization to informa
gration. tion.
In a stock-market simulation, concrete-
level groups minimized diversity, both
M easurem ent and Research
from the environment and within the team;
Harvey and colleagues (1961) emphasized sought less information; and were less ac
multiple measurement methods to identify tive, as well as less cohesive (Tuckman,
the conceptual systems level that particular 1964). Other studies looked at social percep
individuals attained. These included what tion, attitude change in response to persua
would normally be considered experimen sive messages, and behavioral rigidity (e.g.,
tal manipulations (such as criticizing par response modification after critical feedback
ticipants and interpreting their reactions and generalization-extinction curves) as a
to negative comments, including them in a function of system level. Applied researchers
conformity experiment, and so on), which have used measures of system-level function
seems a circular method. Conceptual level ing in studies of trainer-trainee and teacher-
is inferred from the participant’s behavior, student relationships (e.g., Hunt, 1966;
which is attributed to the characteristics of Hunt &c Joyce, 1967), and Tuckman’s (1965)
thinking at that particular level. Theoreti theory of group development proposed four
cal constructs that were thought to overlap stages that are essentially the same as H ar
with conceptual systems could also provide vey et al.’s four systems.
the basis of what we might call measure In one experiment, Harvey (1963) found
ment by analogue. That is, measures of au that System IV participants were able to
thoritarianism, dogmatism, and rigidity tap construct and present counterattitudinal
some of the characteristics of System I; a speeches better than the other three groups,
Machiavellianism scale and some responses regardless of whether or not they expected
on a measure of field independence indi the speeches to be heard by a committee that
cate System II; and so on. There were also had the power to make decisions concerning
paper-and-pencil tests designed to measure the position espoused in the speech. This
systems-level functioning, including a Sen finding was interpreted as reflecting high
tence Completion Test that was the forerun er tolerance for cognitive dissonance and
ner of the later Paragraph Completion Test, greater cognitive flexibility at this level of
described later. structure. System II participants performed
Much of the research mustered in the the worst when they expected to speak pub
Harvey and colleagues (1961) book to sup licly but not otherwise, presumably because
port systems theory had been conducted in of their distrust of authority and unwilling
different theoretical contexts. For exam ness to expose their products to the com
ple, Harvey and colleagues cited research mittee. System III participants performed
on field independence, cognitive rigidity, the worst in the private condition, but better
24. In teg rativ e C o m p lexity 357
than Systems I and II in the public condi them are developed. At the highest level, one
tion, perhaps because they trusted authority finds alternative combinations, general laws,
and had experienced positive reinforcement more diversity, self-generated rules, and flex
in the past in their dealings with people in ibility.
authority.
Measurement
Conceptual C om plexity T h eo ry 2 Although several methods for assessing con
ceptual complexity were described by Schro
Basic Concepts
der et al. and other researchers (see Streufert
Although the systems theory generated a & Streufert, 1978, for a comprehensive re
respectable volume of research, some of the view), the most frequently used instrument
scientists who developed the theory and the was, and remains, the Paragraph Comple
studies soon moved in a different direction. tion Test (PCT). The PCT is a semiprojective
In conceptual complexity theory (Schroder measure in which the individual is presented
et al., 1967), the developmental aspect of with a series of sentence stems consisting of
the original theory was dropped, as were the one or a few words and is asked to complete
stages. More important, the modified theory a sentence starting with the stem and then
primarily concerned the structure of thought to continue writing on the same topic until
rather than its content. time is up (usually 1 -2 minutes), when they
Because this version of the theory viewed go to the next stem.
complexity as a fully structural rather than Two stems are used to represent each of
a content variable, the complexity score was three major areas about which people com
defined by levels of differentiation and inte monly need to make decisions— uncertainty,
gration rather than by the attitude or opin relations to authority, and social rejection.
ion expressed. For example, the beliefs that The specific word or phrase that begins each
rules should always be followed and that item can vary depending on the nature of the
rules are made to be broken are opposite in participant sample. For example, with un
content but equivalent in structure: Neither dergraduate research volunteers, one of the
belief shows any sign that the speaker rec authority stems is “Parents. . . . ” With older
ognizes nuances, contingencies, or different participants, this would not be appropri
arguments (no differentiation), and in the ate, and the researcher would select a stem
absence of differentiation, integration is not that would better represent authority to that
possible. Both statements would be scored group. The completed paragraphs are scored
at the lowest level of conceptual complexity, on a 1 -7 scale in ascending order of com
in contrast with conceptual systems scoring plexity, and the final score for the individual
(Harvey et al., 1961), in which the first state is the mean of the six paragraph scores. As
ment would be scored as reflecting System I the scoring system developed, scores of 1
thinking and the second, System II. Although and 3 came to represent the concrete levels
they thought of complexity as a dimension, of undifferentiated and differentiated but not
Schroder and colleagues (1967) still referred integrated conceptual structures, whereas 5
to four levels of conceptual complexity. and 7 marked abstract, integrated thinking.
However, they acknowledged that these lev Scores of 2 , 4, and 6 were used to mark the
els are actually nodal points along a contin implicit or implied, but not clearly stated,
uum of any number of possible levels. The emergence of the next higher level. A highly
lowest (“concrete”) level is characterized by detailed scoring manual has been developed
compartmentalized, rigid, and absolutistic that is appropriate for scoring both concep
cognitive rules. At a moderately low level, tual (trait) and integrative (state) complex
the individual generates alternative ways of ity (Baker-Brown et al., 1992). In the latter
looking at concepts and recognizes some version, the terms concrete and abstract are
areas of autonomous choice. At the moder dropped, as are verbal labels attached to
ately high, more “abstract,” level, more di specific levels of complexity. Scorers can be
mensions are perceived, and combinatorial trained and qualified either in face-to-face
rules for organizing (matching, comparing) workshops or online.
358 IV. C O G N IT IV E D IS P O S IT IO N S
the problem. Thus, many European diplo rule, more than one qualified scorer works
mats knew in the summer of 1914 that their with the dataset, and reliability between the
world was spiraling toward chaos but could scorers is always calculated.
think of no way to avert disaster, even though
only a few years earlier, they had been able
Research
to generate a clever peaceful solution to an
impending showdown between France and The first study in this aspect of complex
Germany over spheres of influence in North ity (Suedfeld & Rank, 1976) looked at the
Africa. reasons that some leaders of victorious revo
As the example illustrates, another major lutions maintain their eminent positions in
departure from previous aspects of com the postrevolution government, whereas
plexity theory is in the sources of data used. others lose their positions, as well as their
Integrative complexity research focuses on freedom or their lives in many cases. As can
archival materials from biography and au be seen, this is an individual difference with
tobiography, history, the media, and other very serious real-life (as opposed to labora
published or recorded documentation. Stud tory) consequences for the person. It turned
ies have concentrated on how trait (concep out that a low level of complexity predicted
tual) complexity interacts with endogenous long-term success during the armed strug
and environmental factors to determine the gle, with a rise to a significantly higher level
complexity of thinking in the specific situ of complexity when the revolutionary move
ation, as well as on identifying factors that ment became the governing party. Leaders
lead to generally predictable changes in state who failed to show this rise fell by the way
complexity across levels of trait complexity. side after victory (e.g., Trotsky, Guevara),
Integrative complexity researchers usually as did those whose initially high level of
make no special effort to assess trait com complexity made them mistrusted by their
plexity. (Indeed, this would be impossible comrades as being insufficiently committed
in most cases, given their usual sources of to the cause (e.g., Alexander Hamilton), a
data.) Instead, their interest is in the level at common reaction to people who are high in
which the person is operating at some, usu trait complexity.
ally important, time, such as national leaders Later research showed that this pattern is
during a political campaign or international not unique to revolutionaries. In democrat
confrontation, generals before and during ic elections in several countries, campaign
battle, or ordinary people facing life crises. speeches have been found to be generally
lower in complexity than postvictory speech
es. Furthermore, individual differences are
M easurem ent
important. Presidents of the United States
The measurement of integrative complexity who failed to show a substantial increase in
uses the same 1 -7 scale and the same scor complexity from before to after their elec
ing manual as the Paragraph Completion tion were among the least highly regarded
Test. However, the material to be scored, by professional political scientists and histo
although it includes the PCT, is mostly rians (Suedfeld, 1994; Tetlock, 1981).
drawn from archival collections. Data may Integrative complexity while in high office
be taken from books, letters, diaries, media may be related to continued success. Andrei
interviews, speeches to legislatures or the Gromyko, for example, managed to retain
public, memoranda, military orders, audio eminent positions in Soviet foreign policy
or videotapes, and other sources that reflect from the start of his career under Stalin in
a person’s level of thinking. To avoid biased the 1930s through all the vicissitudes of
scoring, the material is collected by a mem history through the reign of Gorbachev in
ber of the research staff who is unaware of the 1980s. He was also the only one among
the hypotheses, who selects paragraphs (the his contemporary statesmen in the United
unit of analysis) at random if sampling from States and the U .S.S.R. whose complexity
a larger population of paragraphs is neces did not diminish— and, in fact, increased—
sary, and who, as far as possible, removes during the domestic and international crises
all identifying information from the materi of that half-century. A similar resistance to
als before passing them to the scorers. As a disruptive stress was shown by several other
2 4. In teg rativ e C o m p lexity 361
long-serving leaders, including the Duke of promise, more complex negotiations are as
Wellington and Canadian Prime Minister sociated with more progress (Liht, Suedfeld,
Lester Pearson (Wallace & Suedfeld, 1988). & Krawczyk, 2005). However, extensive
Whether resistance to disruptive stress is the research on negotiations, as opposed to con
same as high trait complexity, correlated frontations, remains to be done.
with it, or orthogonal to it remains to be in In contrast to Chamberlain, Winston
vestigated. Churchill dogmatically maintained that
High trait complexity, while perhaps en Hitler and the Nazis must be dealt with by
hancing the person’s likelihood of success arms buildups and stern displays of force.
under some difficult conditions, is no pana Churchill claimed that appeasement through
cea under constant and worsening stress. flexible negotiation encouraged further ag
General Robert E. Lee’s pre-Civil War writ gression. Most historians today agree, with
ings showed consistently high complexity, the 20/20 vision that hindsight affords, that
from which high trait complexity may be Churchill was right. Throughout the 1930s,
inferred. In the first years of the war, com Churchill maintained a simple stance toward
manding against Union generals whose state Hitler’s Germany, whereas Chamberlain
complexity was lower than his, he repeatedly continued to discuss the problem at a com
gained victory or at least managed to avoid plex level until shortly prior to the outbreak
severe defeat by superior numbers. However, of war in 1939 (Tetlock & Tyler, 1996).
toward the end of the war, after years of at A drop in complexity just before war
trition in manpower and resources, and for breaks out characterizes national leaders and
the first time facing an enemy commander their subordinates across many international
of equivalent or higher complexity (U. S. confrontations. International crises that end
Grant), Lee’s series of unlikely successes in war are consistently preceded by such a
came to an end. Interestingly, from his sur drop, whereas negotiated solutions come
render at Appomattox to the end of his life, at the end of exchanges that show main
he regained his earlier high complexity level tained or increased complexity. In addition,
(Suedfeld, Corteen, & M cCorm ick, 1986). surprise strategic attacks are forecast by a
Lee is one of the best examples of the in drop in the complexity of communications
terplay between conceptual complexity and from the eventual attacker, but not from the
environmental conditions, an interplay that target. However, the target’s complexity be
determines the level of integrative complex comes as low as the attacker’s immediately
ity- after the attack occurs. This general pattern
As implied earlier, high complexity is not has been reliable from the Russo-Japanese
necessarily a key to success. Under some War through the Iraqi invasion of Kuwait,
conditions, it is in fact counterproductive. during regional wars, world wars, and per
This is probably the case when the situation sistent rivalries (Suedfeld & Bluck, 1988).
requires rapid, clear-cut decisions, such as But within the pattern, some leaders seem to
when the country is under attack or when have chronically higher (or lower) levels of
one is up against an implacable opponent. complexity than others; thus the complex
During the 1938 Munich conference, for ity of a nation’s policies may shift when the
example, Prime Minister Chamberlain’s leadership of that nation changes. For exam
complexity was almost half again as high ple, Mikhail Gorbachev displayed decidedly
as Adolf Hitler’s (Suedfeld, 1988), but the higher complexity scores than his Soviet pre
latter’s intransigence prevailed because of decessors (Tetlock & Boettger, 1989). There
Chamberlain’s willingness to accept succes may also be individual differences in main
sive compromises that redounded to Germa taining complexity under stress and in the
ny’s benefit. Similar patterns may character ability to recognize and act on the need to
ize some negotiations in today’s international shift complexity levels, as mentioned earlier.
system. For example, Tibon (2000) reported Such individual differences within leadership
that strong Israeli supporters of peace ne groups and their influence on group decision
gotiations between Israel and Palestinians making need further research.
were higher in complexity than those who In general, aggressive or otherwise un
were less supportive. It does seem that with compromising strategies, even if they do
parties who are willing to accept some com not result in war, are accompanied by low
362 IV. C O G N IT IV E D IS P O S IT IO N S
complexity among leaders (e.g., Conway, complexity on the one hand and both politi
Suedfeld, & Tetlock, 2001). There are also cal extremism and political interest on the
spread-of-effect phenomena in which poli other. They attribute the discrepancy from
cymakers of nations that are politically or previous findings to differences in the par
geographically remote from the conflict ticipant samples used, but the issue clearly
show less disruption of complexity than calls for further research.
those directly involved or close to the line Although much of integrative complex
of fire. Similarly, although the head of state ity research has focused on political topics,
shows more pronounced effects than subor there have been other issues of interest. For
dinates within the leadership group, war or example, Woike (1994) used the concepts
imminent war produces reduced complexity of differentiation and integration to test the
in a wide variety of elites, even those who theory that these were linked to agency and
have no role in national policy or wartime communion as general orientations, with men
strategy, such as novelists, scientists, edito tending to emphasize the former and women
rial writers, eminent psychologists, and the the latter. Using a revised version of the inte
like (see Suedfeld, 2003). grative complexity measurement technique,
Complexity is also involved in political she reported that in descriptions of a posi
ideology. The old debate about whether au tive or negative life experience, the predicted
thoritarianism in its cognitive incarnation gender difference in the percentages of dif
(rigidity, closed-mindedness, rule-based de ferentiation versus integration was found re
cisions, avoidance of uncertainty and am gardless of whether the event described was
biguity, all-or-nothing thinking, etc.) exists pleasant or unpleasant, but that no such dif
on both the left and the right of the politi ference was found in a neutral condition. A
cal spectrum has been inconclusive, despite second experiment found congruent results
attempts to resolve it by fiat or redefinition when participants high in either intimacy or
(e.g., Altemeyer, 1988; Duckitt, Chapter 20, power motivation were primed to watch for
this volume). However, studies of concep either leadership or friendly cooperation in a
tual complexity have found lower levels of tape of a job interview. Participants high in
complexity among members of ideologically power motivation used differentiation more,
based parties, left or right, than among more and those high in intimacy motivation used
pragmatic ones (Suedfeld, Bluck, Loewen, & integration, regardless of their gender, and
Elkins, 1994). the motivation-priming combination pre
Tetlock (1986) hypothesized that, in dicted differentiation and integration more
Western democracies, liberalism has a spe strongly than gender did.
cial need to reconcile two basic, important, Gruenfeld, Thomas-Hunt, and Kim (1998)
but often mutually contradictory values— tested the finding that members of a majority
freedom and equality. This need to accom generally show higher integrative complex
modate value conflict leads to higher levels ity than a competing minority. Because there
of complexity among center-left parties than has been some controversy as to whether this
among those further from the center in ei difference reflects the structure of thought
ther direction, which hold one or the other or only impression management, Gruenfeld
value as central and are therefore willing to and colleagues used an experimental para
compromise on the other. The model is not digm that manipulated a private versus pub
restricted to the political arena but, rather, lic communication condition. Their results
is relevant to any situation in which value showed that the m ajority-minority differ
conflict is a component of ideology. Support ence exists under both conditions, confirm
ing evidence has been found among politi ing inferences from archival research that
cal groups in the U .S.S.R ., the United States, the complexity score is a measure of thought
Britain, and Canada; in the United States and not merely of self-presentation to an au
before the Civil War; in groups involved in dience.
a Canadian controversy about sustainable de Vries and Walker (1987) had student
forest management; and in students writing participants take the PCT and write an essay
value-related essays. Van Hiel and Merviel- defending their own attitude concerning
de (2003), however, reported significant capital punishment. They found that the es
positive correlations between integrative says were higher in complexity than the PCT
2 4. In teg rativ e C o m p lexity 363
responses, possibly because the participants Their self-ratings showed broad interests,
were more interested in the former. This not playing hunches, enjoying difficult prob
would have been predicted by the cognitive- lems (need for cognition?), and not being
manager model (Suedfeld, 1992). The au motivated by money. Perhaps surprisingly,
thors also reported an interesting curvilin complexity scores in the two domains of re
ear function, supporting the value conflict search and teaching had a zero correlation;
model, with those who rated themselves as although domain specificity has been re
neutral on the subject writing more complex ported before, the frequently assumed close
essays than did both strong opponents and relationship between research and teaching
strong proponents of capital punishment. would imply otherwise.
In another study, de Vries, Blando, and One unusual study (Suedfeld, de Vries,
Walker (1995) found that more pleasant Bluck, Wallbaum, & Schmidt, 1996) tested
than unpleasant events were mentioned in whether there exists an intuitive common-
life-review interviews. However, complexity sense understanding of complexity that
was higher when the individual was recalling parallels the everyday, generally accepted
unpleasant, undesirable, intense events or understanding of the concept of intelli
events that had been unexpected, for which gence. Undergraduates (who had not taken
the interviewee was not responsible and to a course that covered cognitive complexity
which the interviewee had not adjusted. theory) completed the PCT. They then com
Studies of scientists have been few, but pared their own responses with two sets of
interesting. Presidents of the American Psy described solutions. One of these sets con
chological Association give less complex sisted of four “prototype paragraphs” writ
presidential addresses during times of na ten by expert scorers to represent general
tional crisis; the presidents who were rated information processing at complexity levels
as more eminent by other psychologists show of 1, 3, 5, and 7 (see the earlier section on
higher complexity than their less eminent the measurement of complexity); the second
counterparts; and the speeches of presidents set comprised actual PCT responses to the
whose areas and scientific approaches are same stems, taken from previous studies and
more in the area of social rather than bio scored 1, 3, 5, or 7. Last, the participants
logical science are more complex (Suedfeld, were asked how a list of 17 factors would
1985). Feist (1994) expanded this approach have affected their responses. The factors in
to study characteristics of professors of cluded both endogenous and environmental
physics, chemistry, and biology, who agreed influences that previous research had shown
to be interviewed about their research and to have reliable effects on integrative com
teaching. Both Suedfeld (1985) and Feist plexity, such as accountability, value con
(1994) showed that their participants were flict, and distraction.
substantially higher in complexity over The results were reassuring if one be
all (means above 3.5) than most research lieves that ecological validity includes some
samples. Feist did find differences across the congruity with what people other than so
three disciplines, but only the overall results cial scientists sense to be “real” qualities of
are mentioned here. Scientists who thought human personality and behavior. Although
in complex ways about their research were the participants were not good at estimating
cited more frequently, were rated by peers the complexity of their own PCT responses
as more eminent, and were rated by observ (i.e., the general prototypes they chose as
ers as exploitative, fastidious, deceitful, ma most similar to their own were generally not
nipulative, and not socially poised, giving, at the same level of complexity), they were
or sensitive to others. They rated themselves quite accurate in matching the prototype
as having high standards, rapid tempo, and paragraphs with actual PCT completions
a narcissistic working style (cf. the findings from previous studies. In fact, some of them
concerning graduate students in business, scored at r - .85 or higher with the experts’
cited previously; Tetlock et al., 1993). Feist scoring, which is the threshold for qualifi
also found that scientists who thought in cation as an independent scorer! Their pick
complex ways about education and teaching of the “most com fortable” response was
were perceived by others as warm, charm reliably more complex than their own para
ing, gregarious, and not condescending. graphs; and they were very accurate in es
364 IV. C O G N IT IV E D IS P O S IT IO N S
R. V. Wagner, 8c D. Winter (Eds.), P e a c e , c o n flic t, Scott, W. A., Osgood, D. W., 8c Peterson, C. (1979).
a n d v io le n c e : P ea c e p s y c h o lo g y fo r th e 2 1 st c e n C o g n itiv e stru ctu re: T h e o r y a n d m ea su re m en t o f
tury (pp. 6 6 - 7 5 ) . Englewood Cliffs, NJ: Prentice- in d iv id u a l d iffe r e n c e s . New York: Wiley.
Hall. Sorrentino, R. M., Roney, C. J . R., 8c Hanna, S. E.
Coren, S., 8c Suedfeld, P. (1995). Personality correlates (1992). Uncertainty orientation. In C. P. Smith (Hd.),
of conceptual complexity. J o u r n a l o f S o c ia l B e h a v M otiv atio n a n d p e r s o n a lity : H a n d b o o k o f th em a tic
io r a n d P erson ality , 10, 2 2 9 - 2 4 2 . c o n te n t an a ly sis (pp. 4 1 9 - 4 2 7 ) . New York: C a m
de Vries, B., Blando, J . , & Walker, L. J. (1995). The re bridge University Press.
view of life’s events: Analyses of content and struc Streufert, S., 8c Streufert, S. C. (1978). B e h a v io r in the
ture. In B. Haight 8c J. Webster (Eds.), T h e a r t a n d c o m p le x en v ir o n m en t. Washington, DC: Winston.
s c ie n c e o f rem in iscin g : T h eo ry , research , m e th o d s , Streufert, S., Streufert, S. C., 8c Castore, C. H. (1968).
a n d a p p lic a t io n s (pp. 1 2 3 - 1 3 7 ) . Washington, DC: Leadership in negotiations and the complexity of
Taylor 8c Francis. conceptual structure. J o u r n a l o f A p p lie d P sy ch o l
de Vries, B., 8c Walker, L. J . (1987). Conceptual/in ogy, 5 2 , 2 1 8 - 2 2 3 .
tegrative complexity and attitudes toward capital Streufert, S., 8c Swezey, R. W. (1986). C o m p le x ity ,
punishment. P erson ality a n d S o c ia l P sy ch olog y m a n a g ers, a n d o rg a n iz a tio n s . New York: Academic
B u lletin , 1 3 , 4 4 8 - 4 5 7 . Press.
Feist, G. J. (1994). Personality and working style Suedfeld, P. (1985). American Psychological Associa
predictors of integrative complexity: A study of tion Presidential addresses: The relation of integra
scientists’ thinking about research and teaching. tive complexity to historical, professional, and
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 67, personal factors. J o u r n a l o f P erson ality a n d S o c ia l
474-484. P sy ch olog y , 4 9 , 1 6 4 3 - 1 6 5 1 .
Frauenfelder, K. J. (1974). Integrative complexity and Suedfeld, P. (1988). Are simple decisions always worse?
extreme responses. P s y c h o lo g ic a l R e p o rts , 3 4 , S ociety , 2 5 , 2 5 - 2 7 .
7 70. Suedfeld, P. (1992). Cognitive managers and their crit
Gruenfeld, D. H., Thomas-Hun t, M . C., & Kim, P. H. ics. P o litica l P sy ch olog y , 13, 4 3 5 - 4 5 3 .
(1998). Cognitive flexibility, communication strat Suedfeld, P. (1994). President Clinton’s policy dilem
egy, and integrative complexity in groups: Public mas: A cognitive analysis. P o litic a l P sy ch o lo g y , IS ,
versus private reactions to majority and minority 337-349.
status. J o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch olog y , Suedfeld, P. (2003). Integrative complexity in political
34, 202-226. contexts. In J. Kawata 8c Y. Araki (Eds.), H a n d
Harris, R. M., 8c Highlen, P. S. (1982). Cognitive co m b o o k o f p o lit ic a l p s y c h o lo g y (pp. 5 2 - 6 2 ) . Tokyo:
plexity and susceptibility to learned helplessness. Hokuju Shuppan.
S o c ia l B e h a v io r a n d P erson ality , 10(2), 1 8 3 - 1 8 8 . Suedfeld, P., 8c Bluck, S. (1988). Changes in integra
Harvey, O. J. (1963). C o g n itiv e d ete rm in a n ts o f ro le tive complexity prior to surprise attacks. J o u r n a l o f
p la y in g (Technical Report No. 3, Contract No. C o n flic t R e so lu tio n , 3 2 , 6 2 6 - 6 3 5 .
1147(07)). Alexandria, VA: Department of De Suedfeld, P., Bluck, S., Loewen, L., & Elkins, D. J.
fense. (1994). Sociopolitical values and integrative co m
Harvey, O. J., Hunt, D., 8c Schroder, H. M. (1961). plexity of members of student political groups.
C o n c e p tu a l sy stem s a n d p e r s o n a lity o r g a n iz a tio n . C a n a d ia n J o u r n a l o f B eh a v io u r a l S cien ce, 2 6,
New York: Wiley. 1 21-141.
Hermann, M . G. (1980). Explaining foreign policy be Suedfeld, P., 8c Coren, S. (1992). Cognitive correlates
havior using the personal characteristics of leaders. of conceptual complexity. P erson ality a n d In d iv id
I n te r n a tio n a l S tu d ies Q u arterly , 2 4 , 7 - 4 6 . u a l D iffe r e n c e s , 13, 1 1 9 3 - 1 1 9 9 .
Hunt, D. E. (1966). A model for analyzing the train Suedfeld, P., Corteen, R. S., 8c M cCorm ick, C. (1986).
ing of training agents. M e r r ill-P a lm e r Q u a rterly o f The role of integrative complexity in military lead
B e h a v io r a n d D e v e lo p m en t, 12, 137 -15 6 . ership: Robert E. Lee and his opponents. J o u r n a l o f
Hunt, D. E., 8c Joyce, B. R. (1967). Teacher trainee A p p lie d S o c ia l P sy ch olog y , 16, 4 9 8 - 5 0 7 .
personality and initial teaching style. A m erica n Suedfeld, P., de Vries, B., Bluck, S., Wallbaum, A. B.
E d u c a tio n a l R e se a rc h J o u r n a l, 4, 2 5 3 - 2 5 9 . C., 8c Schmidt, P. W. (1996). Intuitive perceptions
Kelly, G. A. (1955). T h e p s y c h o lo g y o f p e r s o n a l c o n of decision-making strategy: Naive assessors’ con
stru cts: Vol. I. A th e o r y o f p er so n a lity . New York: cepts of integrative complexity. In te r n a tio n a l J o u r
Norton. n a l o f P sy ch olog y , 3 1 , 1 7 7 - 1 9 0 .
Kruglanski, A. W., 8c Webster, D. M . (1996). M o ti Suedfeld, P., 8c R ank, D. A. (1976). Revolutionary
vated closing of the mind: “Seizing” and “ freezing.” leaders: Long-term success as a function of concep
P sy c h o lo g ic a l R ev iew , 103, 2 6 3 - 2 6 8 . tual complexity. J o u r n a l o f P erson ality a n d S o c ia l
Liht, J ., Suedfeld, P., 8c Krawczyk, A. (2005). Integra P sy ch olog y , 3 4 , 169-178.
tive complexity in face-to-face negotiations between Tadmore, C. T., 8c Tetlock, P. E. (200 6). Biculturalism:
the Chiapas guerrillas and the Mex ic an government. A model of the effects of second-culture exposure on
P o litic a l P sy ch olog y , 2 6 (4), 5 4 3 - 5 5 2 . acculturation and integrative complexity. J o u r n a l o f
Schroder, H. M., Driver, M. J ., 8c Streufert, S. (1967). C ro ss-C u ltu ra l P sy ch olog y , 37(2), 1 7 3 - 1 9 0 .
H u m an in fo r m a tio n p ro cessin g . New York: Holt, Tetlock, P. E. (1981). Pre- to post-election shifts in
Rinehart 8c Winston. presidential rhetoric: Impression management or
Schroder, H. M ., 8c Suedfeld, P. (Eds.). (1971). P e r s o n cognitive adjustment? J o u r n a l o f P erson ality a n d
a lity th e o r y a n d in fo r m a tio n p ro cessin g . New York: S o c ia l P sy ch olog y , 4 1 , 2 0 7 - 2 1 2 .
Ronald Press. Tetlock, P. F.. (1986). A value pluralism model of ideo
366 IV . C O G N IT IV E D IS P O S IT IO N S
M o t iv a t io n a l D is p o s it io n s
I
CHAPTER 25
Conscientiousness
Bren t W. R o berts
J o sh u a J. J a c k so n
J e n n i f e r V. F a y a r d
Gra n t E dmonds
J e n n a M e in t s
369
370 V. M O T IV A T IO N A L D IS P O S IT IO N S
Finally, Goldberg (1993) used the term Big The 36 measures of conscientiousness were
Five for a reason. The Big Five are big. Not best subsumed by six factors: Impulse Con
big in the sense that they are important, but trol, Conventionality, Reliability, Industri
big in the sense that each of the Big Five is ousness, Order, and Virtue. Interestingly,
best considered a broad domain of traits, not there is striking convergence across the
a unitary construct. This point seems to be lexical and questionnaire studies: Industri
increasingly lost on the current generation ousness, Reliability, Order, Impulse Con
of personality inventory consumers, as the trol, and Conventionality replicated across
preference appears to be to use short mea these disparate samples and assessment
sures under the assumption that measuring techniques, suggesting that, at a minimum,
a single dimension of conscientiousness, or these five factors make up the underlying
any of the remaining Big Five, is a sufficient structure of conscientiousness. Two aspects
representation of the domain. This is like ar of this five-facet interpretation of conscien
guing that oranges, apples, and bananas are tiousness are worth noting. First, no existing
interchangeable because they are all fruit. personality inventory includes all five, which
Conscientiousness is clearly not unidimen renders any existing system of assessing
sional and consists of several relatively dis conscientiousness inadequate. M ost inven
tinct facets that, like different fruit, are not tories fail to incorporate the conventional
identical. ity facet, which is often mistakenly identi
W hat, then, is the composition of the fied as an aspect of low openness. Despite
family of traits within the conscientiousness this preconception, across these two studies
domain? Several studies have focused on conventionality was more strongly related
identifying the lower order structure of con to conscientiousness than to openness. Fur
scientiousness using two approaches. One thermore, in both studies the remaining fac
route to identifying the structure of conscien ets of conscientiousness showed good levels
tiousness is to examine lexically derived trait of convergent and discriminant validity with
adjectives, as was done to develop the Big the remaining Big Five, with the exception
Five (e.g., Goldberg, 1993). A second route of the reliability facet. The latter is almost
to identifying the underlying domain of con equivalently correlated with conscientious
scientiousness is an examination of the fac ness and agreeableness.
tor structure of personality inventories that This more differentiated model of consci
measure conscientiousness-related traits. entiousness provides the starting point for
Across both approaches research has arrived documenting the relation of conscientious
at some semblance of consensus on the repli ness to social behaviors. As is seen in the
cable facets of conscientiousness. In a lexical following section, it allows us to organize
study (Roberts, Bogg, Walton, Chernyshen previous research literature in order to dis
ko, & Stark, 20 0 4 ), five components found cover which aspect of conscientiousness is
in previous lexical research on the lower most important for a variety of phenomena.
order structure of conscientiousness were In this chapter we review the association be
identified: industriousness (tenaciousness vs. tween conscientiousness and four domains:
laziness), reliability (dependability vs. unre behavior (social and otherwise), emotion,
liability), orderliness (organization vs. slop motivation, and social cognition. Given the
piness), impulse control (cautiousness vs. fact that conscientiousness is a personality
carelessness), and decisiveness (decisiveness trait that reflects relatively consistent pat
vs. indecisiveness). Unlike previous research, terns of thoughts, feelings, and behaviors, it
two additional, interpretable facets were should come as no surprise that it is associat
found: formalness and conventionality. Both ed with behaviors (health behavior), feelings
of these dimensions appeared to represent (guilt and shame), and thoughts (motivations
blends of conscientiousness, with high and and social cognition). It is the association be
low openness to experience, respectively. tween conscientiousness and these outcomes
In a second study of scales drawn from that may help to explain why conscientious
personality inventories, the factor structure ness, in turn, predicts so many significant life
of 36 different scales assessing aspects of outcomes, such as health, longevity, occupa
conscientiousness was examined (Roberts, tional success, and marital stability (Roberts,
Chernyshenko, Stark, & Goldberg, 2005). Kuncel, Shiner, Caspi, & Goldberg, 2007).
25. C onscien tiou sness 371
problems, such as schoolwork, rather than tribute to partners’ feeling dissatisfied with
neglecting and avoiding these situations their relationship. In contrast, individuals
(O’Brien Sc DeLongis, 1996). higher in conscientiousness tend to be better
Similarly, success in the labor force is re at managing conflicts that inevitably arise in
lated to behaviors that are associated with relationships (Buss, 1992; Finkel Sc Camp
conscientiousness. For example, consci bell, 2 0 0 1 ; Jensen-Campbell & Graziano,
entiousness is one of the best predictors of 2001). Moreover, conscientious individu
job performance (Barrick & M ount, 1991). als might actually provoke fewer disagree
Conscientiousness also predicts a number ments and have fewer conflicts because their
of behaviors related to job performance, behaviors evoke less criticism, as they are
such as absenteeism (Ones, Viswesvaran, Sc generally controlled, organized, responsible,
Schmidt, 2003), decision making and treat and hardworking. These behaviors can re
ment of subordinates (LePine, Hollenbeck, sult in stronger bonds in relationships, which
Ilgen, Sc Hedlund, 1997), leadership skills should contribute to greater marital and re
(Judge, Bono, Ilies, & Gerhardt, 2002), and lationship stability (Baumeister & Leary,
counterproductive work behaviors such as 1995; Lodi-Smith & Roberts, 2007).
stealing and fighting with coworkers (Rob In addition to the significant domains of
erts, Harms, Caspi, & M offitt, 2007). Con work and love, conscientiousness is inverse
scientiousness also influences how individu ly related to a number of maladaptive life
als search for jobs and to what types of jobs outcomes, such as unemployment, home
people apply for, thus shaping the possibili lessness, and being incarcerated (Caspi,
ties for advancement, success, and satisfac Wright, M offitt, Sc Silva, 1998; De Fruyt
tion (Mount, Barrick, Scullen, & Rounds, Sc Mervielde, 1999; Kokko Sc Pulkkinen,
2005). 2 0 0 0 ; Patrick, Hicks, Krueger, Sc Lang,
In terms of relationships, conscientious in 2005). Criminal acts have long been as
dividuals are less likely to get divorced (Rob sociated with the impulsivity facet of con
erts Sc Bogg, 2 0 0 4 ; Roberts, Kuncel, et al., scientiousness (Eysenck Sc Gudjonsson,
2007; Tucker et al., 1998). This makes sense 1989; Krueger, Caspi, M offitt, W hite, Sc
given the fact that relationship satisfaction Stouthamer-Loeber, 1996), which can lead
is predicted by partners’ levels of conscien to a host of problems above and beyond jail
tiousness (Robins, Caspi, & M offitt, 2 0 0 0 ; time, such as difficulties gaining future em
Watson, Hubbard, & Wiese, 200 0 ). Also, ployment. In addition to criminal activity,
a number of key behaviors associated with people low in conscientiousness have trouble
conscientiousness are thought to contribute saving money and have different borrow
to relationship quality. For example, consci ing practices than conscientious individuals
entiousness is related to a number of specific do (Brandstatter & Guth, 2 0 0 0 ; Nyhus &
behaviors that alone are directly related to Webley, 2001). Behaviors associated with
divorce, such as extram arital affairs, spou money extend into purchasing behavior, in
sal abuse, and alcohol abuse (Buss, 1991; which conscientiousness is associated with
Buss Sc Shackelford, 1997). Based on these planning upcoming shopping purchases and
patterns, conscientiousness plays a critical not spontaneously buying unneeded items
role in developing and maintaining success (Verplanken Sc Herabadi, 2001). Addition
ful relationships. ally, low levels of conscientiousness are as
More generally, conscientious behaviors sociated with watching television more often
are likely to shape the quality of long-term (Persegani et al., 2002), which may reflect a
relationships. For example, individuals low lack of responsibility that can lead to mal
in conscientiousness are not as responsible, adaptive outcomes. Interestingly, this lack of
tend to disclose personal information inap responsibility manifests itself in an increase
propriately, are less responsive to their part in accidental injuries (Vollrath, Landolt, &
ners, have poorer social support, act more Ribi, 2003). Furthermore, parents who are
condescendingly, and are not as proficient low in conscientiousness are more likely to
at holding back comments that might cause have children who injure themselves (van
turmoil in a relationship (Buss, 1991; Fin- Aken, Junger, Verhoeven, van Aken, Sc Dek-
kel Sc Campbell, 2 0 0 1 ; Vohs Sc Ciarocco, ovic, 2007).
2004). Over time these behaviors could con These findings illustrate the broad influ
25. C onscien tiou sness 373
erarchy. Optimization refers to steps taken or replace personality traits in the prediction
to enhance or maintain strategies relevant of behavior (Bandura, 1982). For example,
to selected goals. Compensation involves the one’s attitude toward a behavior influences
application of alternate goal-relevant strate that person’s actual behavior. If the indi
gies when previously used strategies or re vidual’s attitude toward the behavior is posi
sources are no longer available. tive, then one is more likely to perform the
Conscientiousness shows moderate cor behavior. Recently, researchers have consid
relations with selection, optimization, and ered the possibility that personality traits are
compensation strategies (Wiese, Freund, & linked to social-cognitive units of analysis
Baltes, 2 0 0 0 ). Bajor and Baltes (2003) fur in a hierarchical structure (e.g., Roberts &
ther demonstrated that SOC variables are Pomerantz, 2004). In this type of hierarchy,
correlated with autonomous goal setting, social-cognitive variables would mediate the
goal expectancies, and goal commitment. relationship of personality traits, such as
More important, they found that SOC strat conscientiousness, and relevant outcomes,
egies mediated the relationship between such as health behaviors. The implication
conscientiousness and job performance. in this conceptualization is that personality
This effect was stronger for managerial traits and social-cognitive units of analysis
jobs. Overall, these results suggest not only should be linked.
that highly conscientious individuals show Several studies have reported linkages be
greater propensities to select and commit tween conscientiousness and various social-
to challenging goals but also that, in situa cognitive units of analysis. Highly conscien
tions in which they have the opportunity to tious individuals have reported higher levels
exercise autonomy, they are more likely to of perceived behavioral control over intended
employ successful strategies for maximizing actions (e.g., Courneya, Bobick, & Schinke,
performance. 1999), less influence from perceived situ
There are many ways that the connection ational constraints (Gerhardt et al., 2007),
between conscientiousness and motivation and less effect from perceived stress (Besser
can be conceptualized and studied. We have & Shackelford, 2007) than individuals low
described how conscientiousness is related in conscientiousness. Highly conscientious
to goal content, appraisals of those goals, individuals also report fewer externalizing
and expected goal-related outcomes. Addi and attention problems (Jensen-Campbell
tionally, all of these goal-relevant variables & M alcolm , 2 0 0 7 ), higher trait emotion
are related to important outcomes. Having al intelligence (e.g., Petrides & Furnham,
and committing to goals are demonstrative 2001), and stronger locus of control and
of one level of motivation-relevant behavior. coping skills (Saklofske, Austin, Galloway,
These can be thought of as aspects relat & Davidson, 2007). Additionally, highly
ing to the desire component of motivation. conscientious individuals possess higher
Managing and selecting goals and resourc levels of self-management (Gerhardt et al.,
es represents a higher order perspective on 2 0 0 7 ), of promotion and prevention foci at
motivation and its connection to conscien work (Wallace & Chen, 20 0 6 ), and of self-
tiousness. Further developing an explicit regulatory learning strategies (Bidjerano &
taxonomy of motivational constructs related Yun Dai, 2007). Conscientiousness has also
to conscientiousness will likely allow us to been associated with numerous attitudes,
better understand how this important per such as having a positive attitude toward ex
sonality variable leads to so many beneficial ercise behaviors (Courneya et al., 1999) and
long-term outcomes. toward health-protective behaviors (Conner
& Abraham, 2001), as well as a negative at
titude toward arriving late to work (Foust,
Conscientiousness and Elicker, & Levy, 2006).
Social Cognition O f all the social-cognitive variables, self-
efficacy is by far the most commonly studied
Social-cognitive models have historically and influential (Gist & Mitchell, 1992). Ac
theorized that social cognition influences cording to Bandura, “perceived self-efficacy
and predicts human behavior and that is concerned with judgments of how well
social-cognitive units of analysis supersede one can execute courses of action required
376 V. M O T IV A T IO N A L D IS P O S IT IO N S
to deal with prospective situations” (1982, in a Singaporean sample (Tay, Ang, &C Van
p. 122). Conscientiousness has been associ Dyne, 2006).
ated with self-efficacy in the study of health Beyond self-efficacy for health outcomes
outcomes through its incorporation into the and vocational activities, conscientiousness
theory of planned behavior (Ajzen, 1985). has been significantly associated with gener
For example, conscientiousness was sig alized self-efficacy in work settings (Burke,
nificantly correlated with attitudes toward M atthiesen, & Pallesen, 2 0 0 6 ; Judge &
health-protective behaviors and health- Ilies, 2002). These studies indicate that
protective self-efficacy in one sample of Brit highly conscientious individuals report high
ish university students, whereas it was signif er levels of self-efficacy for succeeding at all
icantly correlated only with attitude toward tasks encountered in a work setting without
health-protective behaviors and exercise in a regard to the tasks’ nature or requirements.
second sample (Conner & Abraham, 2001). Moreover, there is some evidence that highly
Similarly, conscientiousness was positively conscientious individuals report even higher
correlated with instrumental attitude toward levels of self-efficacy for succeeding at high-
exercise, as well as affective attitude toward complexity tasks than at low-complexity
exercise, but not exercise self-efficacy, in a tasks (Chen, Casper, & Cortina, 2001). In a
study of American students (Rhodes, Cour- meta-analysis of the effect of training on task
neya, &C Hayduk, 2002). performance, conscientiousness was a sig
Additionally, conscientiousness has been nificant predictor of pretraining self-efficacy
consistently associated with various types and posttraining self-efficacy (Colquitt, LeP-
of vocational self-efficacy modeled with ine, & Noe, 2000). Subsequent research has
social-cognitive career theory (Lent, Brown, shown similar associations between consci
& Hackett, 1994). In most of these stud entiousness and pretest learning self-efficacy
ies, conscientiousness has shown significant in a classroom environment (Lee & Klein,
correlations with investigative, social, en 2 0 0 2 ).
terprising, and conventional self-efficacies Bogg (2006) conducted a series of studies
(Hartman & Betz, 2 0 0 7 ; Larson, Wei, Wu, that included an examination of correlations
Borgen, & Bailey, 2 0 0 7 ; Nauta, 2 0 0 4 ; Rot- between the various facets of conscientious
tinghaus, Lindley, Green, & Borgen, 2002). ness and exercise self-efficacy. The construct
These findings indicate that highly conscien of exercise self-efficacy was delineated into
tious individuals report higher levels of abil six subscales that each represented a par
ity than individuals low in conscientiousness ticipant’s self-efficacy to overcome a specific
to succeed at jobs that focus on examining, barrier to exercise, including negative affect,
analyzing, and solving complex problems excuse making, exercising alone, exercise
(investigative); helping, training, and en inconvenience, resistance of others, and bad
lightening others (social); influencing, per weather. The most important facet of con
suading, and leading others (enterprising); scientiousness for exercise self-efficacy was
and working with data, details, and instruc industriousness, which was correlated with
tions (conventional) (Holland, 1997). all six self-efficacy scales. Reliability was
Furthermore, highly conscientious indi significantly correlated with four exercise
viduals report higher levels of self-efficacy self-efficacies: negative affect, excuse mak
than individuals low in conscientiousness ing, exercise inconvenience, and resistance of
for succeeding at tasks that involve science, others, whereas impulse control was signifi
mathematics, writing, helping, teaching, cantly correlated with three self-efficacies:
teamwork, public speaking, leadership, of exercising alone, exercise inconvenience, and
fice services, organizational management, resistance of others. Orderliness was signifi
data management, and project management cantly correlated with only one self-efficacy,
(Hartman & Betz, 200 7 ). Conscientiousness exercise inconvenience, and conventionality
also has been significantly correlated with yielded no significant correlations at all. If
job-search self-efficacy in American samples these results generalize, one would expect
(Brown, Cober, Kane, Levy, 8c Shalhoop, the industriousness facet of conscientious
2 0 0 6 ; Cote, Saks, & Zikic, 2 0 0 6 ), and with ness to be the most consistent predictor of
preinterview and postinterview self-efficacies exercise self-efficacy.
25. C onscien tiou sness 377
The results from studies that have incorpo dividual differences, to see the issue. In the
rated conscientiousness into models includ current tome, conscientiousness is catego
ing self-efficacy are quite consistent with hi rized as a motivational disposition, which
erarchical models of personality that assume is not unreasonable. Nonetheless, as seen
traits, such as conscientiousness, to be broad in our review, conscientiousness, like most
predictors of multiple outcomes, including personality traits, has links to interpersonal
self-efficacy. Conscientiousness predicts self- functioning, emotions, cognitions, and self
efficacies associated with health behaviors, related constructs. Thus conscientiousness
with some evidence for the specific facet could have been categorized into any one or
of industriousness being the core predictor all of the categories that divide this book.
of self-efficacy. Moreover, conscientious A second issue related to the organization
ness is a significant predictor of vocational al scheme of this book (and our field of so
self-efficacy, job-search self-efficacy, and cial/personality psychology) is the fact that
generalized work self-efficacy. As a whole, we take the organizational scheme seriously.
these significant associations between con That is, by labeling our constructs “self,”
scientiousness and self-efficacy suggest that “motivational,” or “emotion,” researchers
conscientiousness should be hierarchically can avoid confronting the fact that they are
incorporated into social-cognitive models all studying highly related constructs. The
that predict various behaviors. distinction, for example, between the con
structs of self-control, self-regulation, and
conscientiousness is difficult to describe.
Conclusion Clearly, self-control and self-regulation (es
pecially behavioral self-regulation) are lower
As we have shown in this chapter, consci order facets of conscientiousness. By over
entiousness has pervasive correlates with looking these links, the field both needlessly
multiple important life outcomes, includ proliferates new constructs and ignores
ing success at work, marital satisfaction and opportunities to leverage complementary
stability, health, and longevity. The link to strengths. As noted earlier, understanding
important life outcomes is predicated on the proximal, state-like aspects of a trait do
and partially explained by the particular be main is critical for understanding the causal
haviors, feelings, and thoughts that consci pathways from trait to outcome (see Roberts,
entiousness predicts. Conscientious people Kuncel, et al., 2007). It also provides a path
behave in ways that facilitate achievement, way to understanding how personality traits
social interaction, and health. They tend develop (Roberts & Caspi, 2003). The proxi
to be more prosocial and hardworking in mal interpersonal, emotional, cognitive, and
achievement settings, more reliable in in motivational mechanisms underlying consci
terpersonal relationships, and more careful entiousness explains why conscientiousness
with health-related behaviors. In turn, these exhibits both continuity and change over
behaviors are the proximal mechanisms that time— issues that are poorly dealt with in
explain better achievement, stable relation classic personality trait psychology (Roberts
ships, and a longer life. Similarly, conscien & Wood, 2006).
tious people tend to experience more adap In sum, conscientiousness is the trait do
tive emotions, motivations, and cognitions. main that sits on the fulcrum between indulg
The pervasive correlates of conscien ing one’s impulses and controlling oneself in
tiousness makes sense when traits are con order to meet higher order ambitions. The
sidered to be hierarchical systems in which trait domain of conscientiousness is multi
the lower order manifestations of traits are faceted, containing components such as self-
the state-like features of the disposition control, industriousness, and conventional
(Roberts & Jackson, 2 0 0 8 ; Roberts & Po- ity. Consistent with a hierarchical model of
merantz, 2004). Yet this poses an interest dispositions, conscientiousness-related traits
ing dilemma for categorizing conscientious predict a whole host of behaviors, emotions,
ness. One needs to go no further than the and thoughts that, in turn, appear to have
scheme used to organize this book, which is functional significance for the well-being,
an excellent representation of the field of in success, and survival of individuals.
378 V. M O T IV A T IO N A L D IS P O S IT IO N S
Mount, M. K., Barrick, M. R., Scullen, S. M., 8c T. E. (200 7). Predicting the counterproductive em
Rounds, J . (2005). Higher order dimensions of the ployee in a child-to-adult prospective study: Evi
Big Five personality traits and the Big Six vocational dence from a 23-year longitudinal study. J o u r n a l o f
interest types. P erso n n el P sy c h o lo g y , 5 8 , 4 4 7 - 4 7 8 . A p p lied P sy ch olog y , 9 2, 14 2 7 -1 4 3 6 .
Nauta, M . (2004). Self-efficacy as a mediator of the Roberts, B. W., 8c Jackson, J. J. (2008). Sociogenomic
relationships between personality factors and c a personality psychology. J o u r n a l o f P erson ality , 76,
reer interests. J o u r n a l o f C a r e e r A s sessm en t, 12(4), 1523-1544.
381-394. Roberts, B. W., Kuncel, N., Shiner, R. N., Caspi, A.,
Noftle, E. E., 8c Robins, R. ( 2 00 7). Personality pre 8c Goldberg, L. (20 0 7). T he power of personality:
dictors of academic outcomes: Big Five correlates A comparative analysis of the predictive validity of
of GPA and SAT scores. J o u r n a l o f P erson ality a n d personality traits, SES, and IQ. P ersp ectiv es in P sy
S o c ia l P sy c h o lo g y , 9 3, 1 1 6 - 1 3 0 . c h o lo g ic a l S cien ce, 2 , 3 1 3 - 3 4 5 .
Norm an, W. T. (1963). Toward an adequate taxonomy Roberts, B. W., O ’Donnell, M., 8c Robins, R. W.
of personality attributes: Replicated factor struc (2004). Goal and personality trait development in
ture in peer nomination personality ratings. Jo u r n a l emerging adulthood. J o u r n a l o f P erson ality a n d S o
o f A b n o r m a l a n d S o c ia l P sy ch olog y , 6 6 , 5 7 4 - 5 8 3 . cia l P sy ch olog y , 87, 5 4 1 - 5 5 0 .
Nyhus, E. K., & Webley, P. (2001). The role of person Roberts, B. W., 8c Pomerantz, E. M . (200 4 ). On traits,
ality in household saving and borrowing behaviour. situations, and their integration: A developmental
E u r o p ea n J o u r n a l o f P erson ality , 15, 8 5 - 1 0 3 . perspective. P erso n a lity a n d S o c ia l P sy ch o lo g y R e
O ’Brien, T. B., 8c DeLongis, A. (1996). T he interaction view , 8(4), 4 0 2 - 4 1 6 .
al context of problem-, emotion-, and relationship- Roberts, B. W., 8c Robins, R. W. (20 00 ). Broad dis
focused coping: T he role of the Big Five personality positions, broad aspirations: The intersection of the
factors. J o u r n a l o f P erson ality , 6 4 , 7 7 5 - 8 1 1 . Big Five dimensions and major life goals. P ers o n a li
Ones, D. S., Viswesvaran, C., 8c Schmidt, F. L. (2003). ty a n d S o c ia l P sy ch o lo g y B u lletin , 2 6 , 1 2 8 4 - 1 2 9 6 .
Personality and absenteeism: A meta-analysis of in Roberts, B. W., 8c Wood, D. (200 6). Personality de
tegrity tests. E u r o p ea n J o u r n a l o f P erson ality , 17, velopment in the context of the neo-socioanalytic
19-38. model of personality. In D. Mroczek 8c T. Little
Patrick, C. J ., Hicks, B. M ., Krueger, R. F., 8c Lang, (Eds.), H a n d b o o k o f p e r s o n a lity d e v e lo p m e n t
A. R. (2 005). Relations between psychopathy fac (pp. 11-39). Mahwah, NJ: Erlbaum.
ets and externalizing in a criminal offender sample. Robins, R. W., Caspi, A., 8c Moffitt, T. E. (20 00 ). Two
J o u r n a l o f P erson ality D iso rd ers, 19, 3 3 9 - 3 5 6 . personalities, one relationship: Both partners’ per
Persegani, C ., Russo, P., Carucci, C., Nicolini, M., Pa- sonality traits shape the quality of a relationship.
peschi, L. L., 8c Trimarchi, M . T. (2 002). Television J o u r n a l o f P erso n a lity a n d S o c ia l P sy c h o lo g y , 79,
viewing and personality structure in children. P er 251-259.
so n a lity a n d In d iv id u a l D iffe r e n c e s , 3 2 , 9 7 7 - 9 9 0 . Rolland, J. P., 8c De Fruyt, F. (2003). T he validity of
Petrides, K. V., 8c Furnham, A. (2001). Trait em o F F M personality dimensions and maladaptive traits
tional intelligence: Psychometric investigation with to predict negative affects at work: A six month pro
reference to established trait taxonomies. E u ro p ea n spective study in a military sample. E u r o p ea n J o u r
J o u r n a l o f P erson ality , IS , 4 2 5 - 4 4 8 . n al o f P erson ality , 17, 101 -1 21 .
Rhodes, R. E., Courneya, K. S., 8c Hayduk, L. A. Rottinghaus, P. J ., Lindley, L. D., Green, M. A., 8c
(2002). Does personality moderate the theory of Borgen, F. H. (2002). Educational aspirations: The
planned behavior in the exercise domain? J o u r n a l contribution of personality, self-efficacy, and inter
o f S p o rt a n d E x e r c is e P sy ch o lo g y , 2 4 (2 ), 1 2 0 - 1 3 2 . ests. J o u r n a l o f V o c a tio n a l B eh a v io r, 61, 1- 19.
Roberts, B. W., 8c Bogg, T. (20 04 ). A 30-year lon Saklofske, D. H., Austin, E. J ., Galloway, J ., 8c D a
gitudinal study of the relationships between vidson, K. ( 2 00 7 ). Individual difference correlates
conscientiousness-related traits and the family of health-related behaviours: Preliminary evidence
structure and health-behavior factors that affect for links between emotional intelligence and co p
health. J o u r n a l o f P erson ality , 7 2 , 3 2 5 - 3 5 4 . ing. P erso n a lity a n d In d iv id u a l D iffe r e n c e s , 42(3),
Roberts, B. W., Bogg, T., Walton, K. E ., Chernyshenko, 491-502.
O. S., 8c Stark, S. E. ( 2 00 4). A lexical investigation Scher, S. J ., 8c Osterman, N. M . (2 002). Procrastina
of the lower-order structure of conscientiousness. tion, conscientiousness, anxiety, and goals: Explor
J o u r n a l o f R e se a rc h in P erson ality , 3 8 , 1 6 4 - 1 7 8 . ing the measurement and correlates of procrastina
Roberts, B. W., 8c Caspi, A. (2003). The cumulative tion among school-aged children. P sy ch o lo g y in the
continuity model of personality development: Strik S c h o o ls , 3 9, 3 8 5 - 3 9 8
ing a balance between continuity and change in Tangney, J. P. (1996). Conceptual and methodological
personality traits across the life course. In R. M. issues in the assessment of shame and guilt. B e h a v
Staudinger 8c U. Lindenberger (Eds.), U n d erstan d io u r R e se a rc h a n d T h era p y , 3 4 , 7 4 1 - 7 5 4 .
in g h u m a n d e v e lo p m e n t : L ife s p a n p s y c h o lo g y in Tay, C., Ang, S., 8c Van Dyne, L. ( 2 0 06 ). Personality,
ex c h a n g e w ith o t h e r d iscip lin es (pp. 1 8 3 -2 1 4 ). D o r biographical characteristics, and job interview suc
drecht, The Netherlands: Kluwer Academic. cess: A longitudinal study of the mediating effects
Roberts, B. W., Chernyshenko, O., Stark, S., 8c Gold of interviewing self-efficacy and the moderating ef
berg, L. (2005). T he structure of conscientiousness: fects of internal locus of causality. J o u r n a l o f A p
An empirical investigation based on seven major p lie d P sy ch olog y , 91(2), 4 4 6 - 4 5 4 .
personality questionnaires. P erso n n el P sy ch olog y , Tellegen, A. (1988). The analysis of consistency in
58, 103-139. personality assessment. J o u r n a l o f P erson ality , 5 6 ,
Roberts, B. W., Harms, P. D., Caspi, A., & Moffitt, 621-663.
25. C onscien tiou sness 381
Achievement Motivation
D a v id E . C o n r o y
A n d r e w J . E l l io t
T o d d M . T h ra sh
382
26. A ch iev em en t M otivation 383
to the pursuit of competence. For example, 1989; Stipek, Recchia, & M cClintic, 1992).
the need for achievem ent was conceived One class of emotions can be distinguished
as “the desire or tendency to do things as for their unique role in self-evaluative pro
rapidly and/or as well as possible” (p. 164). cessing: the self-conscious or social emo
Likewise, the need for infavoidance repre tions (Tracy, Robins, & Tangney, 2007).
sents a desire “to avoid humiliation, to quit These emotions include pride and shame,
embarrassing situations or to avoid condi which are the two exemplars most frequent
tions which may lead to belittlement: the ly posited to be associated with achievement
scorn, derision or indifference of others, to motives. Anticipatory pride in succeeding
refrain from action because of fear of fail was proposed as the basis for the need for
ure” (p. 192). These two needs parallel the achievement (nAch), and anticipatory shame
appetitive and aversive achievement motives in failing was proposed as the basis for fear
that later emerged in the motive-based ap of failure (FF) (Atkinson, 1957; McClelland
proach to achievement motivation. et al., 1953).
Achievemen t imagery" Need for achievem ent and success Adjectives th at positively
Stated need for achiev ement In strum en ta l activity to achieve evaluate perfo rm an ces
In strum en ta l activity (successful, success G o a ls or pe rfo rm a n ce s that
doubtfu l, or unsuccessful) E x p e c t a tio n o f success are described in ways that
Anticipato ry goal states (positive Praise suggest positive evaluation
or negative) Positive af fe ct M en tio n o f winnin g or
O b stacle s or blocks (perso nal or Success theme co m peting w ith others
environ men tal) Failure, doing badly, o r other
N u rtu ra n t press lack of exce llence
Affective states (positive or Unique accomplish men ts
negative)
Achievement thema
‘'Stories in which achievement imagery is altogether absent receive a negative achievement motivation score. Those in
which achievement imagery is doubtful receive a zero score.
the available coding systems because it can seven major coding categories and is theo
be applied to any data that are at least partly retically consistent with prevailing concepts
imaginative (e.g., speeches, conversations, of FF. Working independently of Heckhaus
fictional writing). This system is similar to en, Birney and colleagues (1969) used an
the Heckhausen system in that the number approach similar to that of McClelland and
of coding categories is limited compared colleagues (1953) to develop a system for
with the McClelland and colleagues system; coding Hostile Press in stories. This Hostile
however, the content of the categories is Press score was based on imagery depicting
somewhat unique compared with the other a threat presented by the situation to the
systems. This coding system also focuses participant and interpreted as an indicator
exclusively on approach-based motives and of FF. Not surprisingly, this coding system
does not differentiate them from avoidance- is also vulnerable to concerns about content
based motives— a nuance that may help to relevance. For example, it is not clear from
explain why text concerning “failure, doing a theoretical standpoint why one would
badly, or other lack of excellence” (W in infer high FF from stories that depict suc
ter, 1994, p. 10) is coded positively for the cessful instrumental activity, anticipation of
achievement motive. successful goal attainment, or pleasant af
The categories in the two major coding fective reactions. Overall, we concur with
systems for FF are summarized in Table the conclusions of McClelland (1987) and
2 6 .2 . The Heckhausen (1963) system has Schultheiss (2001)— the Heckhausen coding
system provides the best fantasy-based ap distinguished and even distanced from mo
proach for assessing nAch and FF motives. tives by denying them status as a motive and
The more recently developed fantasy- calling them instead values (e.g., deCharms,
based methods for assessing achievement M orrison, Reitman, & M cClelland, 1955).
motives are quite refined and are less vul McClelland later backed off this position
nerable to methodological criticisms that and recognized the existence of explicit mo
were frequently leveled in the 20th century tives as a separate motivational system (M c
(for details on these improved methods, see Clelland, Koestner, & Weinberger, 1989).
Schultheiss & Pang, 20 0 7 ; Smith, 1992). This theoretical reconciliation was based on
The Picture Story Exercise described by the conclusion that different motivational
Schultheiss and Pang (2007) is one example systems exist: a primitive implicit motive
of a methodologically rigorous protocol for system that is grounded in affective arousal
administering and scoring fantasy-based and a cognitively elaborated system that is
measures that yields psychometrically sound based on an “elaborate system of explicit
scores for motives. In addition to the pro goals, desires, and commitments” (M cClel
jective measures described previously, the land et al., 1955, p. 700). The former system
nAch and FF were also commonly assessed is expressed in fantasy-based measures such
using self-report measures (e.g., Atkinson as the Picture Story Exercise, whereas the
& I.itwin, 1960; Conroy, Metzler, &c W il latter is consciously accessible and may be
low, 2 0 0 2 ; Feather, 1965; Hagtvet & Ben assessed using self-report methods.
son, 1997; Herman, 1990; Jackson, 1974; Schultheiss (2007) linked these moti
Spence & Helmreich, 1983). Examples of vational systems and their corresponding
items used to assess nAch include “I like to assessment methods to different memory
work hard” and “Once I undertake a task, I systems— implicit motives and fantasy-based
persist” (Spence & Helmreich, 1983, p. 42). measures tap into nondeclarative memory
Examples of items used to assess FF include systems of which the individual is not con
“When I am failing, it is embarrassing if oth sciously aware, whereas explicit motives and
ers are there to see it” and “When I am fail self-report questionnaires tap into declara
ing, I believe that my doubters feel that they tive memory systems of which the individ
were right about me” (Conroy et al., 2 0 0 2 , ual is consciously aware. The differences in
p. 90). In our view, the self-report measures these underlying memory systems may help
that presently provide the most valid scores to explain differences in the outcomes pre
for nAch and FF are the W ork-Family O ri dicted by implicit and explicit motives. The
entation Questionnaire (particularly the nondeclarative memories tapped by implicit
work-mastery score; Spence & Helmreich, motives may be linked to procedural learn
1983) and the Performance Failure Appraisal ing and Pavlovian conditioning that likely
Inventory (Conroy et al., 2002), respectively. underlie the acquisition of skills, habits, and
Semiprojective tests have even been proposed emotional associations. In contrast, the de
to try to capitalize on the strengths of both clarative memory system tapped by explicit
projective and self-report assessments (e.g., motives may be linked most directly to out
Schmalt, 1999), although these measures comes based in semantic and episodic mem
have been used less frequently than either ories, such as conscious attitudes, retrospec
projective or self-report measures. tive judgments, and future intentions. Much
One source of great controversy and, ul remains to be learned about why implicit
timately, insight in the achievement motiva and explicit motives differ, but the ideas ar
tion literature is the fact that scores from ticulated by Schultheiss (2007; Schultheiss
projective and self-report measures tend to & Pang, 2007) provide fruitful ground for
correlate less strongly than would be ex theory development and testing.
pected if they were assessing a common mo A significant emerging line of work in this
tive (Spangler, 1992). Critics from either side area examines the factors that influence the
often took this as evidence that the other relationship between implicit and explicit
approach did not yield valid scores of the achievement motives. As Thrash, Elliot, and
relevant motive. In early writings, what we Schultheiss (2007) noted, the conclusion
now call self-attributed or explicit (i.e., ques that implicit and explicit nAch are largely
tionnaire-based) scores were intentionally uncorrelated is reminiscent of early reports
386 V. M O T IV A T IO N A L D IS P O S IT IO N S
of poor consistency between traits and be Several studies have sought a fuller char
havior and between attitudes and behavior acterization of the relationship between
(Mischel, 1968; Wicker, 1969). In both of implicit and explicit nAch by identifying
these prior consistency literatures, research dispositional variables that function as mod
ers subsequently uncovered two types of erators. Thrash and Elliot (2002) examined
evidence that traits or attitudes are more the moderating role of self-determ ination,
systematically related to behavior than had which refers to autonomy or authenticity
been apparent in early research. First, meth (Deci & Ryan, 1985). Thrash and Elliot ar
odological refinements resulted in stronger gued that feelings of self-determination re
consistency coefficients. Second, consistency flect the development of explicit values that
itself was found to vary systematically as a are well aligned with deeply grounded im
function of moderator variables. In paral plicit motivational tendencies. As expected,
lel to the developments in these literatures, self-determination was found to moderate
motive researchers have documented two the relation between implicit and explicit
classes of factors— methodological factors nAch. Implicit nAch was a robust predictor
and moderator variables— that predict the of explicit nAch among individuals high in
degree of association between implicit and self-determination (r = .40) but was unrelat
explicit nAch. ed to explicit nAch among individuals low in
Regarding a methodological factor, self-determination (r = -.0 7 ).
Thrash and colleagues (2007) argued that More recently, Thrash and colleagues
the correlation between implicit and explicit (2007) examined three additional disposi
nAch may have been underestimated in past tional moderators: private body conscious
research due to poor correspondence of con ness, self-monitoring, and preference for
tent between implicit and explicit measures. consistency. Private body consciousness
Many popular measures of explicit nAch are refers to a sensitivity to internal bodily
based on M urray’s (1938) conceptualization processes (Miller, Murphy, & Buss, 1981).
of nAch (e.g., Jackson, 1974), whereas M c Thrash and colleagues proposed that private
Clelland and colleagues’ (1953) widely used body consciousness may promote congru
coding system for implicit nAch was derived ence between implicit and explicit nAch,
empirically by examining how achievement because the effects of implicit motive arous
imagery changes when the motive is and is al are embodied and may be perceptible
not aroused. One unintended consequence as diffuse gut feelings or surges of energy.
of this approach was that the coding system Self-m onitoring is the tendency to monitor
deviates from Murray’s conceptualization of the social environment and to adjust one’s
the nAch (Koestner & McClelland, 1990). behavior or attitudes accordingly (Snyder
For example, the scoring system used by & Gangestad, 1986). Self-monitoring was
McClelland and colleagues counts negative posited to impede congruence, because the
anticipatory goal-state imagery toward the achievement values internalized from the
nAch score (e.g., “The boy thinks he just social environment are less likely to cor
can’t make it through college”; p. 129)— respond to one’s implicit motives than are
this content is exclusive of the achievement internally generated values. Preference for
need described by Murray (1938). Thrash consistency refers to a tendency to seek con
and colleagues reported that implicit nAch, sistency among cognitions (Cialdini, Trost,
assessed using Schultheiss’s (2001) transla & Newsom, 1995). Preference for consis
tion of Heckhausen’s coding system, was tency was expected to predict greater con
uncorrelated with three existing measures gruence, because individuals high in this
of explicit nAch (rs = .0 0 , .0 0 , and .02); in trait would be more motivated to reconcile
contrast, it was significantly correlated with discrepancies between explicit motives and
a new measure of explicit nAch (r = .17) that any rudimentary knowledge of one’s implicit
was designed to closely match the implicit motives. Results showed that all three traits
nAch coding system in content. This finding moderated the association between implicit
indicates that implicit and explicit nAch are and explicit nAch. Moreover, all three traits
systematically related, albeit weakly, when functioned as independent moderators, sug
assessed with measures that are properly gesting that multiple, distinct processes are
matched for content. responsible for motive congruence.
2 6 . A ch iev em en t M otivation 387
In related research on the congruence be tualized with respect to both the specific
tween implicit motives and explicit goals, achievement context and time. In theory,
Brunstein (2001) reported that state-oriented motives may be “canalized,” or channeled
individuals, who have a tendency toward into specific achievement contexts (e.g.,
indecisiveness and hesitation (as opposed sports or classroom achievement) in differ
to action-oriented individuals, who have a ent individuals (Thrash & Elliot, 2001), but
tendency toward decisiveness and initiative), researchers have generally not exploited this
are more likely to adopt goals that are incon- fact to maximize the predictive validity of
gruent with implicit motives. More recently, their instruments (Thrash & Hurst, 2008).
Baumann, Kaschel, and Kuhl (2005) report From a methodological perspective, it
ed that state orientation predicted incon should also be noted that it can be difficult to
gruence between implicit and explicit nAch interpret many findings in the achievement-
only when individuals were under stress. In motives literature. Researchers often fo
addition, motive incongruence led to lower cused their analyses on “resultant motiva
well-being and partially mediated the effect tion” scores that represented the difference
of the state orientation x stress interaction between standardized scores for nAch and
on well-being. FF in a sample. Large positive and large
negative resultant scores have clear interpre
tations (i.e., high scores for one motive and
Sum m ary o f M otive-Based Approaches
low scores for the other), but it is less clear
The motive-based approaches to achieve what resultant scores of zero indicate about
ment motivation are based on relatively the level of individual motives. Participants
stable individual differences in affective as may have scored high for both motives, av
sociations with success and failure. Motives erage for both motives, or low for both mo
exist at two levels of analysis— implicit mo tives. In contemporary research, it is prefer
tives that are grounded in deeply rooted af able to examine main and interactive effects
fective structures and are not readily acces of the achievement motives instead of losing
sible to awareness and explicit motives that valuable information by calculating a resul
are grounded in consciously held values, tant motivation score.
beliefs, or attitudes. These motive systems
do not necessarily converge for all individu
als, and the available evidence indicates that G oal-Based Approaches
they predict quite different outcomes. R e to A chievem ent M otivation
cent research has shown that implicit and
explicit nAch are not strictly independent An alternative approach to studying achieve
and that methodological and dispositional ment motivation emerged in the form of
factors influence the association between achievement goal theory. Achievement goal
them. Poor alignment between implicit and theory grew from the observation of two
explicit nAch is associated with low levels of very different patterns of responses to fail
well-being. ure among young children: a m astery re
Strengths of this motive-based approach spon se , characterized by low-effort attri
to achievement motivation include the focus butions, persistence, increased competence
on how behavior is energized (via learned expectancies, selection of challenging tasks,
anticipatory affect, particularly involving and improved performance; and a helpless
pride and shame) and the general distinc response , characterized by low-ability at
tion between approach and avoidance ori tributions, unpleasant affect, decreased
entations for achievement behavior (Elliot, competence expectancies, selection of easy
1997). Two major limitations of this ap tasks, and reduced performance (Diener &
proach have also been identified: (1) It does Dweck, 1978, 1980; Dweck, 1975). Dweck
not differentiate beyond omnibus approach (1986; Dweck & Elliott, 1983; E. S. Elliott &
or avoidance strivings and (2) as decontextu- Dweck, 1988) proposed that these responses
alized constructs, motives are not well suited reflected different goals that children adopt
for predicting context-specific processes and in achievement pursuits. Some view achieve
outcomes (Elliot, 1997). This latter point is ment pursuits as opportunities to learn and
important because motives are decontex- to increase their competence (learning goals);
388 V. M O T IV A T IO N A L D IS P O S IT IO N S
others view achievement pursuits as oppor learning goals and mixed consequences for
tunities to establish their standing with re ego involvement and performance goals.
spect to intelligence or ability in comparison For example, whereas task involvement and
with their peers (performance goals). Learn learning goals exhibit consistent positive re
ing goals were presumed to facilitate mastery lations with intrinsic motivation for a task,
responses because they orient the person to ego involvement and performance goals typ
the process of learning and improving. In ically exhibit a mixed profile of null and neg
contrast, performance goals were thought ative relations. To resolve ambiguities about
to engender helpless responses because they the consequences of this goal, Elliot (1997;
orient the person to factors outside of his or Elliot & Harackiewicz, 1996) proposed that
her control and create a threatening environ it was necessary to consider the valence of
ment for achievement pursuits. goals in addition to how competence is de
A similar approach grew from the work fined in the goal.
of Nicholls (1976, 1978, 1984) on devel The valence of an achievement goal refers
opmental changes in children’s conceptions to whether the individual is focused on suc
of ability. In early childhood, children pos ceeding (an approach goal) or on not failing
sess an undifferentiated concept of abil (an avoidance goal). Early goal theorists in
ity that equates competence with learning timated that avoiding incompetence may be
and effort. By trying hard, they are able a relevant achievement goal (e.g., Nicholls,
to improve and therefore feel competent. Patashnick, Cheung, Thorkildsen, & Lauer,
Around age 12, children begin to differen 1989); however, research in the dichoto-
tiate between two primary internal sources mous-goals tradition focused explicitly on
of achievement outcomes: effort and ability. approach-valenced achievement goals that
This differentiated concept of ability leads differed only in how competence was de
to changes in how children construe com fined. Crossing the definition o f com petence
petence. Ability is now inferred from the (task- or self-referenced competence vs. nor-
amount of effort required to produce a suc matively referenced competence) with the
cessful performance— outperforming a peer valence o f the com peten ce-based possibility
while exerting minimal effort would lead represented in the goal (e.g., being compe
to perceptions of greater ability than if one tent vs. avoiding incompetence) yields the 2
had to work very hard to outperform a peer. x 2 achievement-goal framework proposed
Nicholls (1984) extended these ideas about by Elliot (1999; see also Elliot & McGregor,
different conceptions of ability by proposing 2001) and depicted in Figure 26.1.
that they are the basis for two major achieve M astery-approach (MAp) goals focus the
ment goals. People who pursued competence person on performing a task as well as possi
in an undifferentiated sense— meaning that ble (task-referenced competence) or surpass
they focused on effort and learning— were ing his or her previous level of performance
said to be in a state of task involvement. on that task (self-referenced competence).
People who pursued competence in a differ For example, a student with an MAp goal
entiated sense— meaning that they focused could strive to ace an exam or to exceed
on demonstrating ability by outperforming his or her score on previous exams in that
others with an economy of effort— were said course. M astery-avoidance (MAv) goals
to be in a state of ego involvement. These focus the person on not making mistakes
task and ego states of involvement represent (avoiding task-referenced incompetence) or
ed the purpose of achievement behavior and on maintaining a previously established level
overlap considerably with the aims or foci of of performance (avoiding self-referenced in
behavior associated with learning and per competence). A politician with an MAv goal
formance goals, respectively. might be focused on not making a mistake
These converging lines of work pro in a speech or on not doing worse than
vided the foundation for what has come she or he did while practicing the speech.
to be known as the dichotomous model of P erform an ce-approach (PAp) goals focus
achievement goals. The dichotomous model the person on outperforming others (nor-
of achievement goals inspired a large volume matively referenced competence), such as the
of research that consistently demonstrated salesperson who is focused on producing the
adaptive qualities of task involvement and best sales figures in her or his division. Fi
2 6 . A ch iev em en t M otivation 389
Definition of Competence
M astery Perform ance
(self- or task - (norm atively
referen ced) referen ced )
Mastery- Performance-
Approach
.C (striving for
Approach Approach
co m p e ten c e) Goals Goals
CO
O
Q)
^ Avoidance
Mastery- Performance-
.0) (striving aw ay
4® from Avoidance Avoidance
in com peten ce) Goals Goals
F IG U R E 26.1. The 2 x 2 achievement goal framework. From Elliot and McGregor (2001, p. 502).
Copyright 2001 by the American Psychological Association. Adapted by permission.
nally, p erform an ce-avoidan ce (PAv) goals that these four goals have unique profiles of
focus the person on not being outperformed antecedents and consequences.
by others (avoiding normatively referenced The strengths and weaknesses of the
incompetence), such as the swimmer whose goal-based approach to achievement mo
primary objective is to avoid finishing last tivation generally complement those of the
in his or her qualifying heat during a meet. motive-based approach reviewed earlier (El
An emerging body of evidence from social- liot, 1997). Recall that the motive-based
personality, educational, sports, and indus approach emphasizes the energization of
trial/organizational psychology has made achievement behavior but offers only gen
it increasingly apparent that considering eral insight into how such behavior is di
both dimensions of achievement goals (i.e., rected (e.g., toward competence, away from
definition of competence and goal valence) incompetence). The goal-based approach of
enhances the predictive power of the goal fers little with respect to the energization of
construct (for a review, see M oller & Elliot, achievement behavior, but it specifically ac
2006). counts for the different ways that individuals
can orient their achievement behavior to feel
competent (e.g., definitions of competence).
Sum m ary o f G oal-B ased Approaches
The dynamic nature of the goal construct
Goal-based approaches to achievement itself also makes it possible to account for
motivation are based on the different com- intraindividual variability in the quality of
petence-based aims or purposes of achieve achievement strivings that is more difficult
ment strivings. Early research focused on a within the motive-based tradition.
dichotomous model of goals that emphasized
the distinction between mastery- and perfor-
mance-based definitions of competence. Re The H ierarchical Model
cent work has convincingly demonstrated the o f A chievem ent M otivation
conceptual and predictive value of attending
to the approach-avoidance valence of goals. The hierarchical model of achievement mo
The corresponding 2 x 2 achievement-goal tivation was proposed to integrate these
framework has received substantial atten complementary approaches and to increase
tion, and results consistently demonstrate the conceptual clarity of the achievement
390 V. M O T IV A T IO N A L D IS P O S IT IO N S
consistent with the finding that implicit FF was linked with prosocial and noncompliant
was associated with greater impression man behaviors in the workplace (Puffer, 1987).
agement during a purported creativity test Supervisors in a chain of retail stores rated
(Cohen &c Teevan, 1974). Birney and col employees who were high in nAch as dem
leagues (1969) also reported a series of stud onstrating more prosocial behaviors, such
ies demonstrating that FF was linked with as assisting coworkers and pursuing solu
greater conformity to others’ judgments tions to customer service problems. They
and opinions, but that this association ex also rated these employees as demonstrat
ists only when the person is in a social con ing fewer noncompliant behaviors, such as
text. Collectively, these results suggest that complaining about work conditions, lying
relational concerns and insecurities are in to customers, and taking excessive breaks.
tertwined with implicit FF. Young adults ap In another study, high nAch participants al
pear to regulate these concerns with appeas located rewards to a partner based on the
ing behaviors. In contrast, maternal reports partner’s performance instead of the part
indicate that children high in FF engage in ner’s reward-allocation strategy (O’Malley
more attention-seeking behavior than chil & Schubarth, 1984). These findings are con
dren low in FF (Singh, 1992). sistent with proposals that the nAch orients
Overall, these results present a picture people toward efficient and just behaviors
of two implicit motives with quite different in their competence pursuits; however, the
social consequences. Implicit nAch appears study neither evaluated nor controlled for
to facilitate successful social interactions, the influence of FF.
although the achievement pursuit may draw Explicit FF has been linked to self-
the individual’s attention to the task, may protective behavior. Children high in FF en
reduce the accuracy of person perception, gage in cheating more frequently than peers
and may enhance social status. On the other low in FF, presumably to enhance their
hand, implicit FF may inhibit social behav probability of avoiding failure (Monte &
ior in different ways at different points in Fish, 1987; Shelton & Hill, 1969). In college
life. Children high in FF may act out and en students, FF has been shown to negatively
gage in problem behaviors to solicit parental predict students’ likelihood of telling their
attention, whereas young adults may inhibit parents about their performance on a task
agentic behavior because their concerns over they just completed if they failed at the task
social approval and acceptance take prece and to positively predict their likelihood of
dence over genuine competence. telling their parents if they succeeded at the
task (McGregor & Elliot, 2005).
Emerging results from our research also
E xplicit Motives
suggest that achievement motives have dis
Compared with the implicit-motives litera tinct relations with different forms of inter
ture, considerably less evidence is available personal problems. Anticipatory pride (i.e.,
regarding links between explicit achievement explicit nAch) has a very limited association
motives and social behaviors. The following with interpersonal problems; if anything,
review is limited to studies that focused on low levels of the nAch may be associated
nAch or FF; related constructs such as test with submissive interpersonal problems
anxiety are beyond the scope of this coverage. (Conroy, Elliot, & Pincus, 2009). On the
Studies that focused on resultant motivation other hand, anticipatory shame (i.e., explicit
(i.e., standardized nAch minus standardized FF) is associated with significant interper
FF) also were excluded, because it is impos sonal distress. This distress is reported by
sible to interpret which motive is responsible individuals high in FF themselves, as well as
for any observed effects. Unfortunately, this being reported by knowledgeable peers. Al
delimitation leads us to exclude some very though self-reported FF was not associated
interesting studies concerning achievement with specific interpersonal problems, peers
motivation and leadership (e.g., Sorrentino, described friends high in FF as being more
1973; Sorrentino & Field, 1986; Sorrentino exploitable, overly nurturant, and intrusive
& Sheppard, 1978). than friends low in FF.
In one study that specifically examined ex In another study that focused on college
plicit motives and social behavior, the nAch students with high FF, two clusters of self
392 V. M O T IV A T IO N A L D IS P O S IT IO N S
Dominant
F IG U R E 26.2. The interpersonal circle with illustrative exam ples o f social behaviors that have been
linked to achievem ent goals.
judges’ ratings of respectful and supportive PAp goals (Duda, Olson, & Templin, 1986).
communication from members of a triad Self-reported use of aggression to gain a
whose achievement pursuit was disrupted. competitive advantage has also been linked
When the triad was given a difficult goal, with high PAp goals in athletes (Stornes &
PAp goals were negatively associated with Ommundsen, 2004). Based on these lim
judges’ ratings of respectful and supportive ited results, it appears that normative defi
communication; PAp goals were unassoci nitions of competence (performance-based
ated with judges’ ratings when triads were goals) are associated with the potential for
given an easy goal. In the sport domain, increased aggressive behavior— perhaps be
MAp goals have been positively associated cause individuals with performance-based
with athletes’ respect for their opponents, goals are so preoccupied with social status
rules, and officials, whereas PAp goals have that, in their basest moments, they resort to
been negatively associated with these impor primitive means of attaining it.
tant indicators of sportspersonship (Stornes
& Ommundsen, 2004). The PAp goal ef
Hostile to Hostile—Submissive Behavior:
fects were somewhat moderated by athletes’
Social Loafing
perceptions of the motivational climate; a
strong mastery-motivational climate weak Social loafing refers to the phenomenon of a
ens relations between PAp goals and poor “decrease in individual effort due to the so
sportspersonship. cial presence of other persons” (Latane, W il
liams, & Harkins, 1979, p. 823). This be
havior is submissive because the individual
Dominant Behavior (Hostile to Friendly):
is reducing her or his efforts to influence the
Leadership
group or the group’s performance. The fact
One of the central challenges of leadership that such behavior may harm the group sug
involves influencing others. Although there gests that it may be a hostile interpersonal
are many ways of exerting influence (see process, although that need not always be
House & Singh, 1987), leadership in its var the case. Social loafing in academic work
ied styles remains a prototypically agentic has been positively associated with students’
interpersonal behavior. Yamaguchi (2001) PAp goals, but not their MAp goals (Linnen-
used a qualitative analysis to compare the brink, 2003). From a different perspective,
leadership styles that emerged in 10 groups athletes with high PAp goals are more likely
of children working on a task. Groups that to report that their teammates are withhold
were given MAp goals at the beginning of ing effort during performances; however,
the task exhibited a shared leadership style goals were not associated with athletes’ re
between the members, whereas groups that ports that they themselves would withhold
were given PAp goals exhibited a dominant effort if they perceived teammates to be
leadership style in which one member “over loafing (Hoigaard & Ommundsen, 2007).
took and overpowered the [task] and group Performance-based achievement goals may
processes” (p. 683). This effect is consistent lead to effort reductions because ability is
with other findings that children who adopt inferred in part from the amount of effort an
performance-based goals are exceedingly individual must expend to be successful at
focused on social status (e.g., Levy et al., a task (Nicholls, 1984). In other words, the
2004). hostile-submissive act of withholding effort
may actually be a strategy for demonstrat
ing competence in group work. One would
Hostile—Dominant Behavior: Aggression
assume that this effect would be more pro
Aggressive behavior involves an immediate nounced for individuals focused on avoid
intent to injure another individual (Anderson ing incompetence than for those focused on
& Bushman, 2002). Limited research exists being competent.
on links between achievement goals and ag
gressive behavior. An early study in the sport
G roup Processes
domain found that aggressive behavior was
perceived as more legitimate by athletes who There are many possible examples of social
had low MAp goals and moderately high behaviors that do not map directly onto the
26. A ch iev em en t M otivation 395
interpersonal circle. Three relevant examples with the use of epistemic conflict-regulation
that have been linked to achievement goals strategies, whereas PAp goals have been as
involve information exchange, conflict regu sociated with the use of relational conflict-
lation, and role structure adaptation within regulation strategies (Darnon, Muller,
groups. Schrager, Pannuzzo, & Butera, 2006).
plex phenomenon, and a relatively narrow which they can anchor their measures of so
range of behavior has been studied. Some of cial behavior in broad nomological networks
the seemingly simple behaviors that were re to facilitate future theorizing.
viewed may have multiple components (e.g., Finally, it seems appropriate to conclude
help seeking) (Nelson-LeGall, 1985), and it by returning to a fundamental point in
may be simplistic to cast all of these behav individual-differences research. Both the
iors into a single interpersonal circumplex person and situation are important factors
octant. Second, few studies have examined to consider when predicting social behav
the social impact of avoidance goals. It will ior. Our best chance for understanding how
be important to determine how this charac achievement motivation influences social
teristic of achievement goals influences so behavior will require us to engage in more
cial behaviors. Finally, most research in this process-focused research that highlights
area has focused on individual differences in consistencies in behavioral variability as a
goals and has not considered how the m o function of situational characteristics (e.g.,
tivational climate might influence social be Mischel & Shoda, 1995).
havior (either as a main effect or in an inter
action with states of goal involvement).
R eferen ces
Anderson, C. A., & Bushman, B. J . (20 02). Human ag
Future D irections
gression. A n n u al R e v ie w o f P sy ch olog y , 5 3, 2 7 - 5 1 .
Atkinson, J . W. (1957). Motivational determinants
This chapter opened with the proposal that of risk-taking behavior. P sy c h o lo g ic a l R e v iew , 6 4 ,
competence strivings frequently involve so 359-372.
cial behavior and that achievement motiva Atkinson, J . W., & Litwin, G. (1960). Achievement
motive and test anxiety conceived as motive to ap
tion theories should speak to these processes proach success and motive to avoid failure. J o u r n a l
as well as proximal achievement processes o f A b n o r m a l a n d S o c ia l P sy ch olog y , 6 0 , 5 2 - 6 4 .
and outcomes. The evidence reviewed herein Bakan, D. (1966). T h e d u a lity o f h u m an ex is te n c e :
clearly indicates that individual differences Is o la tio n a n d co m m u n io n in W estern m an . Boston:
Beacon Press.
in achievement motivation are associated
Bandura, A. (1997). S e lf-e ffic a c y : T h e e x e r c is e o f c o n
with different patterns of social behavior. trol. New York: Freeman.
Looking forward, we see great potential Baumann, N., Kaschel, R., & Kuhl, J . (2 005). Striving
for using achievement motivation theories to for unwanted goals: Stress-dependent discrepancies
explain social behavior during competence between explicit and implicit achievement motives
reduce subjective well-being and increase psychoso
pursuits. As noted at the beginning of this matic symptoms. J o u r n a l o f P erson ality a n d S o c ia l
chapter, many approaches have been em P sy ch olog y , 8 9, 7 8 1 - 7 9 9 .
ployed in achievement motivation research, Berlew, D. E., & Williams, A. F. (1964). Interpersonal
and this chapter focused on two specific ap sensitivity under motive arousing conditions. Jo u r n a l
o f A b n o r m a l a n d S o c ia l P sy ch olog y , 6 8 , 1 5 0 - 1 5 9 .
proaches that have been integrated in the hi
Birney, R. C., Burdick, H ., & Teevan, R. C. (1969).
erarchical model of achievement motivation F e a r o f failu re. New York: Van Nostrand Reinhold.
(Elliot, 1999). Other approaches, such as Brunstein, J. C. (2001). Personliche Ziele und Han-
those that focus on attributions for achieve dlungs- versus Lageorientierung: Wer bindet sich
ment outcomes or implicit theories of intel an realistische und bediirfniskongruente Ziele?
[Personal goals and action versus state orientation:
ligence and ability, also seem to hold great W ho builds a commitment to realistic and need-
promise for explaining social behavior dur congruent goals?]. Z e its c h r ift fu r D iffe r e n tie lle u n d
ing competence pursuits. D ia g n o stisc h e P sy ch o lo g ie , 2 2 , 1 - 1 2 .
One of the challenges in moving this lit Butler, R., & Neuman, O. (1995). Effects of task and
ego achievement goals on help-seeking behaviors
erature forward will be the sheer scope of
and attitudes. J o u r n a l o f E d u c a tio n a l P sy ch olog y ,
possible social behaviors that can be inves 8 7 ,2 6 1 - 2 7 1 .
tigated. The interpersonal circumplex may Cheung, P. C., M a , H. K., & Shek, D. T. L. (1998).
provide a valuable framework for generating Conceptions of success: Their correlates with proso
hypotheses and organizing findings in this cial orientation and behaviour in Chinese adoles
cents. J o u r n a l o f A d o le s c e n c e , 21, 3 1 - 4 2 .
complex domain. Not every social behavior Cialdini, R. B., Trost, M . R., & Newsom, J . T. (1995).
is neatly captured by this model, and we do Preference for consistency: T he development of a
not advocate limiting investigations to inter valid measure and the discovery of surprising be
personal behaviors alone. Nevertheless, we havioral implications. J o u r n a l o f P erso n a lity a n d
S o c ia l P sy ch o lo g y , 69, 3 1 8 - 3 2 8 .
encourage researchers to consider ways in
26. A ch iev em en t M otivation 397
Cohen, R. J ., & Teevan, R. C. (1974). Fear of failure tion and achievement goals. E d u c a tio n a l P s y c h o lo
and impression management: An exploratory study. gist, 3 4 , 169-189.
P sy c h o lo g ic a l R e p o rts , 3 5, 1332. Elliot, A. J. (2005). A conceptual history of the achieve
Conroy, D. E., 8c Elliot, A. J. (20 04 ). Fear of failure ment goal construct. In A. J. Elliot &c C . S. Dweck
and achievement goals in sport: Addressing the (Eds.), H a n d b o o k o f c o m p e te n c e a n d m o tiv a tio n
issue of the chicken and the egg. A n x iety , Stress, (pp. 5 2 - 7 2 ) . New York: Guilford Press.
a n d C o p in g , 17, 2 7 1 - 2 8 5 . Elliot, A. J ., 8c Church, M . A. (1997). A hierarchical
Conroy, D. E., Elliot, A. J., & Pincus, A. L. (2009). The model of approach and avoidance achievement m o
expression of achievement motives in interpersonal tivation. J o u r n a l o f P erson ality a n d S o c ia l P sy c h o l
problems. J o u r n a l o f P erson ality , 77, 4 4 5 - 5 2 6 . ogy, 7 2, 2 1 8 - 2 3 2 .
Conroy, D. E., Metzler, J. N., 8c Willow, J . P. (2002). Elliot, A. J . , 8c Dweck, C. S. (2 005). Competence and
Multidimensional fear of failure measurement: The motivation: Competence as the core of achieve
Performance Failure Appraisal Inventory. J o u r n a l ment motivation. In A. J. Elliot & C. S. Dweck
o f A p p lie d S p o rt P sy c h o lo g y , 14, 7 6 - 9 0 . (Eds.), H a n d b o o k o f c o m p e te n c e a n d m o tiv a tio n
Dapra, R. A., Zarrillo, D. L., Carlson, T. K., 8c Te (pp. 2 - 1 2 ) . New York: Guilford Press.
evan, R. C. (1985). Fear of failure and indices of Elliot, A. J ., 8c Harackiewicz, J. M . (1996). Approach
leadership utilized in the training of R O T C cadets. and avoidance achievement goals and intrinsic m o
P sy c h o lo g ic a l R e p o r ts , 5 6 , 2 7 - 3 0 . tivation: A meditational analysis. J o u r n a l o f P er
Darnon, C., Muller, D., Schrager, S. M ., Pannuzzo, N., so n a lity a n d S o c ia l P sy ch o lo g y , 70, 9 6 8 - 9 8 0 .
&C Butera, F. ( 2 0 06 ). Mastery and performance goals Elliot, A. J ., 8c McGregor, H. A. (1999). Test anxiety
predict epistemic and relational conflict regulation. and the hierarchical model of approach and avoid
J o u r n a l o f E d u c a tio n a l P sy ch o lo g y , 9 8 , 7 6 6 - 7 7 6 . ance achievement motivation. J o u r n a l o f P ers o n a l
deCharms, R ., Morrison, H. W., Reitman, W., 8c ity a n d S o c ia l P sy ch olog y , 76, 6 2 8 - 6 4 4 .
McClelland, D. C. (1955). Behavioral correlates Elliot, A. J ., 8c McGregor, H. A. (2001). A 2 x 2
of directly and indirectly measured achievement achievement goal framework. J o u r n a l o f P e r s o n a l
motivation. In D. C. McClelland (Ed.), S tu dies in ity a n d S o c ia l P sy ch olog y , 8 0 , 5 0 1 - 5 1 9 .
m o tiv a tio n (pp. 4 1 4 - 4 2 3 ) . New York: Appleton- Elliot, A. J . , & Murayama, K. (20 08 ). On the measure
Century-Crofts. ment of achievement goals: Critique, illustration,
Deci, E. L., & Ryan, R. M . (1985). In trin sic m o tiv a and application. J o u r n a l o f E d u c a tio n a l P sy ch o l
tion a n d s e lf-d e te r m in a tio n in h u m a n b eh a v io r . ogy, 100, 6 1 3 - 6 2 8 .
New York: Plenum Press. Elliott, FL S., & Dweck, C. S. (1988). Goals: An ap
Diener, C. I., & Dweck, C. S. (1978). An analysis of proach to motivation and achievement. J o u r n a l o f
learned helplessness: Continuous changes in per P erson ality a n d S o c ia l P sy ch olog y , 5 4 , 5 - 1 2 .
formance, strategy, and achievement cognitions Feather, N. (1965). The relationship of expectation of
following failure. J o u r n a l o f P erso n a lity a n d S o cia l success to achievement and test anxiety. J o u r n a l o f
P sy ch olog y , 3 6 , 4 5 1 - 4 6 2 . P erson ality a n d S o c ia l P sy ch olog y , 1, 1 1 8 - 1 2 6 .
Diener, C. I., 8c Dweck, C. S. (1980). An analysis of Gilbert, P., & McGuire, M. T. (1998). Shame, status,
learned helplessness: II. The processing of success. and social roles: Psychobiology and evolution. In P.
J o u r n a l o f P erson ality a n d S o c ia l P sy ch o lo g y , 39, Gilbert 8c B. Andrews (Eds.), S h a m e: I n te r p e r s o n a l
940-952. b eh a v io r , p s y c h o p a th o lo g y , a n d cu ltu re (pp. 9 9 -
Doise, W., 8c Mugny, G. (1984). T h e s o c ia l d e v e lo p 125). New York: Oxford University Press.
m en t o f th e in tellect. Oxford, UK: Pergamon Press. Hagtvet, K., 8c Benson, J. (1997). The motive to avoid
Duda, J. L., Olson, L. K., 8c Templin, T. J. (1986). The failure and test anxiety responses: Empirical sup
relationship of task and ego orientation to sports port for integration of two research traditions.
manship attitudes and the perceived legitimacy of A n x iety , S tress, a n d C o p in g , 10, 3 5 - 5 7 .
injurious acts. R e se a rc h Q u a rterly fo r E x e r c is e a n d Harter, S. (1983). Developmental perspectives on the
S p o rt, 6 2 , 7 9 - 8 7 . self-system. In P. H. Mussen (Series Ed.) 8c E. M.
Dweck, C. S. (1975). The role of expectations and at Hetherington (Vol. Ed.), H a n d b o o k o f c h ild p sy
tributions in the alleviation of learned helplessness. c h o lo g y : Vol. 4. S o c ia liz a tio n , p e r s o n a lity a n d s o
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o lo g y , 31, cia l d e v e lo p m e n t (4th ed., pp. 2 7 5 - 3 8 6 ) . New York:
674-685. Wiley.
Dweck, C. S. (1986). Motivational processes affecting Heckhausen, H. (1963). H o ffn u n g u n d F u rch t in d er
learning. A m er ic a n P sy ch o lo g ist, 4 1, 1 0 4 0 - 1 0 4 8 . L eistu n g s m o tiv a tio n [ H o p e a n d fe a r c o m p o n e n ts
Dweck, C. S. (1999). S e lf-th e o rie s: T h eir r o le in m o t i o f a c h ie v e m e n t m o tiv a tio n /. Meisenheim am Gian,
v a tio n , p e r so n a lity , a n d d e v e lo p m e n t. Philadelphia: Germany: Anton Hain.
Psychology Press. Heckhausen, H. (1984). Emergent achievement behav
Dweck, C. S., 8c Elliott, E. S. (1983). Achievement m o ior: Some early developments. In J. Nicholls (Ed.),
tivation. In P. H. Mussen (Series Ed.) & E. M . Heth- A d v a n ces in m o tiv a tio n a n d a c h ie v e m e n t : T h e
erington (Vol. Ed.), H a n d b o o k o f c h ild p sy c h o lo g y : d e v e lo p m e n t o f a c h ie v e m e n t m o tiv a tio n (Vol. 3,
Vol. 4. S o c ia liz a tio n , p er so n a lity , a n d s o c ia l d e v e l pp. 1 -3 2). Greenwich, CT: JA I Press.
o p m e n t (4th ed., pp. 6 4 3 - 6 9 1 ) . New York: Wiley. Herman, W. (1990). Fear of failure as a distinctive
Klliot, A. J. (1997). Integrating the “classic” and “con personality trait measure of test anxiety. Jo u r n a l
temporary” approaches to achievement motivation: o f R e se a rc h a n d D ev e lo p m e n t in E d u c a tio n , 2 3 ,
A hierarchical model of approach and avoidance 180-185.
achievement motivation. In M. Maehr 8c P. Pintrich Hoigaard, R., 8c Ommundsen, Y. (20 07 ). Perceived so
(Eds.), A d v a n c es in m o tiv a tio n a n d a c h ie v e m e n t cial loafing and anticipated effort reduction among
(Vol. 10, pp. 143-17 9). Greenwich, CT: JA I Press. young football (soccer) players: An achievement goal
Elliot, A. J . (1999). Approach and avoidance motiva perspective. P sy c h o lo g ic a l R e p o rts , 1 0 0 , 8 5 7 - 8 7 5 .
398 V. M O T IV A T IO N A L D IS P O S IT IO N S
House, R. J ., & Singh, J . V. (1987). Organizational McClelland, D. C., Atkinson, J. W., Clark, R. A., 8c
behavior: Some new directions for I/O psychology. Lowell, E. L. (1953). T h e a c h ie v e m e n t m o tiv e. East
A n n u al R e v ie w o f P sy ch olog y , 3 8 , 6 6 9 - 718. Norwalk, CT: Appleton-Century-Crofts.
Jackson, D. N. (1974). M a n u a l f o r th e P erson ality McClelland, D. C., Koestner, R ., 8c Weinberger, J.
R e se a rc h F orm . Goshen, NY: Research Psychology (1989). How do self-attributed and implicit motives
Press. differ? P sy c h o lo g ic a l R e v iew , 9 6 , 6 9 0 - 7 0 2 .
Karabenick, S. A. (2003). Seeking help in large college McGregor, H. A., 8c Elliot, A. J . (2005). The shame of
classes: A person-centered approach. C o n t e m p o failure: Examining the link between fear of failure
rary E d u c a tio n a l P sy ch olog y , 2 8 , 3 7 - 5 8 . and shame. P erson ality a n d S o c ia l P sy ch o lo g y B u l
Kiesler, D. J. (1996). C o n te m p o r a r y in te rp e rs o n a l letin , 3 1 ,2 1 8 - 2 3 1 .
th e o r y a n d re se a r c h : P erson ality , p s y c h o p a th o lo g y , Middleton, M. J . , & Midgley, C. (1997). Avoiding the
a n d p sy c h o th e ra p y . New York: Wiley. demonstration of lack of ability: An underexplored
Koestner, R., 8c McClelland, D. C. (1990). Perspec aspect of goal theory. J o u r n a l o f E d u c a tio n a l P sy
tives on competence motivation. In L. A. Pervin c h o lo g y , 89, 7 1 0 - 7 1 8 .
(Ed.), H a n d b o o k o f p e r s o n a lity : T h e o r y a n d r e Miller, L. C., Murphy, R., 8c Buss, A. H. (1981). C on
s e a r c h (pp. 5 2 7 - 5 4 8 ) . New York: Guilford Press. sciousness of body: Private and public. J o u r n a l o f
Latane, B., Williams, K., & Harkins, S. (1979). Many P erson ality a n d S o c ia l P sy ch o lo g y , 4 1, 3 9 7 - 4 0 6 .
hands make light the work: T he causes and con Mischel, W. (1968). P erson ality a n d a s se ss m e n t. New
sequences of social loafing. J o u r n a l o f P erson ality York: Wiley.
a n d S o c ia l P sy c h o lo g y , 37, 8 2 2 - 8 3 2 . Mischel, W., 8c Shoda, Y. (1995). A cognitive-affective
Leary, T. (1957). In te r p e r s o n a l d ia g n o s is o f p e r s o n a l system theory of personality: Reconceptualizing sit
ity. New York: Oxford University Press. uations, dispositions, dynamics, and invariance in
LePine, J. A. (2 005). Adaptation of teams in response personality structure. P s y c h o lo g ic a l R e v iew , 1 0 2 ,
to unforeseen change: Effects of goal difficulty and 246-268.
team composition in terms of cognitive ability and Moller, A. C., 8c Elliot, A. J . ( 2 0 06 ). The 2 x 2 achieve
goal orientation. J o u r n a l o f A p p lied P sy ch olog y , ment goal framework: An overview of empirical re
9 0 , 11 5 3 -1 1 6 7 . search. In A. V. Mitel (Ed.), F ocu s on ed u c a tio n a l
Levy, I., Kaplan, A., 8c Patrick, H. (200 4). Early ad o p s y c h o lo g y (pp. 3 0 7 - 3 2 6 ) . Hauppauge, NY: Nova
lescents’ achievement goals, social status, and atti Science.
tudes towards cooperation with peers. S o c ia l P sy Monte, C. F., 8c Fish, J. M . (1987). T he fear-of-failure
c h o lo g y o f E d u c a tio n , 7, 1 2 7 -1 5 9. personality and academic cheating. In R. Schwarzer,
Lewin, K., Dembo, T., Festinger, L., 8c Sears, P. S. H. M . van der Ploeg, 8c C. D. Spielberger (Eds.), A d
(1944). Level of aspiration. In J. M . Hunt (Ed.), v an ces in test a n x ie ty re se a r ch (Vol. 6, pp. 8 7-1 03 ).
P erson ality a n d th e b e h a v io r d is o r d e r s : A h a n d Amsterdam: Swets 8c Zeitlinger.
b o o k b a s e d o n e x p e r im e n ta l a n d c lin ic a l resea rch Murray, H. A. (1938). E x p lo r a tio n s in p erso n a lity .
(pp. 3 3 3 - 3 7 8 ) . New York: Ronald Press. New York: Oxford University Press.
Lewis, H. B. (1971). S h a m e a n d gu ilt in n eu rosis. New Murray, H. A. (1943). T h e m a tic A p p e r c e p tio n Test.
York: International Universities Press. Cambridge, M A : Harvard University Press.
Lewis, M ., Alessandri, S. M ., 8c Sullivan, M . W. Nelson-Le Gall, S. (1985). Help-seeking behavior in
(1992). Differences in shame and pride as a func learning. In E. W. Gordon (Ed.), R e v ie w o f resea rch
tion of children’s gender and task difficulty. C h ild in ed u c a tio n (Vol. 12, pp. 5 5 - 9 0 ) . Washington, DC:
D e v e lo p m e n t, 6 3 , 6 3 0 - 6 3 8 . American Educational Research Association.
Lewis, M ., Sullivan, M . W., Stanger, C., 8c Weiss, M. Nicholls, J ., Patashnick, M ., Cheung, P., Thorkildsen,
(1989). Self-development and self-conscious em o T., 8c Lauer, J . (1989). Can achievement motivation
tions. C h ild D e v e lo p m e n t, 6 0 , 1 4 6 - 1 5 6 . succeed with only one conception of success? In F.
Lifshitz, M . (1974). Achievement motivation and co p Halisch 8c J. Van der Bercken (Eds.), In te r n a tio n a l
ing behavior of normal and problematic preadoles p er sp e c tiv e s on a c h ie v e m e n t m o tiv a tio n (pp. 187—
cent kibbutz children. J o u r n a l o f P erso n a lity A s 208). Amsterdam: Swets 8c Zeitlinger.
se ssm en t, 3 8 , 1 3 8 - 1 4 3 . Nicholls, J . G. (1976). Effort is virtuous, but it’s better
Linnenbrink,E. A. (2003). T h e d ilem m a o f p e r fo r m a n c e to have ability: Evaluative responses to perceptions
g o a ls : P ro m o tin g stu d e n ts ’ m o tiv a tio n a n d learn in g of effort and ability. J o u r n a l o f P erson ality a n d S o
in c o o p e r a t iv e g ro u p s. Unpublished doctoral disser c ia l P sy ch o lo g y , 3 1, 3 0 6 - 3 1 5 .
tation, University of Michigan, Ann Arbor. Nicholls, J . G. (1978). T he development of concepts
Linnenbrink, E. A. (2 005). The dilemma of of effort and ability, perception of own attainment,
performance-approach goals: T he use of multiple and the understanding that difficult tasks require
goal contexts to promote students’ motivation and more ability. C h ild D e v e lo p m e n t, 4 9 , 8 0 0 - 8 1 4 .
learning. J o u r n a l o f E d u c a tio n a l P sy ch olog y , 97, Nicholls, J . G. (1984). Achievement motivation: C o n
197-213. ceptions of ability, subjective experience, task
Maehr, M. L., 8c Nicholls, J. G. (1980). Culture and choice, and performance. P sy c h o lo g ic a l R e v iew ,
achievement motivation: A second look. In N. W a r 91, 3 2 8 - 3 4 6 .
ren (Ed.), S tu d ies in c ro ss-cu ltu ra l p s y c h o lo g y (Vol. O ’Malley, M . N., 8c Schubarth, G. (1984). Fairness
3, pp. 2 2 1 - 2 6 7 ) . New York: Academic Press. and appeasement: Achievement and affiliation m o
Mandler, G., 8c Sarason, S. B. (1952). A study of a n x i tives in interpersonal relations. S o c ia l P sy ch o lo g y
ety and learning. J o u r n a l o f A b n o r m a l a n d S o c ia l Q u arterly , 47, 3 6 4 - 3 7 1 .
P sy ch olog y , 47, 1 6 6 - 1 7 3 . Poortvliet, P. M ., Janssen, O., Van Yperen, N. W., 8c
McClelland, D. C. (1987). H u m an m o tiv a tio n . New Van de Vliert, E. ( 2 00 7). Achievement goals and
York: Cambridge University Press. interpersonal behavior: How mastery and perfor
2 6 . A ch iev em en t M otivation 399
mance goals shape information exchange. P ers o n Stornes, T., &c Ommundsen, Y. (200 4). Achievement
a lity a n d S o c ia l P sy c h o lo g y B u lletin , 3 3 , 1 4 3 5 - goals, motivational climate and sportspersonship:
1447. A study of young handball players. S ca n d in a v ia n
Puffer, S. M. (1987). Prosocial behavior, noncompliant J o u r n a l o f E d u c a tio n a l R esea rc h , 4 8 , 2 0 5 - 2 2 1 .
behavior, and work performance among commis Teevan, R. C., Diffenderfer, D., Sc Greenfield, N.
sion salespeople. J o u r n a l o f A p p lied P sy ch olog y , (1986). Need for achievement and sociometric sta
72, 6 15-621. tus. P s y c h o lo g ic a l R e p o r ts , 5 8 , 4 4 6 .
Ryan, A. M ., & Pintrich, P. R. (1 997). “ Should I ask Terhune, K. W. (1968). Motives, situation and in
for help?”: The role of motivation and attitudes in terpersonal conflict within prisoners’ dilemma.
adolescents’ help seeking in math class. J o u r n a l o f J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 8,
E d u c a tio n a l P sy ch olog y , 8 9, 3 2 9 - 3 4 1 . 1-24.
Schmalt, H.-D. (1999). Assessing the achievement m o Thrash, T. M ., Sc Elliot, A. J. (2001). Delimiting
tive using the grid technique. J o u r n a l o f R e se a rc h in and integrating achievement motive and goal con
P erson ality , 3 3 , 1 0 9 - 1 3 0 . structs. In A. Efklides, J. Kuhl, & R. M . Sorrentino
Schultheiss, O. C. (2001). M an u al f o r th e a s se ss m e n t (Fids.), T ren ds a n d p r o s p e c ts in m o tiv a tio n research
o f h o p e o f su c c ess a n d fe a r o f failu re. Unpublished (pp. 3 - 2 1 ). Boston: Kluwer.
manuscript, University of Michigan. Thrash, T. M ., & Elliot, A. J. (2 002). Implicit and
Schultheiss, O. C. (200 7). A memory-systems approach self-attributed achievement motives: Concordance
to the classification of personality tests: Comment and predictive validity. J o u r n a l o f P erson ality , 70,
on Meyer and Kurtz (200 6 ). J o u r n a l o f P erson ality 729-755.
A sse ss m e n t, 89, 1 9 7 - 2 0 1 . Thrash, T. M ., Elliot, A. J ., & Schultheiss, O. C.
Schultheiss, O. C., & Pang, J. S. ( 2 0 07 ). Measuring (2007). Methodological and dispositional predic
implicit motives. In R. W. Robins, R. F. Fraley, & R. tors of congruence between implicit and explicit
Krueger (Fids.), H a n d b o o k o f re se a r ch m e t h o d s in need for achievement. P erson ality a n d S o c ia l P sy
p e r s o n a lity p s y c h o lo g y (pp. 3 2 2 - 3 4 4 ) . New York: c h o lo g y B u lletin , 3 3 , 9 6 1 - 9 7 4 .
Guilford Press. Thrash, T. M., 8c Hurst, A. (2008). Approach and
Shelton, J ., & Hill, J. P. (1969). Effects of cheating on avoidance motivation in the achievement domain:
achievement anxiety and knowledge of peer perfor Integrating the achievement motive and achievement
mance. D e v e lo p m e n ta l P sy ch olog y , 1, 4 4 9 - 4 5 5 . goal traditions. In A. J. Klliot (Ed.), H a n d b o o k o f
Singh, S. (1992). Hostile Press measure of fear of fail a p p r o a c h a n d a v o id a n c e m o tiv a tio n (pp. 2 1 5 - 2 3 1 ) .
ure and its relation to child-rearing attitudes and New York: Psychology Press.
behavior problems. J o u r n a l o f S o c ia l P sy ch olog y , Tracy, J . T., Robins, R. W„ Sc Tangney, J . P. (2 007).
1 3 2 ,3 9 7 - 3 9 9 . T h e se lf-c o n s c io u s e m o tio n s : T h e o r y a n d research .
Smith, C. P. (1992). M otiv atio n a n d p e r s o n a lity : H a n d New York: Guilford Press.
b o o k o f th em a tic c o n te n t an aly sis. Cambridge, UK: Weiner, B., Sc Kukla, A. (1970). An attributional anal
Cambridge University Press. ysis of achievement motivation. J o u r n a l o f P e r s o n
Snyder, M ., & Gangestad, S. (1986). On the nature of a lity a n d S o c ia l P sy ch olog y , 15, 1 - 2 0 .
self-monitoring: Matters of assessment, matters of Wicker, A. W. (1969). Attitudes versus actions: T he re
validity. J o u r n a l o f P erson ality a n d S o c ia l P sy c h o l lationship of verbal and overt behavioral responses
og y , S I, 1 2 5 - 1 3 9 . to attitude objects. J o u r n a l o f S o c ia l Issu es, 2 5 ,
Sorrentino, R. M . (1973). An extension of theory of 41-78.
achievement motivation to the study of emergent Wiggins, J . S. (1991). Agency and communion as
leadership. J o u r n a l o f P erson ality a n d S o c ia l P sy conceptual coordinates for the understanding and
c h o lo g y , 2 6 , 3 5 6 - 3 6 8 . measurement of interpersonal behavior. In D. Cic-
Sorrentino, R. M ., & Field, N. (1986). Emergent lead chetti 8c W. M . Grove (Eds.), T h in k in g clea rly
ership over time: T he functional value of positive a b o u t p s y c h o lo g y : E ssa y s in h o n o r o f P au l E.
motivation. J o u r n a l o f P erson ality a n d S o c ia l P sy M e eh l (pp. 8 9 - 1 1 3 ) . Minneapolis: University of
c h o lo g y , SO, 1 0 9 1 - 1 0 9 9 . Minnesota Press.
Sorrentino, R. M ., & Sheppard, B. H. (1978). Effects Winter, D. G. (1994). M a n u a l f o r s c o rin g m o tiv e im
of affiliation-related motives on swimmers in in a g ery in ru n n in g tex t (4th ed.). Unpublished manu
dividual versus group competition: A field experi script, University of Michigan, Ann Arbor.
ment. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l Winter, D. G. (1999). Linking personality and “scien
og y , 3 6 , 7 0 4 - 7 1 4 . tific” psychology: The development of empirically
Spangler, W. D. (1992). Validity of questionnaire and derived Thematic Apperception Test measures. In
T A T measures of need for achievement: Two meta L. Gieser 8c M . I. Stein (Eds.), E v o c a tiv e im ag es:
analyses. P sy c h o lo g ic a l B u lletin , 112, 1 4 0 - 1 5 4 . T h e T h e m a tic A p p e rc e p tio n T est a n d th e a r t o f p r o
Spence, J . T., & Helmreich, R. L. (1983). Achievement- je c tio n (pp. 1 0 7 -1 2 4 ). Washington, DC: American
related motives and behaviors. In J. T. Spence (Ed.), Psychological Association.
A c h iev e m en t a n d a c h ie v e m e n t m o tiv es: P s y c h o lo g Wright, A. G. C., Pincus, A. I.., Conroy, D. E., Sc El
ic a l a n d s o c io lo g ic a l a p p r o a c h e s (pp. 1 0 -7 4 ). San liot, A. J. (in press). The pathoplastic relationship
Francisco: Freeman. between interpersonal problems and fear of failure.
Stipek, D., Recchia, S., Sc M cClin tic, S. (1992). Self- J o u r n a l o f P erson ality .
evaluation in young children. M o n o g ra p h s o f the Yamaguchi, R. (2001). Children’s learning groups: A
S o c iety f o r R e se a rc h in C h ild D ev e lo p m e n t, 5 7 (Se study of emergent leadership, dominance, and group
rial No. 226). effectiveness. S m all G r o u p R esea rch , 3 2 , 6 7 1 - 6 9 7 .
CHAPTER 27
• A
Belonging Motivation
M a r k R . Leaky
K r is t in e M . K elly
400
27. B elo n g in g M otivation 401
subscales that measure these four bases of Belonging motivation explicitly involves
affiliation motivation (Leary et al., 2008). values and goals that entail establishing so
This finding may indicate that people seek cial connections with other people. For ex
belonging for the same kinds of reasons that ample, a study of the basic life values that
motivate them to affiliate with others— that are associated with belonging motivation
is, to obtain support, attention, stimulation, showed that people who score high in be
or social comparison information. We sus longing motivation place a greater value on
pect, however, that although people do af friendship, love, and social recognition than
filiate for the four reasons that Hill identi those who score low, outcomes that clearly
fied, the primary reason may be to obtain reflect a focus on one’s social connections.
acceptance and belonging. Similarly, we have unpublished data show
Clearly, people who score high in belong ing that belonging motivation is strongly as
ing motivation tend to be more sociable and sociated with the goal of “pleasing others.”
affiliation motivated than those who score In contrast, belonging motivation is not as
low. But do people high in belonging motiva sociated with the degree to which people
tion actually spend less time alone? The an endorse values that are irrelevant to being
swer appears to be “no,” for Need to Belong accepted by other people, such as the value
scores did not predict the frequency with that they place on inner harmony, freedom,
which participants in one study reported wisdom, and pleasure (Leary et al., 2008).
engaging in solitary behaviors during the Although little research has examined
past week. However, although participants how the values of people who are high in
with higher scores reported being alone just belonging motivation relate to specific inter
as often as those with lower scores, they ex personal behaviors, the available evidence
pressed a stronger dislike for doing things by shows that people who are high in belong
themselves. ing motivation are more attuned to cues that
Taken together, these patterns show that, involve evaluation and rejection, as well as
although belonging motivation correlates to opportunities to connect socially with
with extraversion, sociability, and affiliation other people, than people who are low in be
motivation, it is a distinct construct. Because longing motivation. For example, M orrison,
people cannot easily achieve acceptance and Wheeler, and Smeesters (2007) reported
belonging without affiliating with other small correlations between the need to be
people, one would expect a small to mod long and self-monitoring— the tendency to
erate correlation between belonging motiva monitor and control one’s behavior in order
tion and motives to affiliate and socialize, to behave appropriately to the social context
which is what is found. and to make desired impressions on others.
Similarly, people who scored higher on
the Need to Belong Scale were more accu
Establishing Social Connections
rate in identifying emotional expressions de
At its most basic level, belonging motivation picted in pictures of angry, happy, fearful,
stimulates people to establish and main and sad faces than participants who scored
tain relationships with other individuals lower (Knowles, Gardner, Pickett, & Turn
and groups. Accordingly, one would expect er, 2004). They were also better at interpret
strong belonging needs to be linked with ing paralanguage by accurately recognizing
larger social networks. Our own unpub positive and negative vocal tones, even when
lished data support this hypothesis. Individ the tone of voice was incongruent with the
uals who score high on the Need to Belong valence of the word spoken. Finally, Need
Scale reported having more close friends and to Belong scores were related to greater em
a larger social support network and were pathic accuracy in that people scoring high
more likely to use Facebook (the online com in need to belong were more often correct in
munity) as a social networking tool (Carton, their construal of what another person was
Young, & Kelly, 2 0 0 8 ; Kelly, 2008). Overall, thinking or feeling than those who scored
these data demonstrate that people’s actual lower.
social connections and feelings of belonging In another set of studies, Kelly and her
and acceptance coincide with their belong colleagues examined the relationship be
ing motivation. tween belonging motivation and interper
27. B elo n g in g M otivation 403
sonal perception (Kelly & Tee, 2 0 0 5 , 2 0 0 6 ; need to belong reported experiencing less
Kelly, Tee, & Ferry, 2005). Participants first personal discrimination due to their gender
completed the Need to Belong Scale and then than participants who scored low in need
engaged in various interpersonal perception to belong, while at the same time reporting
tasks. Analyses revealed that belonging mo that, as a group, members of their gender
tivation was positively correlated with ac experienced greater discrimination. This
curacy in identifying intimate relationships difference in judgments of personal versus
between other people, as measured by the group discrimination held even after con
Interpersonal Perception Task. However, trolling for individual differences in stigma
no differences were found between people consciousness, gender identity, and public
who scored low versus high in belonging collective self-esteem. According to Carvallo
motivation in identifying other types of re and Pelham, these patterns suggest that “the
lationships, such as those based on kinship, drive for social acceptance colors people’s
status, or competition. In subsequent stud judgments of others in ways consistent with
ies (Kelly & Tee, 2 0 0 6 ), college students en the belief that one will not be subject to in
gaged in a 1-minute social interaction with terpersonal rejection” (p. 103). People high
a stranger and then completed personality in the need to belong do not like to think of
measures for both their interaction partners themselves as being personally discriminat
and themselves. Results indicated that par ed against and thus may downplay or ignore
ticipants high in belonging motivation were indications that they have been devalued or
more accurate at judging conscientiousness mistreated.
in others, an interesting finding given that Given that people often cooperate with
people are typically better at judging more others as a way to gain approval and accep
visible traits, such as extraversion, that are tance and that noncooperators are often dis
easily observed during social interactions liked and rejected (Danheiser & Graziano,
(Funder & Dobroth, 1987). In another study 1982), one might expect that people who are
of interpersonal perception, Kelly, Tee, and high in belonging motivation cooperate more
Ferry (2005) found that participants who than those who are low. In a study that tested
were high in belonging motivation more ac this hypothesis, De Cremer and Leonardelli
curately identified people who told lies ver (2003) found that participants who scored
sus truths (on videotape) than those who high on the Need to Belong Scale cooperated
were low in belonging motivation. Overall, more than those low in the need to belong in
the results of studies of interpersonal per a public-goods dilemma that involved split
ception reveal that belonging motivation is ting money between themselves and a group,
associated with patterns of social sensitivity but only when they believed that they were
that may be particularly relevant to one’s ac participating in a large, eight-person group.
ceptance by others. People high in belonging When they thought they were in a smaller,
motivation were more accurate in identify four-person group, belonging motivation
ing close relationships (but not those based was not related to participants’ contribu
on kin, status, or competition), as well as tions. The authors explained this pattern
more proficient at detecting deception and by suggesting that because larger groups
judging other people’s conscientiousness, afford fewer opportunities for the kinds of
characteristics that are especially important interactions that promote acceptance and
in close relationships. belonging, participants relied more heavily
Paradoxically, although people high in on cooperation to increase their acceptance
belonging motivation are generally more ac in larger groups. Interestingly, Need to Be
curate in decoding others’ expressions and long scores were also related to the decision
paralanguage, they may nonetheless under al frustration that participants experienced
estimate rejection cues in certain situations. when trying to decide how much money
Consistent with other research showing that to contribute to the group versus keep for
people’s motives, goals, and anxieties some themselves, possibly because they were more
times bias their interpretations of social cues conflicted by the tradeoff between personal
(Hilton & Darley, 1991; Stevens & Fiske, gain and social belonging.
1995), Carvallo and Pelham (2006) found Individual differences in belonging moti
that participants who scored high in the vation also have implications for the degree
404 V. M O T IV A T IO N A L D IS P O S IT IO N S
to which people are responsive to the goals Indirect M eans o f F eeling Accepted
that other people have for them. M orrison
People cannot always obtain the degree of
and colleagues (2007, Study 3) found that
acceptance and belonging that they desire,
among participants who had low achieve
either because opportunities for acceptance
ment aspirations but whose mothers had
are not currently available or because they
achievement goals for them, those who were
have been explicitly rejected. In such in
high in belonging motivation performed
stances, people may use tactics that make
better on an achievement-relevant task after
them feel accepted even when actual accep
being primed to think about their mothers.
tance is unavailable. Research suggests that
Those who scored low in the need to belong,
in contrast, did not perform better after people who are high in belonging motivation
use such tactics more commonly than those
being primed. To say it differently, scores
on the Need to Belong Scale correlated posi who are low in it.
tively with performance in the mother-prime Some people seem to derive emotional
condition but not in the control-prime con benefits from the parasocial relationships
dition. Thus people high in belonging moti that they have with actors, newscasters,
vation may be especially willing to strive to and celebrities that they see on television.
meet the goals that significant others have Research on parasocial relationships shows
for them. that T V viewers regard favorite television
Another indication that belonging moti performers as emotionally closer to them
vation leads people to seek acceptance is the than an acquaintance but not as close as a
finding that belonging motivation is associ friend (Koenig & Lessan, 1985), reflecting a
ated with the degree to which people value notable degree of interpersonal connection.
having wealth and material possessions, Theorists have assumed that people form
presumably because they believe that money parasocial relationships with public figures
and possessions increase their chances of to fill unmet social needs and reduce loneli
social acceptance. Indeed, scores on both ness (Koenig & Lessan, 1985; Rubin, Perse,
an ad hoc measure of belonging motivation & Powell, 1985), but little is known regard
and the Need to Belong Scale correlated ing when people use parasocial relationships
with materialist values and with the belief to bolster feelings of acceptance. A series
that buying material goods facilitates accep of studies by Knowles and Gardner (2003,
tance. Furthermore, the belief that material 2008) showed that people who scored high
ism promotes acceptance fully mediated the in belonging motivation have closer and
relationship between belonging motivation more intense attachments to their favorite
and materialism (Rose & Dejesus, 2007). television characters, even seeking “social
Along these same lines, people who are support” from television characters, who
high in belonging motivation seem to be keep them company when they are alone.
more likely to engage in risky behaviors that Oddly, people high in belonging motivation
might facilitate social inclusion, approval, were even more likely to seek the company
and acceptance. For example, college stu and support of cartoon characters. They are
dents who scored higher in belonging m o also more likely to endorse God as a source
tivation reported drinking alcohol signifi of social support (Carton et al., 2008), which
cantly more than those who scored lower suggests that they capitalize on the interper
(Mathes, Kelly, &c Carton, 2008). Results sonal elements of a variety of contexts.
further indicated that this effect may occur Similarly, people sometimes satisfy their
because students who use alcohol heavily desire for acceptance indirectly in ways that
are rewarded by being invited to go out and do not actually increase their acceptance by
have a good time, whereas those who ab other people. Gardner, Pickett, Jefferies, and
stain from drinking are less well liked and Knowles (2005) suggested that, in the same
sometimes excluded from social gatherings way that people may snack in order to tide
(Carton, Kelly, Serra, & Mathes, 2008). them over to the next full meal, people who
Whether people who are high in belonging feel inadequately connected may “snack” on
motivation also engage in other risky behav symbolic reminders of their social connec
iors that promote acceptance is an important tions until they can engage in actual sup
topic for future research. portive interactions. Social snacking may
27. B elo n g in g M otivation 405
take the form of rereading letters or e-mail left out and to the tendency to experience
messages from friends and loved ones, remi negative emotions.
niscing about previous times when one was
accepted or loved, daydreaming about sig
Perceived Rejection
nificant others, or looking at photographs
and L ack o f Connections
of family, friends, or romantic partners. Im
portantly, social snacking is more common Leary and colleagues (2008) presented con
among people who score high in belonging siderable evidence that belonging motivation
motivation (Gardner, Knowles, & Jefferies, is not related to the degree to which people
in press). believe that they are accepted and belong.
As we have seen, people fulfill their desire Specifically, no relationships were found be
for acceptance and belonging by seeking ac tween scores on the Need to Belong Scale
ceptance in face-to-face interactions, as well and seven distinct measures of perceived ac
as via surrogate communications (e.g., pho ceptance, received social support, perceived
tographs, e-mails) and real and imagined re belonging, loneliness, alienation, and related
lationships (e.g., with beloved celebrities and constructs (all rs < 1.101). Perhaps most nota
God). We have additional data showing that bly, no relationship between Need to Belong
people remember (or misremember) the na scores and loneliness was found in three dif
ture of their social connections differently, ferent samples (Leary et al., 2 0 0 8 ; Walker,
depending on their belonging motivation. Green, Richardson, & Hubertz, 1996).
In a study of belonging motivation, social Null results are always equivocal, but to
exclusion, and social networking websites, the extent that these particular findings can
Kelly (2008) found that people who are high be trusted, they suggest that people who
in belonging motivation who were primed typically perceive that they lack acceptance,
with a social task (using Facebook) estimat belonging, social support, or adequate social
ed that they had significantly more friends networks do not necessarily score higher in
than those with lower belonging motivation the disposition to desire acceptance and be
and those who were primed with a nonsocial longing. Thus, although state-like desires for
task (using Wikipedia, the online encyclope acceptance may increase when people feel in
dia). Furthermore, after experiencing social adequately accepted at a particular moment
exclusion, people who were high in belong in time (e.g., Baumeister & Leary, 1995), in
ing motivation indicated that they spend dividual differences in belonging motivation
three times more hours per week browsing do not appear to arise from a perceived lack
Facebook than those with lower belonging of social connections.
motivation. Overall, it appears that having
a strong motive to be accepted and belong
F ea r o f Rejection
evokes construals of personal social memo
and O ther Negative Em otions
ries in ways that help to satisfy that motive.
People who are highly motivated to obtain
a particular outcome typically worry more
Social D eprivation that the motive will not be fulfilled than
people who are less motivated to obtain it.
M ost motives tend to become stronger, or Thus we might expect that people high in be
at least more salient, when they remain un longing motivation are more sensitive to and
satisfied. Along these lines, Baumeister and worried about possible rejection and more
Leary (1995) suggested that the degree to likely to experience negative emotions that
which people desire acceptance and belong are linked to interpersonal concerns, such as
ing increases when their need for belonging social anxiety and hurt feelings. However,
is unmet, as does their experience of nega the evidence for such a link is mixed.
tive emotions. Although Baumeister and First, belonging motivation has been
Leary were discussing state-like changes in found to be unrelated to rejection sensitiv
belonging motivation, their analysis raises ity (Downey &C Feldman, 1996). Thus being
the question of whether stable individual motivated by acceptance and belonging
differences in belonging motivation are tied does not relate to the tendency to anxiously
to feeling disconnected, rejected, lonely, or expect rejection in interpersonal situations.
406 V. M O T IV A T IO N A L D IS P O S IT IO N S
However, this conclusion is clouded by the making mistakes and the failure to live up
fact that scores on the Rejection Sensitiv to one’s own expectations (Findley & Kelly,
ity Questionnaire are a function of both 2008b). In each case, these emotions reflect
the expectation that one will be rejected concerns that one may make an undesired
and the degree to which one is concerned impression on others, an impression that
or anxious about being rejected. As a result, might lead to rejection (Leary & Buckley,
we do not know whether fears of rejection, 2 0 0 0 ).
unconfounded with people’s interpersonal Belonging motivation correlates particu
expectations, are related to belonging m o larly highly with the propensity to experi
tivation. ence hurt feelings. One theory of hurt feel
On the face of it, these findings involving ings suggests that people’s feelings are hurt
rejection sensitivity appear to be contradict when they do not believe that others value
ed by data showing that belonging motiva having relationships with them as much as
tion correlates moderately with scores on the they wish (Leary, Springer, Negel, Ansell,
Fear of Criticism and Rejection factor of the & Evans, 1998). Thus people who greatly
Sociotropy-Autonomy Scale (Beck, Epstein, desire acceptance feel more hurt when they
Harrison, & Emery, 1983). However, this perceive that others do not adequately value
finding, too, must be interpreted with cau their relationships.
tion because of uncertainly regarding what
the Fear of Criticism and Rejection subscale
actually measures. Inspection of the items Excessively H igh and Low
from this subscale suggests that they do not Belonging M otivation
predominantly measure “fear” of rejection
and criticism. Rather, they deal with the The desire for social acceptance and be
desire to be liked, attentiveness to signs of longing is presumably a highly adaptive
social approval, feelings of unease and dis characteristic that facilitated survival and
tress when one is uncertain of obtaining reproduction throughout human evolution
approval, and behaviors that people enact (Baumeister & Leary, 1995). Even so, as is
in order to be liked and accepted (such as the case with most motives, excessively low
being nice, trying not to hurt other people’s and high levels of belonging motivation ap
feelings, and doing things to please other pear to be associated with emotional and
people). Thus this subscale appears to assess behavioral difficulties.
thoughts, feelings, and behaviors that are
broadly associated with desiring approval
H igh Belonging Motivation
and acceptance, coupled with concerns re
garding disapproval and rejection. Thus this As noted, people who are very highly moti
subscale appears to measure roughly the vated to be accepted are more likely to expe
same general construct as the Need to Be rience negative emotions when they are not
long Scale, with which it correlates (Leary certain that others will perceive and accept
et al., 2008). them as they would like. Research shows
Although people who are high in belong that people who are high in belonging moti
ing motivation do not perceive that they are vation score higher on measures of audience
generally less accepted than those who are anxiousness, shyness, and embarrassabil
low, they nonetheless show emotional indi ity than those who score low, and they are
cations that they are concerned about accep also more likely to have their feelings hurt
tance. High belonging motivation is associ by other people. More generally, belonging
ated with emotions that reflect concerns with motivation is associated with the tendency
other people’s impressions, evaluations, and to experience negative affect, as reflected in
approval. For example, scores on the Need measures of negative affectivity and neuroti
to Belong Scale correlate with the tendency cism. Although unpleasant, these feelings
to experience social anxiety when speaking are not necessarily dysfunctional unless they
or performing in front of an audience, with become strong enough to interfere with ef
feelings of shyness in social encounters, and fective behavior or to lead people to avoid
with embarrassability (Leary et al., 2008). interactions and relationships in which they
They are also associated with concern over fear rejection.
27. B elo n g in g M otivation 407
Although direct evidence on this point perspective, healthy people should have at
does not exist, we suspect that very high be least a moderate desire for acceptance and
longing motivation is associated with social belonging, particularly given that evolution
anxiety disorder (or social phobia). Social ary pressures would have favored those who
anxiety disorder is characterized by over made every effort to be integrated into their
whelming anxiety, often accompanied by social groups. Even in the modern world, a
excessive self-consciousness, in social situ person who has no desire for acceptance or
ations (American Psychiatric Association, belonging would fare poorly in most social,
1994). Social anxiety disorder can be lim occupational, and romantic pursuits.
ited to only one particular type of situation Evidence that excessively low belonging
(such as speaking in public or interacting motivation may include a dysfunctional side
with members of the other sex) or so broad is provided by the finding that low need to
that a person experiences anxiety in most belong correlates with tendencies toward
of their dealings with other people. At the schizoid personality disorder, which is
core, people with social anxiety disorder are characterized by a chronic pattern of social
concerned about the degree to which others detachment (Leary et al., 2008). People di
will accept or reject them (Leary & Buckley, agnosed with schizoid tendencies find little
20 0 0 ), and to that extent very high belong value in close relationships, do not enjoy
ing motivation can sometimes be debilitat- their relationships, and lack close friends
ing. and confidants other than first-degree rela
Additionally, exceedingly high need to be tives (American Psychiatric Association,
long is associated with decisional procrasti 1994). They also appear indifferent to praise
nation (Findley & Kelly, 2 0 0 8 a , 2008b), pos or criticism from other people, which might
sibly because people fear making “wrong” reflect a lack of concern about being valued
decisions that might lead others to devalue and accepted. To be sure, people with schiz
or reject them. Given that difficulty in mak oid personality disorder typically show a
ing decisions has been linked not only to broad array of problems (for example, they
failure to complete important tasks but also often appear emotionally cold or show flat
to provoking anger and rejection from oth tened affect), but low belonging motivation
ers (Ferrari, 1994), procrastination seems to appears to be among them.
be self-defeating, especially for people who
are highly motivated to affiliate with and be
accepted by others. Conclusions
differences in the broad tendency to seek rejection sensitivity for intimate relationships.
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 70,
acceptance and belonging predict people’s
1327-1343.
goals, thoughts, emotions, and behaviors in Edwards, A. L. (1954). M a n u a l— E d w a rd s P erso n a l
important ways. Even so, we do not doubt P referen ce S c h ed u le. New York: Psychological C o r
that an individual may be notably more mo poration.
tivated to be accepted in some relationships Ferrari, J. R. (1994). Dysfunctional procrastination
and its relationship with self-esteem, interpersonal
and contexts than in others.
dependency, and self defeating behaviors. P ers o n a l
M ost of the content of social psychology is ity a n d In d iv id u a l D iffe r e n c e s , 17, 6 7 3 - 6 7 9 .
predicated on the fact that human beings de Findley, M . B., & Kelly, K. M . (2 0 0 8 a , March). Pro
sire to develop social connections of various crastination as an indicator of inclusionary sta
tus: Delaying work inhibits social connections. In
kinds with one another. Indeed, little else in
J. Ferrari (Chair), R ev ea lin g th e p ro cra s tin a to r's
the field makes sense without the recognition s e lf: S o cia l, p e r so n a lity , co g n itiv e, a n d p e r c e p tu a l
that people have a desire for acceptance and p er sp e c tiv e s. Symposium presented at the annual
belonging. Likewise, individual differences meeting of the Eastern Psychological Association,
in a broad array of cognitive, emotional, Boston.
Findley, M. B., & Kelly, K. M . ( 2 00 8b , May). T h e ro le
and behavioral phenomena appear to reflect o f p e r fe c tio n is m in fu lfillin g th e n e e d to b elon g .
differences among people in the strength of Paper presented at the annual meeting of the M id
their belonging motivation and in the idio western Psychological Association, Chicago.
syncratic ways that they attempt to foster Funder, D. C., & Dobroth, K. M. (1987). Differences
between traits: Properties associated with inter
social acceptance and belonging.
judge agreement. J o u r n a l o f P erso n a lity a n d S o c ia l
P sy ch olog y , 5 2 , 4 0 9 - 4 1 8 .
Gardner, W., Jefferies, V. E., & Knowles, M . L. (in
R eferen ces press). Never alone: The interdependent self as a
buffer from rejection. J o u r n a l o f P erso n a lity a n d
American Psychiatric Association. (1994). D ia g n o stic S o c ia l P sy ch olog y .
a n d sta tistic a l m a n u a l o f m e n ta l d iso rd er s (4th ed.). Gardner, W. L., Pickett, C. L., Jefferies, V., & Knowles,
Washington, DC: Author. M . (2 005). On the outside looking in: Loneliness
Baumeister, R. F., & Leary, M. R. (1995). The need and social monitoring. P erson ality a n d S o c ia l P sy
to belong: Desire for interpersonal attachments as a c h o lo g y B u lletin , 3 1 , 1 5 4 9 - 1 5 6 0 .
fundamental human motivation. P sy c h o lo g ic a l B u l Gifford, R. (1982). Afilliativeness: A trait measure in
letin , 117, 4 9 7 - 5 2 9 . relation to single-act and multiple-act behavioral
Beck, A. T., Epstein, N., Harrison, R. P., & Emery, G. criteria. J o u r n a l o f R e se a rc h in P erson ality , 26,
(1983). D e v e lo p m e n t o f th e S o c io t r o p y - A u to n o m y 128-134.
S c a le: A m ea su re o f p e r s o n a lity fa c t o r s in p s y c h o Hill, C. A. (1987). Affiliation motivation: People who
p a th o lo g y . Unpublished manuscript, University of need people . . . but in different ways. J o u r n a l o f P er
Pennsylvania, Philadelphia. s o n a lity a n d S o c ia l P sy ch olog y , 5 2 , 1 0 0 8 - 1 0 1 8 .
Carton, A. D., Kelly, K. M ., Serra, R. N., & Mathes, Hilton, J . L., & Darley, J. M . (1991). T he effects of
E. W. ( 2 0 0 8 , May). L a s c iv io u s a n d in e b r ia te d : C o l interaction goals on person perception. In M. P.
leg e stu d en t e n fo r c e m e n t o f s e x a n d a lc o h o l n orm s. Zanna (Ed.), A d v a n ce s in e x p e r im e n ta l s o c ia l p s y
Paper presented at the annual meeting of the M id c h o lo g y (Vol. 2 4 , pp. 2 3 5 - 2 6 7 ) . San Diego, CA:
western Psychological Association, Chicago. Academic Press.
Carton, A. D., Young, M . S., & Kelly, K. M. (2008). Jack son, D. N. (1967). P erson ality R e se a rc h F orm
C h a n g es in p e r c e iv e d s o c ia l s u p p o r t a n d q u a lity m a n u al. Goshen, NY: Research Psychologists
o f r e la tio n sh ip s a m o n g fo r m e r ly h o m e le s s p er so n s Press.
rec eiv in g a s s e rtiv e co m m u n ity tre a tm e n t serv ices. Kelly, K. M . (1999). M e a su r em en t a n d m a n ifes ta tio n
Manuscript submitted for publication. o f th e n e e d to b elo n g . Unpublished doctoral disser
Carvallo, M., & Pelham, B. W. (20 06 ). When fiends tation, University of Tennessee, Knoxville, T N .
become friends: The need to belong and perceptions Kelly, K. M . (2008). |Fulfilling belonging needs with
of personal and group discrimination. J o u r n a l o f Facebook], Unpublished raw data.
P erso n a lity a n d S o c ia l P sy ch olog y , 9 0 , 9 4 - 1 0 8 . Kelly, K. M., & Tee, A. J. (20 05 ). [Need to belong and
Cheek, J . M., & Buss, A. H. (1981). Shyness and so interpersonal perception]. Unpublished raw data.
ciability. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l Kelly, K. M ., & Tee, A. J. (200 6 ). [Need to belong and
og y , 4 1 , 3 3 0 - 3 3 9 . personality judgments.] Unpublished raw data.
Danheiser, P. R., & Graziano, W. G. (1982). Self Kelly, K. M ., Tee, A. J . , & Ferry, S. ( 2 0 0 5 , January).
monitoring and cooperation as a self-presentational B elo n g in g n ess a n d th e d e te c tio n o f lies. Paper pre
strategy. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o l sented at the meeting of the Society for Personality
ogy, 42, 497-505. and Social Psychology, New Orleans.
De Cremer, D., & Leonardelli, G. J. (2 003). Co opera Knowles, M . L., & Gardner, W. L. (2 0 0 3 , May). W hen
tion in social dilemmas and the need to belong: The th e Friends a r e y o u r fr ie n d s : P a ra s o c ia l r e la tio n
moderating effect of group size. G r o u p D y n a m ics: sh ip s a m o n g in d iv id u a ls w ith a high n e e d to b elon g .
T h eo r y , R esea rc h , a n d P ractice, 7, 16 8-17 4. Paper presented at the meeting of the Midwestern
Downey, G., & Feldman, S. (1996). Implication of Psychological Association, Chicago.
27. B elo n g in g M otivation 409
Affiliation Motivation
C r a i g A . H il l
perceptions. Advocates of this perspective, systems would account for the differences in
therefore, have called self-reported motiva patterns of correlations that have been found
tions self-attributed motives. Self-attributed for the TAT and self-report questionnaires.
motives are thought to be activated and to
influence behavior when explicit demand is
T h e T A T M easure
strong in a social situation, that is, when a
o f Affiliation Motivation
social incentive is salient. Drawing on con
cepts generated within the learning theory The administration of the TAT, sometimes
tradition, McClelland and his colleagues called the Picture Story Exercise (PSE), in
(1992) proposed that implicit motives reflect volves having people write stories concern
operant tendencies (self-generated behavior ing situations portrayed in four to six pic
focused on attaining rewards), whereas self tures. Scores are assigned to the content of
attributed motives reflect respondent ten the stories by individuals trained in the cod
dencies (reacting to stimuli in the environ ing scheme, which was developed by com
ment). paring responses of individuals who were as
The basis for the distinction between im sumed to be in a state of aroused affiliation
plicit and self-attributed motives is twofold. motivation with the responses of individuals
First, McClelland and his colleagues (1992) who were thought to be in a state in which
have concluded that scores derived from the affiliation motivation was not aroused (At
Thematic Apperception Test (TAT) assessing kinson, Heyns, & Veroff, 1954; Rosenfeld
a specific motive (such as affiliation motiva & Franklin, 1966; Shipley & Veroff, 1952).
tion, achievement motivation, power motiva For example, in the initial studies (Atkinson
tion) do not correlate to a great extent with et al., 1954; Shipley & Veroff, 1952), re
scores on self-report questionnaires that as searchers hypothesized that affiliation mo
sess the same motive. However, Emmons tivation was aroused for a group of college
and McAdams (1991) disputed this conclu men who had just undergone an evaluation
sion. The second basis for the distinction by their peers; this was thought to activate
between implicit and self-attributed motives thoughts and feelings concerning others, as
is that measures of the two types of moti well as to create desires related to positive
vation tend to correlate with different kinds relations with others. The coding system in
of behavior. According to McClelland and volves detecting themes in the stories related
colleagues, “implicit motives predict sponta to concern about initiating, establishing,
neous behavioral trends over time, whereas and maintaining positive social interactions
self-attributed motives predict immediate and relationships. The underlying type of
specific responses to specific situations or relationship in the themes of the story may
choice behavior” (p. 52). The conclusions be characterized as a concern about having
reached by McClelland and his colleagues, friendly interactions (Heyns, Veroff, & At
however, are at odds with those of theorists kinson, 1992).
and researchers who have found stable cor Scores on the TAT affiliation motivation
relations between self-reported dispositional scale are generally sufficiently stable, as in
motivations, including affiliation motiva dicated by test-retest correlations, for rela
tion, with emotions over a 3-year period tively short periods of time. Coefficients for
(Izard, Libero, Putnam, & Haynes, 1993; scores measured approximately a year apart
Wong &c Csikszentmihalyi, 1991b). in two studies were .56 (Lundy, 1985) and
Perhaps the most useful perspective for .66 (Koestner & Franz, 1989). However,
understanding the distinction between mo the correlation for scores compared over a
tivations assessed by the two methods is to 10-year period was only .30 (Koestner &
consider them to be analogous to implicit Franz, 1989). Similarly, internal consistency
and explicit attitudes (Fazio & Olson, 2 0 0 3 ; coefficients are frequently not very high for
Wilson, Lindsey, & Schooler, 2 0 0 0 ). Implic the scoring systems of many of the motive
it and explicit attitudes have been conceived dimensions, including affiliation motivation.
as separate and independent systems within Although not specific to affiliation motiva
the model of dual attitudes (Wilson et al., tion, a review by Entwisle (1972) of evidence
200 0 ). Conceiving of implicit and explicit regarding the reliability of the achievement
motivation as distinct, relatively independent motivation scoring system indicates that co
28. A ffiliation M otivation 413
People with greater affiliation motivation attaining or receiving benefits from the per
are more sensitive to social demands and son and who is undemanding and nonintru-
the reactions of others. They are more likely sive. Being-Love involves joy and satisfaction
to comply with requests (Walker Sc Heyns, in the mere experience of the relationship
1962) and to avoid competitive situations with the other person rather than having an
(McClelland, 1975; Terhune, 1968), and endpoint at which the desire for involvement
they perform more poorly when in compe with the other person is fulfilled and begins
tition (Karabenick, 1977). In general, they to subside. In contrast, Deficiency-Love is
prefer to avoid conflict with others (Exline, concerned with filling a deficit or void in a
1962). Such behaviors are consistent with the person’s life, and involvement with another
interpretation of the TAT affiliation motiva person is sought to obtain the resource that
tion dimension as reflecting a concern about is absent. The intimacy motivation that M c
negative evaluation by others and a fear of Adams (1980, 1982, 1992) has conceptual
not being liked. Furthermore, the dimension ized and assessed through the TAT scoring
may be characterized as general insecurity system is focused on the desire to experience
and dependency. People with high scores Being-Love.
are sensitive to evaluative feedback related The procedure employed in measuring
to their performance on tasks and desire intimacy motivation within McAdams’s per
relationship-relevant feedback rather than spective involves creating stories in response
competence-related feedback in group activ to pictures used within the TAT tradition.
ity settings. They also prefer to work with As with affiliation motivation, the coding
friends rather than with experts (French, scheme was created by comparing responses
1956). Again, this pattern indicates a con of individuals who were assumed to be in
cern about receiving information about their a state of aroused motivation with those
competence; friends are more likely to be of individuals who were in a state in which
considerate of, and concerned about, their intimacy motivation was not aroused. Four
feelings and to avoid conveying unfavorable samples were initially employed to develop
information about their performance. Along the coding scheme. In one, the individuals
these lines, people high in affiliation motiva thought to be in a state of aroused intimacy
tion tend to perform better if the focus is on motivation were those who had just been in
an affiliative outcome. They even obtain bet ducted into a fraternity or sorority “during
ter grades in classes led by supportive, car friendly, joyful celebrations” (McAdams,
ing instructors (McKeachie, 1961). 1992, p. 225). The other samples involved
individuals at a large dance party, dating
couples who had obtained high scores on
Intim acy M otivation as an A lternate a self-report measure of being in love, and
to Affiliation M otivation those who participated in a series of games
and discussions that were designed to pro
Because the TAT affiliation motivation scor mote intimacy (McAdams, 1992).
ing system assesses the desire to avoid rejec Test-retest reliability of TAT intimacy
tion by others, a type of social anxiety, M c motivation scores was reported to be .48 for
Adams (1980) developed a scoring system high school students assessed over a 1-year
intended to focus more directly on a positive period (Lundy, 1985), a moderate level of
desire for contact with others, as originally stability at best. As is the case for all TAT
conceived by Murray (1938). Intimacy moti motive scoring systems, proponents of the
vation is the desire for “a mutually enjoyed, TAT method maintain that traditional tech
reciprocal, egalitarian union” (McAdams, niques for assessing reliability are not rel
1980, p. 135). According to McAdams evant to evaluating TAT scoring systems
(1982, 1992), the theoretical basis for the (Atkinson, 1992).
intimacy scoring system was the conceptu
alization of B eing-Love {B -L ov e ) compared
Correlates o f T A T Intim acy
with D eficiency-Love (D -L ov e ) proposed
Motivation Scores
by Maslow (1954, 1968). Being-Love is the
desire to share emotional communion with In contrast to people who score high in TAT
another person who is not concerned with affiliation motivation, those with high inti
28. A ffiliation M otivation 415
macy motivation scores are not less popular motivation may feel more secure and able to
with their peers. Rather, they are perceived explore what they are like as individuals.
as friendly, sincere, loving, affectionate, and
cooperative (McAdams, 1980; McAdams &
Losoff, 1984; McAdams & Powers, 1981). The Jackson Perspective
People with high intimacy motivation scores
display more nonverbal behaviors that pro Jackson (1 9 8 4 ,1 9 8 9 ) developed a self-report
mote intimacy and positive feelings, such as questionnaire, the Personality Research
eye contact, smiling, and laughter. They are Form (PRF), to measure 20 of the funda
also more likely to include other people in mental motivations or needs proposed by
the way they talk (using more words such Murray (1938). Jackson characterized the
as we and us rather than I and me), as well fundamental motivations as traits, specifi
as more likely to engage all members of a cally motivational dispositions, that influ
group in discussions. Furthermore, individu ence the behaviors in which individuals
als with high intimacy motivation tend to be engage to attain their goals. A number of re
less controlling and directive in group proj searchers have employed the PRF to measure
ects (McAdams & Powers, 1981). individual differences in affiliation motiva
Intimacy motivation as measured by the tion to examine issues relevant to the desire
TAT also relates to positive feelings about for closeness with others, but a systematic
relationships and higher quality interactions. program of research has not been conduct
In an experience-sampling study in which ed within this perspective to document the
participants reported their feelings and be theoretical framework of dispositional affili
haviors throughout the day, greater intima ation motivation.
cy motivation was associated with a greater
percentage of thoughts about other people
T h e M easurem ent o f Dispositional
and relationships, involvement in a greater
Affiliation Motivation
number of conversations, and more posi
tive feelings when around others (McAdams The PRF is a theory-based measure of 20 of
& Constantian, 1983). W ithin friendships, the needs proposed by Murray, comprising
individuals with high intimacy motivation 352 items in the version called PRF-E (that
report engaging in more one-on-one interac is, form E). After the creation of an initial
tions, as well as self-disclosing and listening pool of items, the instrument was developed
more in conversations with others (McAd based on psychometric principles to hone its
ams, Healy, & Krause, 1984). empirical properties and construct validity.
Furthermore, intimacy motivation is as The items consist of statements to which re
sociated with psychological well-being and spondents indicate true or false with respect
adjustment, as well as psychological health to whether a given statement describes their
over long periods of time. In one study, in typical proclivity or behavior. Examples of
timacy motivation scores obtained for Har- statements measuring affiliation motivation
vard men in the early 1950s were strongly- are “I try to be in the company of friends
related to their job satisfaction and happi as much as possible,” “Sometimes I have
ness with marriage 17 years later (McAdams to make a real effort to be social” (reverse
& Vaillant, 1982). Intimacy motivation has scored), and “I spend lots of time visiting
been found in a large study to correlate with friends.”
self-reported satisfaction with women’s roles The PRF-E scales were constructed by nar
in life and general happiness (McAdams & rowing an initial pool of 100 items per scale
Bryant, 1987). Men high in intimacy moti down to 20 items each (Helmes & Jackson,
vation were less likely to report anxiety, sub 1977). Statements were eliminated based on
stance abuse, and psychosomatic symptoms. empirical criteria to reduce social desirabil
High levels of intimacy motivation may ben ity and redundancy, maximize the internal
efit women and men psychologically in dif consistency of each scale, and minimize the
ferent ways, according to McAdams (1989, similarity of items on one scale to those of
1992). Women with a strong desire for in other scales (enhancing discriminant valid
timacy may have a greater sense of identity ity). Factor analysis of the PRF has dem
and self-value, whereas men high in intimacy onstrated extremely high factorial integrity
416 V. M O T IV A T IO N A L D IS P O S IT IO N S
of the instrument in that all but two of the ity/Arrogance and Exploitive/Entitlement
items loaded on the theoretically expected subscales of the Narcissistic Personality
factor and did not load substantially on the Inventory (Raskin & Hall, 1979). The af
other factors. The mean factor loading of af filiation motivation scale was additionally
filiation motivation items on the affiliation negatively associated with the N arcissism -
motivation scale was .39, whereas the mean Hypersensitivity Scale (Serkownek, 1975)
factor loading of these items on other scales and the Ego Sensitivity Scale, both mea
was .09. sures of covert narcissism (Pepper & Strong,
1958).
Although people generally prefer cowork
Correlates o f the P R F Affiliation
ers who are highly affiliative, people with
Motivation Scale
strong affiliation motivation have even
A number of studies employing the PRF in greater preference for working with af
general (as opposed to focusing on the affili filiative colleagues (Tett & Murphy, 2002).
ation motivation scale) are concerned with Furthermore, affiliation motivation af
psychometric issues involving the compari fects a range of issues related to working
son of personality measures with one an with others on tasks. Klein and Pridemore
other and establishing the construct validity (1992) examined the effects of cooperating
of measures. A bibliography available from with another person on a learning project
the Research Psychologists Press, which among a sample of education majors in col
publishes the PRF, attests to the largely psy lege. Participants who scored high in affili
chometric focus of much research (SIGMA ation motivation performed more poorly on
Assessment Systems, 20 0 8 ). Only a few a test of knowledge following the learning
studies have examined affiliation motivation project when they worked on the project by
in particular. themselves compared with working with a
The PRF affiliation motivation scale has partner. The authors suggested that the in
been shown to correlate with the expressivi volvement of other people influences the
ty scale of the Personal Attributes Question task-related motivation of learners with
naire (Spence, Helmreich, & Stapp, 1975), high affiliation motivation. Social interac
a measure of expressiveness or communion tion may make the task more interesting for
(Moneta & Csikszentmihalyi, 2003). Com those who desire close interaction with oth
munion was discussed previously as one ers. As an aside, individuals with low affilia
of two core aspects of personality, the one tion motivation spent more time working on
with which affiliation motivation is strong the project overall and performed better on
ly aligned conceptually and theoretically. the test of knowledge, regardless of whether
Characteristics constituting the measure of they worked alone or with a partner, possi
expressiveness are understanding o f others, bly as a result of more time investment.
aw are o f the feelings o f others, kind, able to In a follow-up study (Klein & Schnack-
devote self com pletely to others, helpful to enberg, 20 0 0 ), participants with low affili
others, gentle, warm in relations to others, ation motivation expressed a greater desire
and em otional. As would be expected, high to work alone on projects in the future,
school students with stronger dispositional whereas those with high affiliation motiva
affiliation motivation as measured by the tion indicated more interest in working with
PRF more often wished to be with friends a partner on future projects. Dyads high in
and less often wanted to be alone (Wong & affiliation motivation also engaged in more
Csikszentmihalyi, 1991a). task-related group behaviors than dyads low
On the other hand, the affiliation motiva in affiliation motivation, although they also
tion scale is negatively related to measures of engaged in more behaviors that were irrel
narcissism, which involves being dominant, evant to task progress. In this study, no ef
feeling a sense of entitlement, possessing in fect was found for affiliation motivation on
flated self-worth, and yet being dependent knowledge attained or on a task involving
on social approval and sensitive to being application of knowledge.
slighted or rejected (Sturman, 2 0 0 0 ). Scores Affiliation motivation is also related to
on the PRF affiliation motivation scale were people’s expectations about a task, even
negatively correlated with the Superior before beginning it. Stronger affiliation mo
2 8 . A ffiliation M otivation 417
tivation was associated with a tendency to levels of both commitment and exploration.
emphasize the affiliative aspects of an up Those with low implicit affiliation motiva
coming project in which individuals were to tion did not tend to differ in terms of identity
teach others about incorporating software status, regardless of their standing in terms
packages in workplace and school settings of explicit affiliation motivation. Explicit
(Griner & Smith, 20 0 0 ). This occurred de affiliation motivation was positively associ
spite the fact that the project was designed ated with identity achievement, although
to involve both achievement-related features this was not the case for implicit motivation.
and affiliation-related features. However, Congruence of implicit and explicit motives
people with stronger affiliation motiva may permit people to develop a self-concept
tion additionally appraised the task-related and identity that straightforwardly and ac
(achievement) aspects as especially impor curately reflects the characteristics that they
tant and anticipated experiencing greater possess.
interest and less boredom during the project Congruence of motivations between peo
than participants who were low in affiliation ple are likewise important with respect to in
motivation. terpersonal relationships. Meyer and Pepper
Affiliation motivation as measured by the (1977) found that married couples charac
PRF predicted improvement in psychological terized by greater adjustment were those in
symptoms of 33 inpatients over the course which the individuals were more similar to
of a 3-week group psychotherapy program one another with respect to motives for af
(Ratto & Hurley, 1995). Psychological filiation, aggression, autonomy, and nurtur-
symptoms— as indicated by two measures— ance. No evidence was found to support the
declined over the course of psychotherapy. notion of complementarity, that individuals
Dispositional affiliation motivation was who have opposite types of needs are better
negatively associated with both measures suited for one another in relationships.
of posttest symptoms at a substantial level.
Ratto and Hurley (1995) speculated that in
patients who were high in affiliation moti The Mehrabian
vation may have felt less threatened in the Tw o-M otive Perspective
group therapy setting, contributing to a
heightened sense of security, compared with Mehrabian (1970) developed two measures
patients low in affiliation motivation. Indi that are potentially relevant to understand
viduals low in affiliation motivation may ing affiliation motivation— one designed to
need special assistance in developing trust in assess affiliative tendency (M AFF) and the
such interpersonal situations. other to assess sensitivity to rejection (MSR).
Reprising the issue of implicit and explicit Affiliative tendency is the disposition to have
motivation, Hofer and colleagues (Hofer, generalized positive expectations about so
Busch, Chasiotis, & Kiessling, 2 0 0 6 ) dem cial relationships, to anticipate that social
onstrated that congruence between implicit interactions will be pleasant and rewarding,
affiliation motivation and explicit affilia and to engage in behavior based on these
tion motivation was related to individuals’ positive expectations. In contrast, sensitivity
level of identity achievement. Implicit moti to rejection is the disposition to have gen
vation was assessed with the TAT method, eralized negative expectations about social
and explicit motivation was measured with relationships, to be fearful and apprehensive
the PRF. Specifically, people with high lev that social interactions will produce rejec
els of both implicit and explicit motiva tion and pain, and to engage in behavior
tion were more likely to be classified in the based on these negative expectations. The
identity achievement status, as opposed to two dimensions are conceptualized as inde
moratorium, foreclosure, and diffusion sta pendent rather than as opposing tendencies
tuses (M arcia, 1994). Identity achievement (Mehrabian, 1994).
involves having a meaningful commitment However, empirical evidence indicates
to one’s own values and goals after a period that sensitivity to rejection is not an element
of exploration. The other statuses involve of, nor relevant to, affiliation motivation at
lower commitment or less personal explo all. Instead, it is virtually equivalent to dis
ration to achieve the commitment, or lower positional dominance, or possibly power
418 V. M O T IV A T IO N A L D IS P O S IT IO N S
motivation, in the reverse direction. Conse affiliative behaviors during spontaneous so
quently, it is not considered further in this cial interaction (Ksionzky & Mehrabian,
section. 1980; Mehrabian, 1971; Mehrabian & Dia
The M A FF is a self-report questionnaire mond, 1971; Mehrabian & Ksionzky, 1972).
that was developed from an initial pool of This is true under conditions of anxiety-
items drawn from other measures and new provoking social situations, as well, such
items written specifically for the scale. The as when pairs of men and women believed
final items were selected based on several that they would evaluate each other regard
rounds of analysis and modification of items. ing their social attractiveness. In a study that
It consists of 26 items that may be grouped evaluated social behavior in such a setting,
into a number of correlated factors. The people who scored high in affiliative ten
M AFF possesses adequate internal consis dency engaged in more positive affiliative
tency and substantial test-retest reliability behaviors. In the process, they made their
and is only slightly correlated with measures partners more relaxed, were liked more, and
of social desirability (Mehrabian, 1994). elicited more affiliative behaviors from their
partners (Mehrabian & Ksionzky, 1985).
Correlates o f the M A F F
The M A FF correlates with other measures A M ultidim ensional Model
of affiliation motivation, as well as other o f Affiliation M otivation
measures that are conceptually related to
affiliation motivation. These include scales The approaches discussed thus far view af
that measure liking for other people in gen filiation motivation as a monolithic trait, but
eral, belief in the goodness of people, con evidence suggests that people may be moti
cern for the welfare of others, and willing vated to affiliate with others for a number
ness to self-disclose. The scale is also highly of distinct reasons that have implications
correlated with the Jackson PRF affiliation for understanding their behavior. To ad
motivation scale. Scales inversely related to dress this possibility, Hill (1987) proposed
the M A FF assess the inability to experience the multidimensional model of affiliation
pleasure in social relationships, loneliness or motivation, which maintains that motiva
the absence of social contacts, social alien tion for interpersonal closeness and inti
ation, avoidance of social contacts related to macy consists of four different, but related,
discomfort around others, and general anxi desires, specifically the desires for (1) posi
ety. The M A FF is also somewhat correlated tive stimulation, (2) emotional support, (3)
with psychological well-being (Mehrabian, social comparison, and (4) attention. The
1994). rationale for this more detailed view of af
People with higher scores on the M A FF filiation motivation is that understanding
tend to assume that others will have similar of motivation is enhanced by distinguishing
attitudes to theirs (Mehrabian & Ksionzky, among factors that motivate a general class
1971) and to perceive that individuals with of behavior (Buss, 1986; Foa &C Foa, 1974;
whom they have interacted have similar atti Spence & Helmreich, 1983; Veroff, 1986).
tudes (Mehrabian & Ksionzky, 1985). They As Atkinson (1966) observed, “The names
likewise assume that they are compatible of given motives— such as achievement—-are
with a variety of different types of individu really classes of incentives which produce
als (Mehrabian 8c Ksionzky, 1971), possibly essentially the same kind of experience of
as a result of the tendency to perceive others satisfaction. ... The general aim of one class
as similar to them. In the same vein, people of motives . . . is to maximize satisfaction of
who are high in affiliative tendency perceive some kind” (p. 13). Veroff (1986) noted ad
another person who has been portrayed in ditional advantages of proposing a more nu-
an ambiguous way with respect to affiliative anced view of subtypes of motives, such as
tendencies as more affiliative (Solar & M eh documenting developmental trends in which
rabian, 1973). one aspect of a motive increases or decreases
As would be expected of those with stron while other aspects change in different ways
ger affiliation motivation, people who have or remain the same. Additionally, differenti
higher scores on the M A FF engage in more ation among various motive subtypes is like
2 8. A ffiliation M otivation 419
ly to permit more accurate prediction of the and men. (Additionally, a fifth set of items
influence of social situations on behavior. was constructed to represent a dimension of
Murray (1938) and a variety of social psy interpersonal skills to establish that motiva
chologists, as well as related empirical re tion is different from social ability. Factor
search, provide a basis for identifying desire analyses supported this proposal in that the
for positive stimulation, emotional support, interpersonal skill items loaded distinctly on
social comparison, and attention as specific a separate factor in all analyses.) The four
aspects of affiliation motivation (see Hill, scales are moderately correlated with one
1987, for a discussion of the theoretical is another, ranging from .27 to .58. Internal
sues justifying the four proposed motive consistency coefficients (alphas) in the initial
subtypes). study (Hill, 1987) ranged from .70 for the
The positive stimulation aspect of affilia Social Comparison scale for men to .86 for
tion motivation is the desire to obtain pleas the Emotional Support scale for men; most
ant affective and cognitive stimulation from were .78 or above.
contact and interaction with others. This is
the desire to receive gratification from har
Correlates o f the F o u r IO S Scales
monious relationships and a sense of commu
nion. It involves a desire for affection, love, All four scales were correlated with socia
intimacy, and belongingness. The em otion al bility, the tendency to self-disclose during
support aspect of affiliation motivation is interactions with others, emotional vulner
the desire to obtain relief from stressful or ability (being needful of others’ approval),
fearful situations by receiving sympathy, and self-monitoring (the tendency to attend
compassion, and nurturance from others. to social cues and to modify behavior to con
The social com parison aspect of affiliation form with social expectations). The Positive
motivation is the desire to reduce uncertain Stimulation scale and the Emotional Sup
ty, ambiguity, and confusion through ob port scale were especially strongly associ
taining information about others’ behavior, ated with sociability, as well as measures of
attitudes, opinions, and expectations. The expressivity (or communion; Bakan, 1966)
attention aspect of affiliation motivation is and empathic concern (the tendency to expe
the desire to be held in high regard and to re rience the same emotions others are believed
ceive praise and adulation from other people to be experiencing); the Social Comparison
(Hill, 1987). scale and Attention scale were not correlated
with expressivity or empathic concern. On
the other hand, the Social Comparison and
A M easure o f the F o u r Aspects
Attention scales were associated with pub
o f A ffiliation Motivation
lic self-consciousness, the tendency to focus
The Interpersonal Orientation Scale (IOS), on oneself when in the presence of others;
a self-report questionnaire, was designed to the other two affiliation motivation scales
measure each of the four subtypes of affili were not correlated with this dimension
ation motivation (Hill, 1987). The develop (Hill, 1987). All four scales were moderately
ment process involved writing statements positively correlated with sociotropy (Clark,
that reflected the four proposed motive sub- Steer, Beck, & Ross, 1995), the tendency to
types. The response format was a 5-point be extremely committed to the importance
Likert scale with response anchors ranging of positive social interactions (Beck, 1987).
from “Not at all true” to “Completely True” The discriminant validity of the IOS scales
in reference to the extent to which each was established by the absence of correla
statement describes a respondent’s typical tions with measures that were theoretically
feelings. Through a series of analyses of the and conceptually unrelated to affiliation
items, subsequent revision of the wording motivation. These included a measure of
of a few statements, and elimination of oth instrumentality or agency (the counterpart
ers, a final set of 26 items was obtained. The to communion; Bakan, 1966), achievement
proposed model of four affiliation motive motivation, self-esteem, shyness, and social
subtypes was supported by factor analyses desirability (need for approval).
of responses from two large samples. Factor The discriminant validity of each scale
solutions were extremely similar for women was also established in the initial study by
420 V. M O T IV A T IO N A L D IS P O S IT IO N S
Hill (1987) through a procedure in which in support and interpersonal skill on the nature
dividuals imagined how they would behave of interaction between pairs of unacquainted
in four situations described in vignettes. individuals. The interactions focused on per
Each vignette portrayed a situation that is sonal problems that each individual was cur
specifically relevant to one of the affiliation rently experiencing or that were still of con
motivation subtypes but less relevant to the cern to them. The emotional support aspect
other affiliation motivation subtypes. As of affiliation motivation was associated with
predicted, the vignette about a party situ a greater tendency to express understand
ation was most highly associated with the ing, offer encouragement, and compliment
Positive Stimulation scale, whereas the vi their partners, three responses that reflect
gnette about a stressful, frightening situa emotional support (Cohen & W ills, 1985).
tion was most strongly correlated with the Analyses indicated that individuals with a
Emotional Support scale. The vignette about strong need for emotional support provided
a job interview emphasizing one’s capabili greater levels of emotional support to their
ties relative to others was most highly cor partners, regardless of whether the partners
related with the Social Comparison scale. reciprocated or not.
Finally, the vignette about concern over re Highly interpersonally skilled partici
ceiving recognition for one’s contribution on pants were more likely to discuss ways for
a project at work was most strongly associ the partners to deal effectively with emo
ated with the Attention scale. tions related to their problems as well as to
All four aspects of affiliation motivation offer more suggestions, a type of inform a
are correlated with affect intensity (Blank- tional support. Interpersonally skilled indi
stein, Flett, Koledin, & Bortolotto, 1989). viduals engaged in both types of behaviors
Those with stronger affiliative needs have (discussing ways to deal with negative emo
a tendency to experience more intense emo tions and making suggestions) to a greater
tions. This finding supports the position that extent regardless of whether the partners
interaction with others provides a source of reciprocated with similar types of behavior
stimulation that some people desire more that focused on the interpersonally skilled
strongly. individuals’ problems.
One line of research has focused on the af This set of findings indicates that people
filiation motivation component of desire for who desire emotional support are not self-
emotional support because of its direct rel centered or self-indulgent, focusing only
evance to coping with stress and enhancing on their own problems. In fact, this type of
well-being. Individuals with stronger desire individual provided greater emotional sup
for emotional support were shown to express port to his or her partner despite the fact
greater desire to discuss personal problems that some partners did not offer emotional
with another person only when that person support to them. People with strong inter
was warm and compassionate. When the personal skills focused primarily on strate
other person was viewed as not at all warm gies for helping the partners deal effectively
and compassionate, those with strong emo with their negative emotions. This finding
tional support need were extremely averse to indicates that interpersonally skilled indi
talking about their problems (Hill, 1991). In viduals possess the sensitivity, confidence,
contrast, those who were low in the desire and expertise necessary to carry out helping
for emotional support did not distinguish behaviors effectively.
between the two types of confidants, being Dispositional affiliation motivation has
only moderately interested in talking with also been shown to relate to cancer patients’
either person. A similar effect was obtained interest in receiving emotional support from
for the Positive Stimulation scale. The results their spouses, specifically, having their
of the study are consistent with the incentive spouses show concern and provide com
view of motivation in that individuals with fort (Manne, Alfieri, Taylor, & Dougherty,
high motivation are interested primarily in 1999). Affiliation motivation was not associ
interactions that provide desired incentives ated with an interest in receiving instrumen
(such as warmth and support). tal support (providing assistance with treat
A follow-up study (Hill, 1996) examined ment, doing chores), however. This study
the influence of both the desire for emotional examined the joint effects of the Emotional
2 8 . A ffiliation M otivation 421
Support scale, the Positive Stimulation scale, tional support was found for officers with
and the Attention scale in a path analysis lower affiliation motivation.
model.
The dispositional motivation to affiliate
for emotional support is additionally as Gender Differences
sociated with perceptions of the quality of
emotional support that people receive from Evidence regarding gender differences in dis
family members and friends (Hill, 1997). positional affiliation motivation as assessed
Specifically, people with stronger emotional by projective measures is generally incon
support needs were more satisfied with their clusive (Wong & Csikszentmihalyi, 1991a).
support providers, but only if the providers Some studies based on projective measures
were perceived as possessing greater expres have found no differences (Chusmir, 1985;
siveness (communion), such as being warm, Hyland & M ancini, 1985), whereas oth
sensitive, and compassionate. If providers ers have found that women have higher af
were not emotionally expressive, partici filiation motivation than men (Agrawal &
pants expressed less satisfaction than those Upadhyay, 1983; McAdams & Constan-
who had lower motivation to affiliate to ob tian, 1983; McAdams, Lester, Brand, M c
tain emotional support. This pattern is con Namara, & Lensky, 1988; Schroth, 1985).
sistent with the findings of the Hill (1991) Thus a major review of research concerning
study, in that people with a high desire for the TAT measure of affiliation motivation
emotional support are sensitive to and selec led Stewart and Chester (1982) to conclude
tive about the quality of the experiences they that the evidence is inconsistent regarding
have with others. gender differences. Nonetheless, in another
Affiliation motivation in the pursuit of review, Minton and Schneider (1980) con
positive social stimulation also contributes cluded that women obtain higher scores on
to the organizational identification of virtu projective measures. One factor that may
al employees, employees in a company who contribute to inconsistent findings is the fact
work outside of traditional centralized offic that projective measures typically possess
es, such as in their homes. Employees with relatively low internal consistency and test-
a stronger desire for positive social stimula retest reliability (Entwisle, 1972; Koestner
tion report higher levels of organizational & Franz, 1989).
identification. However, even those with a In contrast, studies based on self-report
lower desire for social stimulation express questionnaires of affiliation motivation gen
strong identification when they feel that they erally find gender differences, with women
receive support from colleagues and super scoring higher than men (Minton & Sch
visors (Wiesenfeld, Raghuram, & Garud, neider, 1980; M offitt, Spence, & Goldney,
2001). Other research has demonstrated that 1986; Schroth, 1985; Wong &c Csikszentmi
perceived organizational support— the sense halyi, 1991a). An important caveat to this
that one’s employer values one’s contribu conclusion is that gender differences actually
tions and cares about one’s well-being— occur as a function of the particular aspect
affects work-related performance primarily of affiliation motivation. Within the multi
for those with stronger dispositional affilia dimensional model of affiliation motivation
tion motivation (Armeli, Eisenberger, Fasolo, (Hill, 1987), women usually have higher
& Lynch, 1998). The explanation proposed scores than men on the Positive Stimulation
by the authors for this qualified relationship scale and the Emotional Support scale, al
is that workers who feel valued, and espe though men are typically no different from
cially those who are concerned about social women with respect to scores on the Social
relationships, feel an obligation to repay the Comparison scale and the Attention scale.
organization. In this study, state police patrol The positive social stimulation and emo
officers with stronger desires for emotional tional support aspects are highly similar to
support, attention, and positive stimulation the broader personality dimension of com
who felt supported by their employers made munion or expressivity. The positive social
more arrests for driving under the influence stimulation aspect is most likely the compo
of alcohol and gave more citations for speed nent of affiliation motivation that other, more
ing. No such effect for perceived organiza general measures tap. The social comparison
422 V. M O T IV A T IO N A L D IS P O S IT IO N S
Power Motivation
E u g en e M . Fo d o r
426
29. P o w er M otivation 427
land tradition, and the one researchers have what Woike (1994) characterizes as an emo
studied the most, is need for achievement (n tional “kick,” a surge in pleasurable affect.
Achievement), which translates as a need to For the power-motivated person, feelings of
perform well according to a standard of ex strength, vigor, and energy derive from ac
cellence. tivities prompted by the appropriate activ
An important distinction exists between ity incentive. In partial support of this idea,
PSE and self-report measures of motivation McClelland and colleagues (1989) obtained
(Winter, 1999). PSE measures are inferred, evidence that the act of registering impact
that is, implicit. They are projective tests on others results in the hormonal release of
thought to reflect motives that are largely norepinephrine, which is known to be asso
nonconscious. The research participant does ciated with pleasure. Sometimes an activity
not directly describe the thoughts residing incentive presents itself, but the successful
within but rather tells a story about a pic exercise of power that the power-motivated
ture, from which the rater infers inner mo individual anticipates does not occur. The
tivational dynamics. Self-report inventories, result, according to McClelland (1976),
by contrast, describe what the participant is what he termed pow er stress. Indeed,
believes are the motives that shape his or the mere anticipation of failure to exercise
her intents and actions. They are explicit, power (a G a - in the Winter coding system)
residing squarely within the conscious mind. may be sufficient to cause power stress.
Later I show that these two kinds of mea
sures reflect different aspects of motivation.
Im plicit versus E xp licit Motives
colleague has valued technical skills, only same men, and obtained blood pressure
to meet with a reaction of no interest. Per readings. Men who had shown PSE evidence
sons known to be prone to cardiovascular for the blocked power motive syndrome now
disease, McClelland noted, look in many showed elevated blood pressure readings
ways like persons high in n Power. M ost no compared with men who had then exhibited
tably, they are aggressive and competitively alternative motive patterns. The readings,
striving, seeming to derive satisfaction from on average, were only slightly below the level
outdoing others. Steele (1973) found that a ordinarily regarded as having medical sig
power-arousal manipulation did increase nificance. McClelland, Davidson, Floor, and
epinephrine output as a correlate of sympa Saron (1980) subsequently investigated the
thetic arousal in research participants whose possible role of epinephrine in the blocked
measured n Power increased in response to power motive syndrome, using male prison
the power-arousal manipulation. No such inmates as participants. Those inmates who
effect resulted from an achievement-arousal reported high levels of power stress during
manipulation. The general adaptation syn recent months and who exhibited the blocked
drome (Selye, 1973) is believed to cause car power motive syndrome in their PSE proto
diovascular illness when it is aroused often cols showed high epinephrine concentration
and to a high degree. Epinephrine produc in the urine as compared with inmates with
tion is integrally associated with the general other motive patterns.
adaptation syndrome, as are increased heart Friedman and Rosenman (1974) conduct
rate, increased blood pressure, and glycogen ed extensive studies on what they termed
conversion from the liver into blood glu Type A versus Type B personalities. More
cose. than normal individuals, Type A’s were ir
The McClelland (1979, 1982) hypothesis ritable, always in a hurry, hard driving, and
proposes two separate mechanisms by which tense. Also, they gave evidence of repressed
power stress can occur among persons who anger. Type B’s, by contrast, were much more
are high in n Power. The first is activity inhi relaxed and easygoing, willing to go with
bition — an internal control mechanism that the flow. Friedman and Rosenman found
restricts the outward expression of anger that Type A’s were more likely to experience
and assertiveness. The second is the occur heart attacks than were Type B’s. The “driv
rence of social circumstances that by their en” quality of Type A behavior, they rea
very nature prevent the power motive from soned, produced a more chronically active
expressing itself. An example would be a sympathetic nervous system that was likely
person presenting a strong counterargument to put a strain on the cardiovascular system.
to a position on which the power-motivated Chronic sympathetic activation increases
person has taken a firm stand. Either meth heart rate and releases epinephrine. Both ef
od— one internal, the other external to the fects can damage the cardiovascular system.
self— can thwart the power motive and, ac Suppressed assertiveness and the implica
cording to McClelland, produce cardiovas tion of catecholamine release through the
cular distress if repeated many times. sympathetic nervous system both constitute
McClelland’s (1979) inquiries into the attributes of Type A behavior and thereby
nature of power stress led him to what he connect with aspects of the blocked power
termed the b lock ed pow er m otive syndrom e , motive syndrome (McClelland, 1976, 1985).
namely, high n Power, low n Affiliation, and Power motivation per se, however, does not
high activity inhibition. The role of n Affilia correlate with Type A behavior, nor does n
tion is that it prompts people to seek out the Achievement or n Affiliation (Matthews &
com fort of friends in times of duress, shar Saal, 1978). The power motive must incur
ing their discontent and essentially solicit blockage to its expression for the deleterious
ing the solace and alleviation from anxiety effect on the cardiovascular system to occur
that friends can provide (McClelland, 1985). (McClelland, 1976, 1985).
In an important study of this topic, largely It will be recalled that McClelland (1979)
because it was prospective rather than con hypothesized that power stress could occur
current in its design, McClelland (1979) ex in either of two ways: (1) in the blocked
amined PSE protocols that college men had power motive syndrome, as discussed in the
written 2 0 years previously, sought out these previous paragraphs, and (2) through social
29. P o w er M otivation 431
events that thwart expression of the power Deactivation Adjective Check List at the end
motive. W hat is needed to explore the sec of the power-stress session, higher than su
ond possibility, he wrote, are experiments pervisors low in n Power, and higher than
that present “strong situational challenges” all supervisors in the nonstress sessions. The
(p. 189) to the power motive. I conducted General Activation subscale consists of the
some experiments to pursue M cClelland’s following adjectives: activated, active, en
suggestion. The difference from the M cClel ergetic, full o f pep, lively, peppy, vigorous,
land research was that I regarded n Power w akeful, and w ide aw ake. In validating the
as high if it was in the top third of the over General Activation subscale, Thayer ex
all PSE distribution. For assignment to the posed people to an experimental procedure
blocked power motive syndrome, M cClel that was designed to stimulate physiological
land regarded a T-score of 45 as qualifying arousal. General Activation scores showed
if it was higher than the T-score for n Af a substantial correlation with a composite
filiation (a T-score of 50 is the average for physiological index that combined heart rate
a distribution). The higher cutoff for desig and skin conductance.
nating persons as high in n Power implies In the second industrial simulation ex
a hypothesis that if high enough, power periment, students high or low in power
motivation by itself is sufficient to produce motivation acted as “president” of Modern
power stress given appropriately instigating World Electronics (Fodor, 1985). Presiding
circumstances. over two “managers”— one in marketing
Heeding McClelland’s call for experi and the other in production engineering—
ments that present “strong situational chal the president’s responsibility was to recon
lenges” to the power need, I designed a pair cile any conflict that might arise between the
of industrial simulation experiments using two managers and guide the group toward
male college students as research partici an amicable solution. He further under
pants. The first experiment (Fodor, 1984) stood that his ability to mediate and resolve
simulated for the college student “supervi conflict signified managerial capability. The
sor” a hard-to-manage work crew of high issue concerned whether the company should
school students in the next room. The task, or should not manufacture and market a
ostensibly, was construction of Tinkertoy portable sunlamp. The two managers in the
models from pictured diagrams. Actually, power-stress condition had opposing role
there were no workers. Their production scripts, urging one manager to argue for, the
and voiced comments to the supervisor were other against the proposal. Each had in his
preprogrammed, the voiced comments com script four arguing points for his position.
ing through the intercom system at the end Instructions were that he introduce these ar
of each of the six trials. Worker comments in guments forcibly but not let on that he was
the group-stress condition connoted tension being prompted by the script. Managers as
within the group: concern about not meet signed to the control condition both had role
ing the standard or beating out the compe scripts exhorting them to favor the project.
tition, fear that they would not earn much Electromyographic (EMG) recordings from
money (the supervisor was authorized to the forearm extensor muscle served as the
grant pay increases or decreases), and con stress measure. Presidents high in n Power
cern that they would not be invited back evinced higher EM G readings in the conflict
for another experiment. Instructions to the condition than did presidents low in n Power
supervisor stated that it was his task to use and higher than all presidents placed in the
any and all means at his disposal to improve control condition.
worker performance. His efforts, however, The hypothalamic-pituitary-adrenal (HPA)
were to no avail. Productivity held at rough axis offers clues as to how researchers may
ly the 40th percentile according to the stated effectively examine the physiological conse
norms. The nonstress condition remained at quences of power stress. McClelland (1989)
approximately the 80th percentile through saw cortisol as playing a mediating role in
out, and worker comments were strictly organizing the general adaptation syndrome.
neutral. Student supervisors high in n Power Noting that the HPA axis releases cortisol in
scored distinctly high on the General Activa response to psychological stressors, W irth,
tion subscale of Thayer’s (1978) Activation- Welsh, and Schultheiss (2006) experimen
432 V. M O T IV A T IO N A L D I S P O S IT I O N S
tally induced defeat in a competitive contest. tive rather than n Power or n Affiliation,
They hypothesized that high-power indi McClelland and Winter (1969) described a
viduals are more stressed by a defeat than groundbreaking program for fostering en
low-power individuals and that this should trepreneurial success among businessmen in
manifest itself in increased cortisol release. India. They conducted the initial and most
Specifically, student participants worked on extensive program in two different prov
varying forms of the Number Track Ring inces in India, focusing their efforts toward
Task, a reaction-based cognitive task. H alf inculcating the achievement motive. They
of the participants were led to experience based their design on the extensive research
social defeat, the other half social victory. literature documenting what is known about
Among the various forms that power stress the precursors and correlates of n Achieve
may assume, losing out against another in ment. The businesses of which participants
competition surely ranks high in its impact were a part typically involved only a small
value for the power-oriented person. And so number of employees. Two years after the
the evidence demonstrated. Salivary cortisol training program ended, men trained in the
proved to be high for participants high in n achievement motive attained higher scores
Power who experienced social defeat but not than a control group on a business activity
for those low in the power motive. index. That is, they gave evidence of greater
Winter (1973) found that students high in effort to improve the functioning of their re
n Power chose as close friends students who spective business operations, such as broad
were not well known by other students— ening the product line (e.g., selling women’s
students who posed minimal threat to the saris in addition to bicycles), adding a new
power-motivated students’ assertiveness and salesperson to expand the sales territory, or
desire for control. He also found that power- hiring an accountant to improve bookkeep
motivated men had a preference for unas ing efficiency. In addition, men trained in
sertive, compliant wives. W hat about the n Achievement, more frequently than con
prospect of hiring an assertive, strong-willed trols, created a second business, hired more
subordinate? Would such a person not con people to work for them, and increased busi
stitute a power stress to the power-motivated ness revenue more. A follow-up program on
person, posing a threat to the successful ex a more limited scale further strengthened
ercise of power? We obtained evidence that the emerging impression that n Achievement
this appears to be the case (Fodor et al., was the core ingredient of managerial suc
20 0 6 ). College men imagined themselves in cess. Looking at initial managerial hires at a
terviewing for employment a candidate who major U.S. firm, researchers in the M cClel
came across on video either as strongly asser land group (McClelland & Winter, 1969)
tive (but not unpleasant) or as nonassertive now conducted a version of the program
and compliant. They further were to imagine that controlled for the “Hawthorne effect,”
themselves as that person’s superior should the possibility that all the individual atten
he be hired. Viewing the assertive candidate, tion may have contributed to the success of
high-power participants showed stronger the India program. Managers in the control
EM G readings from the corrugator brow group received instruction in the basic func
(frown) muscle than did participants scoring tions of management (accounting, produc
low in the power motive and stronger read tion, marketing, etc.), thereby matching the
ings than all participants viewing the com group trained in n Achievement for amount
pliant candidate. A scale measuring negative of attention shown to participants. M anag
affect toward the candidate displayed the ers in both groups were tracked over a pe
same pattern and correlated highly with the riod of several years to see who advanced
corrugator EM G readings. most through the managerial hierarchy. The
cadre trained in the achievement motive won
out over the people who learned principles
Leadership and the Pow er Motive of management.
To explore this question further, M cClel
Extrapolating from previous research and land and Burnham (1976) examined sales
then thinking that the essence of leadership division managers at a major corporation.
capability resided with the achievement mo Sales management is a good managerial cat
29. P ow er M otivation 433
egory to study because there are relatively of those divisions high in organizational
unambiguous quantitative measures that clarity and team spirit generally exhibited
one can apply across different divisions to LM P; leaders of divisions low in those cli
determine performance level. McClelland mate dimensions rarely did so.
and Burnham wanted to know whether PSE As a corollary step, McClelland and Boy
profiles for managers of the best-performing atzis (1982) examined managers who partici
divisions differed from those of the worst- pated in a managerial assessment experience
performing divisions. Surprisingly, the when they first came to work for AT&T.
best-performing sales managers were high The PSE was among the many exercises and
in n Power, not n Achievement! Indeed, n tests that were part of the program. M cClel
Achievement was no higher for the best- land and Boyatzis tracked these managers
performing than for the worst-performing over a 16-year period and found that LMP
managers. Moreover, the best-performing correlated with their level of managerial
managers scored low on n Affiliation and placement at the end of that period, whereas
high on activity inhibition. (Activity inhibi those high in n Achievement usually settled
tion is indexed by the number of times the at a lower managerial level. LM P has a num
word not appears throughout a protocol, a ber of features in its favor. People who are
seemingly simple measure that has proven high in power motivation orchestrate the
to have predictive value.) McClelland and performance of others. They are charismatic
Burnham named this constellation of traits (House, Woycke, & Fodor, 1988). Think of
the leadership m otive profile, or LM P (high Jack Welch, the legendary CEO of General
n Power, low n Affiliation, and high activ Electric (Stater, 1999), Winston Churchill,
ity inhibition). The reader will note that this or Franklin Roosevelt. These were people
is the same motive pattern that McClelland endowed with a capacity to inspire others.
and his associates (McClelland, 1979, 1982; Power-motivated managers tend to use oth
McClelland et al., 1980) later found to be ers as a means toward achievement of orga
implicated in power stress. The finding may nizational goals.
have surprised them, because power motiva As McClelland and Burnham (1976) ob
tion conjures up ugly images in the popular served, leadership is an influence game. It re
mind of a tyrannical type of person who is quires manipulative intent, the power of per
insensitive to the needs of subordinates. suasion. Persons high in n Achievement, the
McClelland and Burnham (1976) admin evidence shows, tend to do tasks themselves
istered an organizational climate question rather than make optimal use of others (M c
naire to all of the salespersons in each divi Clelland, 1985). Affiliation motivation de
sion and discovered two climate dimensions tracts from leader effectiveness because it
that appeared to be associated with high places a premium on soliciting the good will
sales performance for a division (i.e., sales and liking of others at the expense of orga
divisions for which LM P characterized the nizational goals. McClelland and Burnham
division managers). These dimensions were emphasized that managers must be univer-
organizational clarity and team spirit. O r salistic in how they make and invoke policy
ganizational clarity means that individual decisions. If managers make exceptions in
employees have a clear definition of what order to cultivate the good will of individual
the organization’s performance expectations employees, overall morale may suffer, erod
are. Team spirit refers to a proclivity to iden ing the all-important organizational climate
tify with the organization, working toward dimensions of organizational clarity and
common goals. The motto of the Three M us team spirit.
keteers— all for one, one for all— captures The significance of high activity inhibition
the essential meaning of team spirit. as an aspect of LM P requires some explana
Because there was no easy way to compare tion. High activity inhibition means control
production divisions against one another ling one’s emotions (“don’t let them see you
(they manufacture entirely different prod sweat”) but also more than that. Persons low
ucts), McClelland and Burnham (1976) ad in activity inhibition too often “say what’s
ministered the organizational climate ques on their minds” and let fly feelings that
tionnaire to all employees within the various cause psychological distress to others and
production divisions. Sure enough, leaders thereby diminish subordinates’ willingness
434 V. M O T IV A T IO N A L D I S P O S IT I O N S
Illustrative of crises that Winter (2007) The practical applications of these findings
examined were two that are familiar to stu are obvious. It would be possible to monitor
dents of American history. One concerned government-to-government communications
United States territorial expansion in the in times of crisis, anticipating escalation or
years 1 8 4 5 -1 8 4 6 , specifically the Mexican even possible movement toward war. Alter
War as it related to U.S. annexation of Texas natively, analysis might reveal a deescalating
and the Oregon boundary dispute with Can temper of mind among conflicting parties,
ada. The M exican War eventuated in mili thereby opening the opportunity for compro
tary confrontation, the Oregon boundary mise and peaceful resolution of differences.
dispute in peaceful compromise. The second Communications from key political figures
crisis was the escalating economic, politi also might contain important clues on what
cal, and social conflict between North and their constituencies can expect from them by
South: the Compromise of 1850 versus the way of aggressive or conciliatory initiatives
Civil War, the latter a truly tragic episode. in dealing with adversarial nations.
Another pair focused on Iraq in Kuwait: The available evidence for small groups
noninvasion in 1961, invasion in 1991. In implies a negative influence from power-
cases of this sort, ratings were applied only motivated leaders. Recall that n Power as
to parties to the conflict, not to peripheral an aspect of the LM P produces a positive
parties or to those who become involved af influence on productivity only in relation
terward. to large aggregates of subordinates, situa
Consistent with prior evidence, power mo tions in which one-on-one interaction with
tivation emerged as the principal predictor subordinates is minimal. The leader in those
of war entry. Surprising at first glance is the instances inspires and motivates from a dis
finding that ratings of responsibility likewise tance, essentially orchestrating the actions
correlated with war entry. W inter’s (2007) of others. We did an experiment that ex
scoring system for responsibility involves plored some implications of Jan is’s (1982)
reference to an abstract moral standard, in concept of groupthink, which he defined as
cluding legality, and fittingness of manners faulty group decision making prompted by
or behavior as appropriate or inappropriate. various social forces that arise within the
Obligation to rules and regulations and sym group (Fodor & Smith, 1982). Janis exam
pathetic concern for others or groups also ined foreign relations fiascos that U.S. gov
figure into the scoring system. By W inter’s ernment and military officials perpetrated in
reasoning, responsibility can take an ugly their advisory group roles at various points
turn in times of crisis when people perceive in history, such as the failure to anticipate
that the other party lacks responsibility and the attack on Pearl Harbor, escalation of
when responsibility demands protection of the Korean War, the Bay of Pigs invasion of
oneself and one’s own group. These cogni Cuba, and escalation of the Vietnam War.
tions can prompt a sense of time urgency, A reading of history (along with W inter’s
coupled with an expressed need to secure the [1987] inaugural-address analyses of the
survival of one’s core values. Power motiva presidents) suggests that the leaders who
tion in combination with high responsibil officiated over the deliberations leading to
ity appears to constitute a potent force that each of these fiascoes were distinctly high in
leads to conflict escalation and war entry but power motivation.
couched in the language of altruistic concern We arranged five-person discussion groups
for one’s own nation. of college students for the imagined purpose
Government communications obviously of determining whether they as managers
constitute a collaborative effort by multiple should recommend the manufacture and
individuals, including speechwriters. W in marketing of a new microwave oven (Fodor
ter (2007) suggested that they reasonably & Smith, 1982). Each “manager” had a role
represent the climate of thinking and value script with six items of information, some
orientation that prevail within an interact items favoring and some opposing the de
ing group at a given time. These communi cision. For example, the following items of
cations can be viewed as proxies that aver information were among those available to
age out prevailing motivational sentiments the group member designated as manager of
that exist among closely interacting persons. production engineering:
436 V. M O T IV A T IO N A L D I S P O S IT I O N S
Merton (1973) wrote of the “race for prior individuals high in n Power generally used
ity” as a key element in the pursuit of scien more power imagery in their recollections of
tific discovery. the happy event (personal strength, control,
An important point is that n Achievement vigor, prestige, and recognition) and indi
also fosters creativity, but the underlying viduals high in n Intimacy wrote memories
mechanism may be different. When they more associated with intimacy imagery (in
have recently experienced negative feedback volving loving, caring, empathy, and close
on a prior effort at creative performance, ness to people) when recalling a happy event.
achievement-motivated persons improve In their questionnaire responses, power-
their creative performance on a subsequent motivated individuals in the pleasant-recall
task, whereas power-motivated persons condition reported more excitement and
show diminished creativity during the sec anger than did individuals high in n Inti
ond task (Fodor &c Carver, 20 0 0 ). This find macy in the same condition. Anger energizes
ing is consistent with research indicating one’s efforts to create impact and therefore
that negative feedback results in improved elicits positive emotions, such as excitement,
task performance in persons who are high in in power-motivated persons (Woike & M c
n Achievement (McClelland, 1985). As W in Adams, 2005).
ter (1973) observed, achievement-motivated Taking account of McClelland’s moti
persons may not like news of failure, but it vational theory (McClelland, 1985; Wein
goads them on to greater effort. In essence, berger & McClelland, 1990), Zurbriggen
it serves as a basis for new information that and Sturman (2002) hypothesized that mo
can modify future performance. Persons tives are linked to specific primary emo
who are high in power motivation, by con tions. They asked people to vividly imagine
trast, may perceive failure as an inability successful experiences at satisfying each of
to impress another person or an aggregate three motives— power, achievement, and
of persons, thereby thwarting their need to affiliation/intimacy— and then report the
register impact and receive acclaim. In other degree to which they experienced various
words, failure to impress may represent a emotional states during the visualizations.
power stress. Winter suggested that in this Instructions for the power visualization
sense the goals of persons high in n Power were to “think about a time when you were
may be short rather than long term. able to persuade someone to do something,
or to convince someone of something (to
bring them around to your point of view).”
Pow er M otivation and Em otion Consistent with M cClelland’s prediction,
high levels of anger were experienced during
Closely integrated with M cClelland’s con the power visualization but not during the
cept of power stress is research evidence achievement or affiliation/intimacy visual
linking power motivation with emotion. izations. Contrary to M cClelland’s reason
McClelland and colleagues (1989) theorized ing, however, excitement was not associated
that individuals engage in behaviors that im with the power visualization but rather with
pact the social environment in ways that link the achievement visualization.
to their dominant motives. Individuals high Building upon M cClelland’s (1989) theo
in n Power, by this reasoning, should enact retical analysis of how incentives activate im
behaviors that create and maintain a feel plicit motives, Schultheiss and Hale (2007)
ing of energy and personal excitement. Ac considered the way that faces expressing
cordingly, Woike (1994) attempted to induce various emotions might affect allocation of
specific affective states by asking power- and attention in power- and affiliation-motivated
intimacy-motivated individuals to vividly individuals. They used as incentives faces
recall an event that had made them very known to express joy, anger, and surprise,
happy or an event that occurred yesterday paired side by side with faces with a neutral
and was ordinary. Immediately thereafter, expression. Previous studies have shown that
research participants completed a question facial expressions consistently elicit certain
naire on their current affective state. Using attributions as to the motives they express
McAdams’s (1982) scoring system for auto (Knutson, 1996). Joyful faces elicit high
biographical memories, Woike found that ratings for both dominance and affiliation,
438 V. M O T IV A T IO N A L D I S P O S IT I O N S
whereas angry faces rate high on dominance well. In presenting his seminal work on the
but low on affiliation. A dominant expres power motive, Winter (1973) commented,
sion from another person, Schultheiss and “A general explanation of the physiological
Hale hypothesized, signals lack of control by mechanisms of all motivated behavior ...
the power-motivated person. Anger and joy, would be valuable if it were complete and
therefore, should figure as disincentives for accurate” (p. 24). Advances in the technol
these persons, constituting cues that should ogy for studying physiological mechanisms
divert their attention. Surprise, on the other that underlie psychological processes are
hand, implies that the power-motivated per opening new research vistas. Illustrative of
son has had impact and therefore should rivet future possibilities is a pilot study by Welsh
the power-motivated person’s attention. (2003) that made use of positron emission
Schultheiss and Hale (2007) employed tomography. Welsh presented aversive im
the dot-probe task (Mogg & Bradley, 1999), ages of mutilated bodies to persons high
which presents an emotional face and a neu in need for intimacy and also persons high
tral face side by side. The computer screen in need for power. These images activated
then masks the faces. A dot next appears in the right fusiform gyrus, a portion of the
the position of either the emotional or the brain that evaluates the socioempathic im
neutral face just previously seen. The par plications of images, as when people pro
ticipant’s response latency in attending to cess faces. Her hypothesis was that intimacy
the dot reflects the degree of attention the motivation would accentuate this process,
participant directed toward the emotional as the neuroimaging data suggested it had.
face. A previous focus on the emotional face With power-motivated persons, on the other
should produce a short latency if the dot hand, these same aversive images activated
replaces that face; a previous focus on the brain regions implicated in defending the
neutral face (and away from the emotional self against threat, namely the right pons, an
face) should result in a long latency, again if area involved in the startle reflex, and the
the dot replaces the emotional face. Power- superior frontal gyrus, an area involved in
motivated individuals attended to the low- deception and angry face processing. These
dominance surprised faces, facial expres findings, although tentative, make sense in
sions that had likely provided reward value terms of what we know about the intimacy
for them in the past. They oriented away and power motives.
from anger faces, as these would pose a Another interesting development is a
threat. Affiliation-motivated persons proved means for determining testosterone level
to be highly vigilant toward anger faces, through analysis of a person’s saliva.
consistent with evidence that persons high Schultheiss and colleagues (2005) found evi
in n Affiliation are rejection-sensitive. dence for a connection between n Power and
testosterone level. Pairs of men competed on
several rounds of a contest that the experi
Som e Concluding Thoughts menters had programmed to result predomi
nantly in success or in failure. For men high
As in so much of the research on the relation in power motivation, more than for those
of personality to social behavior, we look scoring low, success resulted in an increase
back historically with some surprise as we in gonadal steroid testosterone within the
note the directions that research traditions saliva, whereas failure produced a decrease,
have taken. With power motivation, two again more for men scoring high than for
recent trends are especially noticeable, and those scoring low in n Power. The associa
possibly a third. One, as the preceding sec tion between testosterone level in men and
tion suggests, concerns the interconnection predilections toward aggression and domi
between n Power and emotion. Indications nance is well established (Mazur & Booth,
are that this trend will continue into the near 1998). High or rising levels of testosterone,
future. Schultheiss and colleagues theorized, moti
Another growing trend, to which some of vates dominance, thus priming assertive be
the foregoing discussion alludes, is an inter havior. Defeat-induced decreases in testos
est in physiological substrates of the power terone, on the other hand, constitute impact
motive and the other implicit motives as failure and diminish attempts to dominate
29. P o w er M otivation 439
others. Women’s testosterone levels did not Fodor, E. M., & Greenier, K. D. (1995). The power
react in the same way. motive, self-affect, and creativity, J o u r n a l o f R e
se a r c h in P erson ality , 29, 2 4 2 - 2 5 2 .
On a totally different note, W inter’s Fodor, F^. M ., & Smith, T. (1982). The power motive as
(2007) recent analysis of power motivation an influence on group decision making. J o u r n a l o f
as it relates to escalation or deescalation of P erson ality a n d S o c ia l P sy ch olog y , 4 2 , 1 7 8 - 1 8 5 .
Fodor, E. M., Wick, D. P., & Hartsen, K. M. (20 06 ).
conflict offers a possible avenue for explor
The power motive and affective response to asser
ing labor-management negotiations. Should tiveness. J o u r n a l o f R e se a rc h in P erson ality , 4 0 ,
organizational psychologists obtain evidence 598-610.
that parallels what Winter found, they might Friedman, M., & Rosenman, R. H. (1974). T ype A b e
possibly develop diagnostic criteria and in h a v io r a n d y o u r h ea rt. New York: Fawcett.
House, R. J ., Woycke, J., & Fodor, E. M. (1988). Ch ar
tervention techniques that could forestall ismatic and noncharismatic leaders: Differences in
the injurious personal and economic conse behavior and effectiveness. In J. S. Conger & R. N.
quences that accompany the breakdown of Kanungo (Eds.), C h a rism a tic le a d e r s h ip : T h e e lu
negotiations. siv e f a c t o r in o r g a n iz a tio n a l e ffe c tiv e n e s s (pp. 9 8 -
121). San Francisco: Jossey-Bass.
Advances in methodology have been the
Isen, A. M., Daubman, K. A., Sc Nowicki, G. P. (1987).
catalyst for many of the recent findings in Positive affect facilitates creative problem solving.
our quest to better understand power moti J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 5 2,
vation. Theories by themselves merely whet 1122-1131.
Janis, I. L. (1982). G r o u p th in k : P sy c h o lo g ic a l stu d ies
the appetite for further understanding, but
o f p o lic y d ec is io n s a n d fia s c o e s (2nd ed.). Boston:
advances in methodology bring theories to Houghton Mifflin.
fruition. Knutson, B. (1996). Facial expressions of emotion
influence interpersonal trait inferences. J o u r n a l o f
N o n v e r b a l B eh a v io r, 2 0 , 1 6 5 - 1 8 2 .
Matthews, K. A., & Saal, F. FI. (1978). Relationship
References of the Type A coronary-prone behavior pattern to
achievement, power, and affiliation motives. P sy
Amabile, T. M. (1983). T h e s o c ia l p s y c h o lo g y o f c r e
c h o s o m a tic M ed icin e, 4 0 , 6 3 1 - 6 3 6 .
ativ ity . New York: Springer-Verlag.
Mazur, A., & Booth, A. (1998). Testosterone and
Borkenau, P., & Ostendorf, F. (1993). N E O -F iin f-
dominance in men. B eh a v io r a l a n d B rain S cien ces,
F a k to r e n In v en ta r (N E O -F F 1 ) n a ch C o sta a n d M c
21, 353-397.
C ra e: F lan d an iv eisu n g \ N E O -F iv e-F a ctor in v e n
McAdams, D. P. (1982). Experiences of intimacy and
to ry (N E O -F F I) a c c o r d in g to C o sta a n d M cC rae: power: Relationships between personal motives and
M a n u a l|. Gottingen, Germany: Hogrefe. autobiographical memory. J o u r n a l o f P erson ality
Bullock. W. A., & Gilliland, K. (1993). F.ysenck’s a n d S o c ia l P sy ch olog y , 4 2 , 2 9 2 - 3 0 2 .
arousal theory of introversion-extraversion: A co n McClelland, D. C. (1958). Methods of measuring
verging measures investigation. J o u r n a l o f P ers o n human motivation. In J. W. Atkinson (Ed.), M otives
a lity a n d S o c ia l P sy ch olog y , 6 4 , 11 3 - 1 2 3 . in fan tasy , a c tio n , a n d s o c ie ty (pp. 7 - 4 2 ) . Princ
Dacey, J. S. (1 989). F u n d a m en ta ls o f crea tiv e thin kin g. eton, NJ: Van Nostrand.
New York: Lexington Books. McClelland, D. C. (1976). Sources of stress in the drive
deCharms, R., Morrison, H. W., Reitman, W. R ., & for power. In G. Serban (Ed.), P sy c h o p a th o lo g y a n d
McClelland, D. C. (1955). Behavioral correlates h u m a n a d a p t a t io n (pp. 2 4 7 - 2 7 0 ) . New York: Ple
of directly and indirectly measured achievement num Press.
motivation. In D. C. McClelland (Ed.), S tu d ies in McClelland, D. C. (1979). Inhibited power motivation
m o tiv a tio n (pp. 4 1 4 - 4 2 3 ) . New York: Appleton- and high blood pressure in men .J o u r n a l o f A b n o r
Century-Crofts. m a l P sy ch olog y , 8 8, 1 8 2 -1 9 0.
Flowers, M. L. ( 1 97 7 ). A laboratory test of some McClelland, D. C. (1982). The need for power, sympa
implications of J a n i s ’s groupthink hypothesis. thetic activation, and illness. M otiv atio n a n d Flm o-
J o u r n a l o f P e rs o n a lity a n d S o c ia l P sy ch o lo g y , 35, tion , 6, 3 1 - 4 1 .
888-896. McClelland, D. C. (1985). H u m an m o tiv a tio n . Glen
Fodor, E. M. (1984). The power motive and reactivity view, II,: Scott, Foresman.
to power stresses. J o u r n a l o f P erso n a lity a n d S o c ia l McClelland, D. C. (1989). Motivational factors in
P sy c h o lo g y , 47, 8 5 3 - 8 5 9 . health and disease. A m erica n P sy ch o lo g ist, 41,
Fodor, F.. M. (1985). T he power motive, group con 675-683.
flict, and physiological arousal. J o u r n a l o f P ers o n McClelland, D. C „ & Boyatzis, R. E. (1982). The
a lity a n d S o c ia l P sy c h o lo g y , 4 9, 1 4 0 8 - 1 4 1 5 . leadership motive pattern and long-term success in
Fodor, F.. M. (1990). T he power motive and creativity management. Jo u rn a l o f A p p lied P sy ch olog y , 67,
of solutions to an engineering problem. J o u r n a l o f 737-743.
R e se a rc h in P erson ality , 2 4 , 3 3 8 - 3 5 4 . McClelland, D. C., & Burnham, D. H. (1 976, M a r c h -
Fodor, E. M., & Carver, R. A. (20 0 0). Achievement April). Power is the great motivator. H a rv a rd B u si
and power motives, performance feedback, and n ess R e v iew , 5 4 , 1 0 0 - 1 1 0 .
creativity. J o u r n a l o f R e se a rc h in P erson ality , 3 4 , McClelland, D. C., Davidson, R. J ., Floor, E., & Saron,
380-396. C. (1980). Power motivation, catecholamine secre
440 V . M O T IV A T IO N A L D I S P O S IT I O N S
tion, immune function and illness reports. J o u r n a l projective measure of power motivation. J o u r n a l o f
o f H u m an S tress, 6, 11-19. A b n o r m a l P sy ch olog y , 5 4 , 1 - 8 .
McClelland, D. C., Koestner, R., & Weinberger, J. Watson, J . D. (1968). T h e d o u b le h elix . New York:
(1989). How do self-attributed and implicit motives Atheneum.
differ? P sy c h o lo g ic a l R e v iew , 9 6 , 6 9 0 - 7 0 2 . Weinberger, J ., 8c McClelland, D. C. (1990). Cognitive
McClelland, D. C., 8c Winter, D. G. (1969). M o tiv a t versus traditional motivational models: Irreconcil
ing e c o n o m ic a c h ie v e m e n t. New York: Free Press. able or complementary? In E. T. Higgins 8c R. M.
Mednick, S. A., 8c Mednick, M . T. (1967). R e m o t e Sorrentino (Eds.), H a n d b o o k o f m o tiv a tio n a n d
A s s o c ia te s T est: E x p e r im e n t e r ’s m a n u al. Boston: c o g n itio n : Vol. 2. F o u n d a tio n s o f s o c ia l b e h a v io r
Houghton Mifflin. (pp. 5 6 2 - 5 9 7 ) . New York: Guilford Press.
Merton, R. K. (1973). T h e s o c io lo g y o f s c ie n c e. Chi Welsh, K. M. ( 2 0 0 3 , August). Im p licit m o tiv es a n d
cago: University of Chicago Press. b rain a c tiv a tio n in e m o tio n . Paper presented at the
Mogg, K., 8c Bradley, B. P. (1999). Some methodologi convention of the American Psychological Associa
cal issues in assessing attentional biases for threat tion, Toronto, On tario, Canada.
ening faces in anxiety: A replication study using a Winter, D. G. (1 967). P o w er m o tiv a tio n in th o u g h t
modified version of the probe detection task. B e h a v a n d a c tio n . Unpublished doctoral dissertation, H ar
io r R e se a rc h a n d T h era p y , 37, 5 9 5 - 6 0 4 . vard University.
Ritter, K., 8c Medhurst, M . J . (2 003). P resid en tia l Winter, D. G. (1973). T h e p o w e r m otiv e. New York:
sp e ec h w ritin g : F rom th e n ew d e a l to th e R eag an Free Press.
r ev o lu tio n a n d b e y o n d . College Station: Texas A & Winter, D. G. (1987). Leader appeal, leader perfor
M University Press. mance, and the motive profiles of leaders and fol
Schultheiss, O. C., 8c Brunstein, J. C. (2001). Assess lowers. J o u r n a l o f P erson ality a n d S o c ia l P sy c h o l
ment of implicit motives with a research version of og y , 5 2 , 1 9 6 - 2 0 2 .
the TAT: Picture profiles, gender differences, and Winter, D. G. (1992). Power motivation revisited. In C.
relations to other personality measures. J o u r n a l o f P. Smith (Ed.), M o tiv a tio n a n d p e r s o n a lity : H a n d
P erson ality A ssessm en t, 77, 7 1 - 8 6 . b o o k o f th e m a tic c o n te n t an a ly sis (pp. 3 0 1 - 3 1 0 ) .
Schultheiss, O. C., 8c Hale, J . A. ( 2 0 07 ). Implicit m o Cambridge, UK: Cambridge University Press.
tives modulate attentional orienting to facial ex Winter, D. G. (1999). Linking personality and “scien
pression of emotion. M otiv atio n a n d E m o tio n , 31, tific” psychology: The development of empirically
13-24. derived Thematic Apperception Test measures. In
Schultheiss, O. C., W irth, M. M ., Torges, C. M., Pang, L. Geiser 8c M . I. Stein (Eds.), E v o c a tiv e im a g es:
J . S., ViUacorta, M . A., 8c Welsh, K. M . (2005). T h e T h e m a tic A p p e rc e p tio n T est a n d th e a r t o f p r o
Effects of implicit power motivation on men’s and je c t io n (pp. 1 0 7 - 1 2 4 ). Washington, DC: American
women’s implicit learning and testosterone changes Psychological Association.
after social victory or defeat. J o u r n a l o f P erson ality Winter, D. G. ( 2 00 7 ). The role of motivation, respon
a n d S o c ia l P sy ch olog y , 8 8 , 1 7 4 - 1 8 8 . sibility, and integrative complexity in crisis escala
Selye, H. (1973). The evolution of the stress concept. tion: Comparative studies of war and peace crises.
A m eric a n S cien tist, 61, 6 7 2 - 6 9 9 . j o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 9 2,
Stater, R. (1999). J a c k W elch a n d th e G E w ay : M a n 920-937.
a g e m en t in sig h ts a n d le a d e r s h ip sec rets o f th e leg Winter, D. G., J oh n, O. P., Stewart, A. J ., Klohnen, E.
e n d a r y C E O . New York: McGraw-Hill. C., 8c Duncan, L. E. (1998). Traits and motives: T o
Steele, R. S. (1973). T h e p h y s io lo g ic a l c o n c o m ita n ts ward an integration of two traditions in personality
o f p sy c h o g en ic m o tiv e a r o u s a l in c o lle g e m a les. research. P sy c h o lo g ic a l R e v iew , 1 0 5 , 2 3 0 - 2 5 0 .
Unpublished doctoral dissertation, Harvard Uni W irth, M . M ., Welsh, K. M ., 8c Schultheiss, O. C.
versity. (2 00 6 ). Salivary cortisol changes in humans after
Stewart, G. L. (1996). Reward structure as a mod winning or losing a dominance contest depend on
erator of the relationship between extraversion and implicit power motivation. H o r m o n e s a n d B e h a v
sales performance. J o u r n a l o f A p p lie d P sy ch olog y , ior, 49, 3 4 6 - 3 5 2 .
8 1, 6 1 9 - 6 2 7 . Woike, B. A. (1994). Vivid recollection as a technique
Stumpf, H., Angleitner, A., Wieck, T., Jackson, D. N., to arouse implicit motive-related affect. M otiv atio n
8c Beloch-Till, H. (1985). D eu tsc h e P erso n a lity R e a n d E m o tio n , 18, 3 3 5 - 3 4 9 .
se a r c h F o rm (P R F). Gottingen, Germany: Hogrefe. Woike, B. A., 8c McAdams, D. P. (2 005). Motives.
Thayer, R. E. (1978). Factor analytic and reliability In V. J . Derlaga, B. A. Winstead, 8c W. H. Jones
studies on the Activation-Deactivation Adjective (Eds.), P erson ality : C o n te m p o r a r y th e o r y a n d r e
Check List. P sy c h o lo g ic a l R e p o r ts , 4 2 , 7 4 7 - 7 5 6 . se a r ch (pp. 1 5 6 -1 89 ). Belmont, CA: Thomson Wad
Ullman, J. S. (1972). T he need for influence: Develop sworth.
ment and validation of a measure, and compari Zurbriggen, E. L ., 8C Sturman, T. S. (2002). Linking
son with the need for power. G e n e tic P sy ch olog y motives and emotions: A test of McClelland’s hy
M o n o g r a p h s, 8 5 , 1 5 7 - 2 1 4 . pothesis. P erso n a lity a n d S o c ia l P sy ch o lo g y B u lle
Veroff, J . (1957). Development and validation of a tin, 2 8 , 5 2 1 - 5 3 5 .
CHAPTER 30
Social Desirability
R o n a ld R . H o ld e n
J e n n i f e r P a ssey
441
442 V. M O T IV A T IO N A L D I S P O S IT I O N S
values suggested that most items on the For example, does a normatively elevated
Hopelessness scale would fail to meet stan score on a self-report measure of orderliness
dards (e.g., Jackson, 1984, 1989) for being represent a respondent’s true level of being
adequately distinct from response-style vari orderly or the individual’s tendency to an
ance. In general, therefore, the scale and its swer in terms of social desirability?
items appear not to measure anything other In other circumstances, however, social
than a supposedly irrelevant response style. desirability may not be a contaminant but,
It could even be suggested that the scale de rather, a legitimate aspect of the construct
velopers should go back the drawing board. being measured. As an example, it could be
Or should they? Again, empirical studies in theorized that appearing socially desirable
dicate that the Hopelessness scale predicts is an inherent component of the individual-
actual behavior— behavior that is of sub difference variable of affiliation motiva
stantial and societal significance (e.g., death tion (i.e., people who have a greater desire
by suicide). to affiliate may tend to describe themselves
Although concerns about social desir positively to others). Although, for a specific
ability appeared to subside for a while, as construct, such an interpretation could be
personality assessment regained prominence appropriate, applying this reasoning to a
with the ascendance of the Big Five model set of individual differences does raise an
of personality within industrial/organiza important consideration (Holden, 2001).
tional and personnel contexts, social desir If social desirability is a legitimate facet of
ability reemerged as a potential issue. Given many different individual-difference con
that job applicants’ self-report for Big Five structs (e.g., achievement, affiliation, agree
constructs (i.e., Neuroticism, Extraversion, ableness, nurturance) and if negative social
Openness, Agreeableness, Conscientious desirability (i.e., social undesirability) is also
ness) could predict job performance, a con a true component of certain person variables
cern about socially desirable responding, in (e.g., depression, hopelessness, impulsiv
particular impression management, seemed ity, psychopathy), then these constructs are
logical. However, prominent validity stud not conceptually distinct and should not be
ies by Hough, Eaton, Dunnette, Kamp, and theorized, measured, or interpreted as such.
McCloy (1990), Barrick and Mount (1996), Presently, discerning whether social desir
and Piedmont, M cCrae, Riemann, and An- ability is part of the construct being mea
gleitner (2000) seemed to indicate that, at a sured or an interfering response style is not
group level, social desirability does not at an easy task and is not well established ei
tenuate self-report personality scale valid ther theoretically or empirically.
ity, leading some to suggest that the issue How does social desirability relate to
is a “red herring” (Ones, Viswesvaran, & other personality variables? This seemingly
Reiss, 1996). More recently, however, find straightforward question is not answered
ings suggest that this putative red herring readily. Difficulties arise because social de
may be important after all (Holden, 2007, sirability is not well defined theoretically,
2008). Holden (2007) indicated that, aver because measures of social desirability are
aged over personality scales, socially desir quite varied, and because the structure of
able responding can account for 10—15% social desirability is not completely articu
of the variability of the prediction in peer lated. Historically, social desirability scales
criterion ratings. Furthermore, under high- have been associated with personality vari
stakes testing associated with the military, ables of honesty, need for approval, and ego
White, Young, and Rumsey (2001) reported resiliency. However, subscales of social de
that the validity of personality measures that sirability have also been shown to correlate,
predict duty performance and attrition is se with at least a medium effect size (i.e., r >
verely compromised as a function of social .30), with every domain scale of the Big Five
desirability. personality traits (Paulhus, 2002).
Where does this leave us? It appears that More recently, in a novel statistical residu-
social desirability can contaminate self- alization approach to characterizing social
report scales and, consequently, threaten desirability bias as a departure from reality,
scales’ construct validity. In such instances, Paulhus (2002) identified egoistic and moral
scale score interpretations are ambiguous. istic biases as the chief facets of social desir
444 V. M O T IV A T IO N A L D I S P O S IT I O N S
ability bias. The egoistic bias of social desir of their popularity, their psychometric prop
ability focuses on a narcissistic exaggeration erties, and the reputations of their develop
of agency-related traits such as dominance, ers.
courage, emotional stability, intelligence, The Edwards Social Desirability Scale
and creativity. Also associated with this bias (Edwards, 1957) comprises items from the
are personality characteristics of ego resil M M PI. The scale includes 39 true-false
ience, achievement via independence, social items that were selected by judges who
potency, perceived capability, lack of dis unanimously agreed on the direction of key
tress, and personal growth (Paulhus & John, ing when asked to respond in a socially de
1998). Social desirability’s moralistic bias sirable fashion. In assessing scale internal-
exaggerates communion-related traits such consistency reliability, a coefficient alpha of
as duty, agreeableness, and impulse control .79 has been reported with a sample of uni
(Paulhus, 2002). Related personality traits versity students (Holden & Fekken, 1989).
are ego control, achievement via conformity, Sample items include “My hands and feet
nurturance, social closeness, interpersonal are usually warm enough” (true-keyed), “I
sensitivity, restraint, and socialization (Paul am very seldom troubled by constipation”
hus & John, 1998). These innovative char (true-keyed), and “I find it hard to keep my
acterizations of social desirability await fur mind on a task or job” (false-keyed).
ther empirical confirmation. The M arlow e-Crow ne Social Desirability
Because researchers are far from reaching Scale (Crowne & Marlowe, 1960) focuses on
a consensus on the contamination-versus- culturally sanctioned and approved but un
legitimate content issue, test developers and likely behavior. In this way this scale’s items
users should be alert to the potential delete avoid psychopathology content. Items were
rious effects of social desirability. Two sug devised and selected based on consideration
gestions may be particularly relevant. First, of various personality inventories, judged
although many personality dimensions (e.g., social desirability ratings, and item -total
hopelessness) could have content aspects of associations. Holden and Fekken (1989) re
socially desirable responding as part of their ported a coefficient-alpha scale reliability
constructs, if a goal is to construct mul of .78 for an undergraduate sample. Sample
tiscale inventories with relatively distinct items include “Before voting, I thoroughly
and independent scales, then item selection investigate the qualifications of all the can
procedures that minimize social desirability didates” (true-keyed), “There have been oc
will serve to foster greater scale indepen casions when I felt like smashing things”
dence. Conversely, item selection techniques (false-keyed), and “I have never deliberately
that enhance scale orthogonality should at said something that hurt someone’s feelings”
tenuate the socially desirable responding (true-keyed).
associated with the resultant personality The Personality Research Form (Form E)
scales. Second, in situations in which the Desirability Scale (Jackson, 1984) consists of
motivation to distort self-report exists, the 16 true-false items. Items represent hetero
use of a separate scale of socially desirable geneous content and have relatively extreme
responding is recommended. Even if the con values of scaled desirability. Jackson report
tent scale has been developed to reduce the ed a Spearman-Brown corrected split-half
impact of socially desirable responding, an reliability of .68 for college students. Sample
independent index of test-respondent moti items include “I am quite able to make cor
vation can alert test users to particular in rect decisions on difficult questions” (true-
dividuals whose scores may reflect stylistic keyed), “I am never able to do things as well
rather than content responding. as I should” (false-keyed), and “My life is
full of interesting activities” (true-keyed).
The Balanced Inventory of Desirable Re
M easurem ent sponding (Version 7) Self-Deceptive En
hancement Scale (Paulhus, 1998) focuses on
Although a plethora of scales are available to an unconscious favorability bias. This scale
measure social desirability, five of the more comprises 20 items answered on 5-point
noted ones are mentioned here. These scales rating scales ranging from 1 (not true) to
have been and are more prominent because 5 (very true). With items scored dichoto-
30. S ocial D esirability 445
mously, this scale assesses a pervasive lack of of the Edwards and M arlow e-Crow ne so
insight associated with narcissistic overcon cial desirability scale items, concluded that
fidence. Paulhus reported scale coefficient- the two scales measure distinct factors.
alpha reliability coefficients ranging from Holden and Fekken (1989) factor analyzed
.70 for college students to .75 for the gen responses to 92 items from three social de
eral population. Sample items include “My sirability scales and uncovered eight first-
first impressions of people usually turn out order and two higher order factors. In inter
to be right” (positively keyed), “It would be preting their solution, they did not rule out
hard for me to break any of my bad habits” the possibility that factors could represent
(negatively keyed), and “I don’t care what both style and substance and indicated that
other people really think of me” (positively the two higher order social desirability fac
keyed). tors represented a Sense of Own General
The Balanced Inventory of Desirable Re Capability and Interpersonal Sensitivity.
sponding (Version 7) Impression M anage Contained within the Sense of Own General
ment Scale (Paulhus, 1998) consists of 20 Capability factor were facets of focused and
items responded to on 5-point rating scales. realistic thinking (e.g., no difficulty concen
Items are scored dichotomously, and the trating), social integration (e.g., a sense of
scale assesses faking, lying, and dissimula being cared for), self-confidence (e.g., not
tion. Coefficient-alpha reliability for the being self-conscious), hardiness (e.g., seldom
scale is reported by Paulhus as .81 for college being ill), and acceptance of responsibility
students and .84 for prison entrants. Sample (e.g., actions following words). For Inter
items include “I sometimes tell lies if I have personal Sensitivity, subcomponents of con
to” (negatively keyed), “I never cover up my siderateness (e.g., never deliberately hurting
mistakes” (positively keyed), and “There someone), social sensitivity (e.g., always
have been occasions when I have taken ad being courteous), and tolerance (e.g., not re
vantage of someone” (negatively keyed). senting doing favors) were uncovered.
Regardless of whether scales, miniscales,
or items are analyzed, a consensus emerges
Structure that at least two correlated but distinct high
er order factors of social desirability exist,
Although social desirability is commonly re one emphasizing the self and one focusing
garded as a single response style, evidence on others. These factors may include both
suggests that the structure may be more content and style and, consequently, may or
complex. In a factor analysis of 30 scales of may not indicate distortions from accuracy.
stylistic responding (including more than just That is, although departures from veridical-
social desirability scales), Wiggins (1964) re ity will be indicated by extreme scores on
ported three factors associated with social these factors, extreme scores do not neces
desirability: Alpha (unfavorable vs. favor sarily imply invalid responding.
able self-evaluation); Gamma (a lying factor);
and Cautious, Controlled Good Impression.
In also analyzing at the scale level, Paulhus’s R ecen t Investigations
(1984) analysis of various response-style into Social Desirability
measures identified self-deception and im
pression management as two distinct com Given the potential detrimental effects of
ponents of socially desirable responding. social desirability, the extent of its possible
More recently, Paulhus (2002) suggested impact should be considered. The follow
a revision of the structure, now indicating ing is a nonexhaustive review of recent so
four socially desirable responding aspects cial desirability investigations conducted in
(self-deceptive enhancement, agency man prominent research areas. Though many of
agement, self-deceptive denial, and commu these explorations may be preliminary, and
nion management). although in some areas only a handful of
Complementing the scale-level results, studies have been conducted, future studies
lower level analyses also indicate the mul will likely build on these examinations. Thus
tidimensional nature of social desirability. we refrain from drawing general conclusions
O ’Grady (1988), in analyzing six miniscales in anticipation both of future work and of
446 V. M O T IV A T IO N A L D I S P O S IT I O N S
Recent research examining the relation reported that social desirability scale scores
ship between social desirability and emo correlated negatively (r = -.2 5 ) with the
tion has indicated that social desirability is presence of a previous suicide attempt for a
positively associated with emotional intel sample drawn from the general community.
ligence (Mesmer-Magnus, Viswesvaran, Among psychiatric patients, social desir
Deshpande, & Joseph, 20 0 6 ). Furthermore, ability correlated negatively with clinicians’
emotional intelligence was a significant pre ratings of patients’ suicidal desire (r - -.3 4 ),
dictor of social desirability over and above suicide preparation (r - -.2 2 ), and suicide
self-esteem and overclaiming alone. In stud ideation (r = -.3 0 ) (Flolden, Mendonca, &
ies examining responses to emotional adver Serin, 1989). Holden and colleagues (1989)
tising, men reported a less pleasant viewing also found that, for male federal inmates,
experience and a less favorable attitude to social desirability correlated negatively (r
ward the advertisement when a stereotype- = -.2 9 ) with a history of a previous suicide
incongruent advertisement was viewed with attempt. Similarly, Ivanoff and Jang (1991)
another man, whereas their responses were found that, for male state inmates, social
not affected by the presence of another man desirability scale scores correlated negative
when they were exposed to a stereotype- ly with clinician-rated suicide ideation (r =
congruent advertisement (Fisher & Dube, -.4 4 ) and a previous history of suicidal be
2005). Therefore, research has emerged in havior (r = -.3 2 ). Thus it appears that there
several areas showing a significant influence is substantive as well as stylistic variance in
of social desirability on well-being. measures of social desirability. Perhaps this
Other research, however, indicates that is the reason that social desirability may be
scales in the domain of adjustment func evolving into a concept for which some now
tion independently of social desirability. For indicate that, in measuring this response-
example, some measures of life satisfaction style concept, there is a “necessity of demon
appear to be untainted by social desirability strating departure-from-reality” (Paulhus,
in adolescents (Gilman &c Barry, 2003) and 2 0 0 2 , p. 49). Thus many current self-report
younger and older adults (Laicardi, Baldas- scales of social desirability may lack a nec
sarri, & Artistico, 2001). Furthermore, Rog essary exaggeration-from-veridicality aspect
ers, Reinecke, and Setzer (2004) revealed a or at least confound it with true content as
strong association, independent of social sociated with other constructs.
desirability, between childhood attachment
experience and cognitive vulnerability in a
Self-K now ledge and G oals
sample of clinically depressed adults. Like
wise, Cramer (2000) found that, although Just as social desirability may influence self-
social desirability was positively associated reported measures of psychological well
with satisfaction, social support, and men being, it may also affect measures of self-
tal health, the association between these knowledge. Social desirability is positively
latter two constructs remained relatively associated with the evaluative component
unchanged when social desirability was con of self-knowledge: self-esteem (Mesmer-
trolled for. Future research will be needed Magnus et al., 2006). Social desirability also
to examine the full extent and boundaries moderates the relationship between implicit
of the effects of social desirability on self- and explicit self-esteem such that the rela
reported well-being. tionship is stronger under conditions of high
Rather than having a construct definition, or low self-deception depending on the type
social desirability, as measured by self-report of implicit measure used (Riketta, 2005).
scales, is traditionally viewed as a response Self-knowledge has been found to be signifi
style that threatens the construct validity of cantly influenced by social desirability, but
other self-report scales. Nevertheless, social the amount of influence differs as a function
desirability scale scores, in and of them of the self-knowledge domain (Meleddu &
selves, show interesting relationships with Guicciardi, 1998). Specifically, social desir
behavioral and non-self-report measures. In ability effects are weaker for the anxiety
particular, the relationship of social desir domain than for the extraversion domain in
ability and suicide appears to be stable and that the ideal self is a better predictor of the
generalizable. Linehan and Neilsen (1981) actual self for extraversion than for anxiety.
448 V. M O T IV A T IO N A L D IS P O S IT IO N S
These investigations indicate a consistent Cuixia, Jian , and Zhongfang (2003) report
impact of social desirability on the measure ed that Chinese college students’ ratings of
ment of self-knowledge. item social desirability and estimates of the
Social psychology has seen increased in percentage of others who would behave in
terest in studies examining aspects of mo the manner described by these items indi
tivation. Accordingly, recent research has cated self-enhancement and honesty similar
investigated whether social desirability in to that for American college students. Chi
fluences goal orientation related to work nese college students perceived that they did
and academics (Tan & H all, 2005). So more desirable and fewer undesirable activi
cial desirability was negatively associated ties than others and chose to give more hon
with learning goals and positively related est responses to undesirable items that were
to performance-avoidance goals, whereas perceived to be more neutral than desirable
performance-approach goals were uncon items. Future research will need to deter
taminated by social desirability. Grossbard mine whether different samples and meth
and colleagues (2007) reported that so odologies account for these discrepancies or,
cial desirability was negatively associated alternatively, whether self-enhancement is
with ego orientation in adolescent men and an exception to the influence of culture on
women and positively associated with task social desirability.
orientation in women and that it reduced the
relationship between goal orientations and
Relationships, Attraction,
performance anxiety in women. Though the
and Gender Roles
reasons for these complex relationships re
quire further investigation, social desirabil In exploring love styles, Davies (2001) found
ity appears to be related to people’s goals that social desirability was associated with
and motivations. traditional gender-role socialization such that
social desirability negatively correlated with
possessive, dependent love styles [mania) in
Culture
men and women. Furthermore, for men, so
Research suggests that cultural orientation cial desirability was positively related to ro
may be associated with socially desirable re mantic, passionate (eros) and game-playing
sponding. Middleton and Jones (2000) ex (ludus) love and negatively associated with
plored differences in overall social desirability all-giving, selfless love (agape). For women,
across Eastern and Western cultures. Signifi social desirability was positively related to
cant differences existed such that students agape and negatively related to ludus.
from Asian countries reported higher social The susceptibility of attachment-style
desirability than students from the United measures to impression-management and
States and Canada. Across several studies, self-deception aspects of social desirability
European Americans, compared with Asian has been examined in college students (Leak
Americans, scored higher on self-deception & Parsons, 2001). All attachment measures
and lower on impression-management facets examined were influenced by impression-
of social desirability, and individualism was management tendencies, and two of three
positively associated with self-deception, were influenced by self-deception. Maltby
whereas collectivism was positively associ and Day (2000) examined whether the im
ated with impression management (Lalwani, portance of romantic acts was associated
Shavitt, & Johnson, 20 0 6 ). Keillor, Owens, with social desirability and found that en
and Pettijohn (2001) investigated differences dorsement of romantic acts was positively
in social desirability in 15- to 65-year-olds related to social desirability in both men and
across cultures and observed higher social women. Loving and Agnew (2001) developed
desirability in Malaysian compared with a measure of social desirability (including
U.S. and French samples. impression-management and self-deception
Contrary to the research that demon components) for use in relationship re
strates significant relations between cultur search with dating and married couples.
al orientation and social desirability, some Impression-management scores were high
studies demonstrate a lack of cultural dif er in public versus private conditions, and
ferences in self-enhancement. For example, self-deception was associated with several
3 0. S ocial D esirability 449
the impact of self-deception and impression- enced reporting of marijuana and cocaine
management facets of social desirability drug use by arrestees in the United States.
on undergraduate students’ sexuality self- In comparing self-reports with the results
reports. Sexual adjustment was associated of urinalysis tests of drug use, social desir
with self-deception variables for both sexes, ability was related to reporting such that ar
whereas a number of intrapersonal and in restees who tested positive for cocaine were
terpersonal sexual behaviors for women and 15 times more likely to misreport their drug
unrestricted sexual attitudes and fantasies use than those who tested positive for mari
for men were negatively associated with im juana.
pression management. Furthermore, the as Thus a large body of work has demon
sociations between impression management strated that social desirability appears to
and sexuality measures were significant even exert at least some influence on self-reported
after controlling for general personality and measures of physical health and could have
conservatism. impact on several areas that are significant
The influence of self-deceptive social de to society. In addition to the research men
sirability on the accuracy of self-reported tioned here on psychological well-being,
FIIV serostatus was studied in active injec- self-knowledge, motivation, relationships,
tion-drug users (Latkin & Vlahov, 1998). and culture, social desirability continues to
For respondents who scored low on self- be of interest among social and personality
deception, the sensitivity of self-reported psychologists.
H IV serostatus was 81% as compared with
63% for those individuals who scored high
on self-deception. However, for a study ex Sum m ary
amining the relationship between social
desirability and self-reported condom use As research on social desirability contin
behavior in sex workers in the Philippines, ues to evolve, additional promising avenues
no relationship was found (Morisky, Ang, and techniques emerge. Holden and his col
&c Sneed, 2002). Furthermore, Keffala and leagues (Holden & Kroner, 1992; Holden,
Stone (1999) found that social desirability Kroner, Fekken, & Popham, 1992) have
was not related to psychologists’ decisions identified the use of response latencies as
to maintain or break confidentiality of HIV- a behaviorally based method for assess
positive patients across 16 scenarios of vary ing socially desirable responding in general
ing risk, danger, and situation. and faking in particular. In their model, re
W illebrand, Wikehult, and Ekselius sponses that are incongruent with a faking
(2005) found that among former burn pa schema take longer to produce than schema-
tients, high social desirability was associated congruent answers. From the perspective
with poorer perceived health on the burn- of item-response theory, Zickar and his
specific health subscales of heat sensitivity, associates (Zickar, Gibby, & Robie, 2 0 0 4 ;
work, and body image. However, another Zickar & Robie, 1999) have applied this
study examining reported somatic and emo item-response approach toward the model
tional health in adults found no influence ing of induced socially desirable responding.
for social desirability (Sheridan, Mulhern, Their work indicates that patterns of in
& M artin, 1999). duced faking are heterogeneous rather than
Other aspects of physical health have also homogeneous in nature. This, again, sug
been explored. A study of the relationship gests that socially desirable responding is a
between social desirability and self-reported multidimensional phenomenon. Employing
heroin cravings revealed that high social de signal-detection methods, Paulhus and his
sirability was related to lower self-reported collaborators (Paulhus, Harms, Bruce, &
cravings but was not related to physiological Lysy, 2 0 0 3 ; W illiams, Paulhus, & Harms,
craving measures and did not moderate the 2001) have developed an operationalized
relation between self-reported and physi implementation of social desirability. In this
ological craving indices (Marissen, Fran- technique, the overclaiming of knowledge
ken, Blanken, van den Brink, & Hendriks, involves a self-serving distortion that relates
2005). Sloan, Bodapati, and Tucker (2004) to other self-report measures of socially de
examined whether social desirability influ sirable self-deceptive enhancement. Further
3 0. S ocial D esirability 451
more, this approach represents a relatively lessness scale. J o u r n a l o f C o n su ltin g a n d C lin ical
P sy ch olog y , 4 2 , 8 6 1 - 8 6 5 .
subtle and unobtrusive measure of socially
Bourgeois, A. E., Loss, R., Meyers, M . C., &c LeUnes,
desirable responding. A. D. (2003). The Athletic Coping Skills Inventory:
Social desirability is a multidimensional Relationship with impression management and self-
response style that has the potential to com deception aspects of socially desirable responding.
promise the accurate measurement of non- P sy ch o lo g y o f S p o rt a n d E x erc is e, 4, 7 1 - 7 9 .
Brown, G. K., Beck, A. T., Steer, R. A., & Grisham,
cognitive individual differences assessed by J. R. ( 2 0 00 ). Risk factors for suicide in psychiatric
self-report. The extent of this disruption outpatients: A 20-year prospective study. J o u r n a l o f
may be substantial depending on the domain C o n su ltin g a n d C lin ica l P sy ch olog y , 6 8 , 3 7 1 - 3 7 7 .
and context of the assessment. Scales of so Carels, R. A., Cacciapaglia, H. M ., Rydin, S., Doug
lass, O. M ., & Harper, J. (200 6). Can social desir
cial desirability can reflect both content and
ability interfere with success in a behavioral weight
stylistic responding, and, in some domains, loss program? P sy ch o lo g y a n d H ea lth , 2 1 , 6 5 - 7 8 .
the content component of a social desir Chillag, K., Guest, G., Bunce, A., Joh nson, L., Kil-
ability scale can be relevant for the domain m arx, P. H., & Smith, D. K. (20 0 6). Talking about
being measured. Distinguishing relevant sex in Botswana: Social desirability bias and pos
sible implications for HIV-prevention research. A f
content responding from irrelevant stylistic rica n J o u r n a l o f A ID S R esea rch , 5, 1 2 3 - 1 3 1 .
responding is challenging, both theoretically Cramer, D. (200 0). Social desirability, adequacy of so
and operationally. Although new directions cial support and mental health. J o u r n a l o f C o m m u
in social desirability research may assist in nity a n d A p p lie d S o c ia l P sy ch olog y , 10, 4 6 5 - 4 7 4 .
Crowne, D. P., & Marlow e, D. (1960). A new scale of
separating substance from style, in the in
social desirability independent of psychopathology.
terim, prudent scale users should remain J o u r n a l o f C o n su ltin g P sy ch olog y , 2 4 , 3 4 9 - 3 5 4 .
alert to social desirability response styles as Cuixia, I.., Jian, X ., & Zhongfang, Y. (2003). A
potential threats to the construct validity of compromise between self-enhancement and hon
esty: Chinese self-evaluations on social desirability
self-reports.
scales. P sy c h o lo g ic a l R e p o rts , 9 2, 2 9 1 - 2 9 8 .
Dadds, M . R „ Perrin, S., & Yule, W. (1998). Social de
sirability and self-reported anxiety in children: An
A ckn o w led gm en t analysis of the R C M A S Lie Seal e. J o u r n a l o f A b n o r
m a l C h ild P sy ch o lo g y , 2 6 , 3 11-317.
Preparation of this chapter was supported, in Davies, M . F. (2001). Socially desirable responding
part, by a grant from the Social Sciences and H u and impression management in the endorsement of
love styles. J o u r n a l o f P sy ch o lo g y : In terd iscip lin a ry
manities Research Council o f C an ada.
a n d A p p lied , 135, 5 6 2 - 5 7 0 .
Edwards, A. E. (1957). T h e s o c ia l d esira b ility v a r ia b le
in p e r s o n a lity a s s e ss m e n t a n d research . New York:
R eferen ces Dryden.
Edwards, A. L., & Walker, J . N. (1961). A short form
Armitage, C. J., & Conner, M. (1999). Predictive va of the M M P I: The SD Scale. P sy c h o lo g ic a l R e p o rts ,
lidity of the theory of planned behaviour: The role 8, 4 8 5 - 4 8 6 .
o f questionnaire format and social desirability. Ellis, A. (1946). The validity of personality question
jo u r n a l o f C o m m u n ity a n d A p p lie d S o c ia l P sy ch o l naires. P sy c h o lo g ic a l B u lletin , 4 3 , 3 8 5 - 4 4 0 .
og y , 9, 2 6 1 - 2 7 2 . Fisher, R. J., & Dube, L. (2005). Gender differences in
Barrick, M. R., & Mount, M . K. (1996). Effects of im responses to emotional advertising: A social desir
pression management and self-deception on the pre ability perspective. J o u r n a l o f C o n s u m er R esearch ,
dictive validity of personality constructs .J o u r n a l o f 31, 8 5 0 - 8 5 8 .
A p p lied P sy ch olog y , 81, 2 6 1 - 2 7 2 . Frank, L. F. (1948). P ro jectiv e m e th o d s . Springfield,
Baxter, S. D., Smith, A. F., Litaker, M. S., Baglio, M. L., IL: Thomas.
Guinn, C. H., & Shaffer, N. M . ( 2 0 04 ). Children’s Furnham, A., Petrides, K. V., & Spencer-Bowdage, S.
social desirability and dietary reports. J o u r n a l o f (2002). T he effects of different types of social desir
N u trition E d u c a tio n a n d B eh a v io r, 3 6 , 8 4 - 8 9 . ability on the identification of repressors. P ers o n a l
Beck, A. T., Brown, G., Berchick, R. J ., Stewart, B. L., ity a n d In d iv id u a l D iffer en c es , 3 3 , 1 1 9 - 1 3 0 .
& Steer, R. A. (1990). Relationship between hope Gianakos, 1. (2002). Predictors of coping with work
lessness and ultimate suicide: A replication with stress: T he influences of sex, gender role, social
psychiatric outpatients. A m erica n J o u r n a l o f P sy desirability, and locus of control. S ex R o le s, 4 6 ,
c h ia try , 147, 1 9 0 - 1 9 5 . 1 4 9 -1 5 8.
Beck, A. T., Steer, R. A., Kovacs, M ., & Garrison, B. Gilman, R., & Barry, J . (2003). Life satisfaction and
(1985). Hopelessness and eventual suicide: A 10- social desirability among adolescents in a residential
year prospective study of patients hospitalized with treatment setting: Changes across time. R e sid e n tia l
suicidal ideation. A m er ica n J o u r n a l o f P sychiatry, T rea tm en t fo r C h ild ren a n d Y outh, 2 1, 1 9 - 4 2 .
1 42, 5 5 9 - 5 6 3 . Goodenough, F. L. (19 46). Semantic choice and per
Beck, A. T., Weissman, A., Lester, I)., & Trexler, L. sonality structure. S cien ce , 104, 4 5 1 - 4 5 6 .
(1974). The measurement of pessimism: The Hope Gravdal, L., & Sandal, G. M. (20 06 ). The two-factor
452 V. M O T IV A T IO N A L D IS P O S IT IO N S
model of social desirability: Relation to coping and p s y c h o lo g y (pp. 7 7 5 - 7 9 2 ) . Englewood Cliffs, NJ:
defense, and implications for health. P erson ality Prentice-Hall.
a n d In d iv id u a l D iffe r e n c e s , 4 0 , 10 5 1-1 06 1 . Jack son, D. N. (1984). P erson ality R e se a rc h F o rm
Grossbard, J . R., Cumming, S. P., Standage, M ., Smith, m a n u a l (3rd ed.). Port Huron, MI: Research Psy
R. E., & Smoll, F. L. ( 2 0 07 ). Social desirability and chologists Press.
relations between goal orientations and competitive Jack son, D. N. (1989). B a sic P erson ality In v e n to ry
trait anxiety in young athletes. P sy c h o lo g y o f S p o rt m a n u al. Port Huron, M I: Sigma Assessment Sys
a n d E x erc is e, 8, 4 9 1 - 5 0 5 . tems.
Holden, R. R. (2001). Social desirability. In W. E. Jack son, D. N., & Messick, S. (1961). Acquiescence
Craighead & C. B. Nemeroff (Eds.), T h e C o rsin i and desirability as response determinants on the
e n c y c lo p a e d ia o f p s y c h o lo g y a n d b e h a v io r a l s c i M M P I . E d u c a tio n a l a n d P s y c h o lo g ic a l M e a su r e
e n c e (3rd ed., Vol. 4, pp. 1 5 5 7 - 1 5 5 8 ). New York: m en t, 2 1, 7 7 1 - 7 9 0 .
Wiley. Jack son, D. N., Sc Messick, S. (1962a). Response styles
Holden, R. R. (20 07 ). Socially desirable responding and the assessment of psychopathology. In S. M es
does moderate scale validity both in experimental sick & J . Ross (Eds.), M ea su r em en t in p er so n a lity
and in nonexperimental contexts. C a n a d ia n J o u r a n d c o g n itio n (pp. 1 2 9 - 1 5 5 ) . New York: Wiley.
n al o f B e h a v io u r a l S cien ce, 3 9 , 1 8 4 - 2 0 1 . Jackson, D. N., & Messick, S. (1962b). Response styles
Holden, R. R. (2 008). Underestimating the effects on the M M P I : Comparison of clinical and normal
of faking on the validity of self-report personality samples. J o u r n a l o f A b n o r m a l a n d S o c ia l P sy ch o l
scales. P erso n a lity a n d In d iv id u a l D iffer en c es , 4 4 , og y , 6 5, 2 8 5 - 2 9 9 .
311-321. Keffala, V. J . , Sc Stone, G. L. (1999). Role of H IV
Holden, R. R., & Fekken, G. C. (1989). Three common serostatus, relationship status of the patient, ho
social desirability scales: Friends, acquaintances, or mophobia, and social desirability of the psycholo
strangers? J o u r n a l o f R e se a rc h in P erson ality , 2 3 , gist on decisions regarding confidentiality. P sy ch o l
180-191. o g y a n d H ea lth , 14, 5 6 7 - 5 8 4 .
Holden, R. R., & Kroner, D. G. (1992). Relative ef Keillor, B., Owens, D., & Pettijohn, C. (2001). A cross-
ficacy of differential response latencies for detecting cultural/cross-national study of influencing factors
faking on a self-report measure of psychopathology. and socially desirable response biases. In te r n a tio n a l
P s y c h o lo g ic a l A ssessm en t, 4, 1 7 0 - 1 7 3 . J o u r n a l o f M a r k e t R e se a rc h , 4 3 , 6 3 - 8 4 .
Holden, R. R ., Kroner, D. G., Fekken, G. C., & Po- Klesges, L. M., Baranowski, T., Beech, B., Cullen, K.,
pham, S. M. (1992). A model of personality test Murray, D. M ., Rochon, J ., et al. (20 04 ). Social de
item response dissimulation. J o u r n a l o f P erson ality sirability bias in self-reported dietary, physical activ
a n d S o c ia l P sy ch olog y , 6 3, 2 7 2 - 2 7 9 . ity and weight concerns measures in 8- to 10-year-
Holden, R. R., & Mendonca, J . D. (1984). Hopeless old African-American girls: Results from the Girls
ness, social desirability, and suicidal behavior: A Health Enrichment Multisite Studies (GEMS). P re
need for conceptual and empirical disentanglement. ven tiv e M ed icin e: An In te r n a tio n a l J o u r n a l D e v o t
J o u r n a l o f C lin ic a l P sy ch olog y , 4 0 , 1 3 4 2 - 1 3 4 5 . e d to P ractice a n d T h eo r y , 3 8 , S 7 8 - S 8 7 .
Holden, R. R., Mendonca, J. D., & Mazm anian, D. Laicardi, C., Baldassarri, F., & Artistico, D. (2001).
(1985). Relation of response set to observed suicide Unidimensionality of life satisfaction and its rela
intent. C a n a d ia n J o u r n a l o f B e h a v io u r a l S cien ce, tion to social desirability: A confirmatory study of
17, 3 5 9 - 3 6 8 . a short form of the Life Satisfaction Scale. P sy ch o
Holden, R. R., Mendonca, J. D., & Serin, R. C. (1989). lo g ic a l R e p o r ts , 8 8 , 2 5 3 - 2 6 1 .
Suicide, hopelessness, and social desirability: A test Lalwani, A. K., Shavitt, S., & Johnson, T. (2 006).
of an interactive model. J o u r n a l o f C o n su ltin g a n d W hat is the relation between cultural orientation
C lin ic a l P sy ch olog y , 57, 5 0 0 - 5 0 4 . and socially desirable responding? J o u r n a l o f P er
Hough, L. M ., Eaton, N. K., Dunnette, M . D., Kamp, so n a lity a n d S o c ia l P sy ch olog y , 9 0 , 1 6 5 - 1 7 8 .
J. D., Sc McCloy, R. A. (1990). Criterion-related va Latkin, C. A., & Vlahov, D. (1998). Socially desirable
lidities of personality constructs and the effect of response tendency as a correlate of accuracy of self-
response distortion on those validities. J o u r n a l o f reported H IV serostatus for H IV seropositive injec
A p p lied P sy ch olog y , 75, 5 8 1 - 5 9 5 . tion drug users. A d d ic tio n , 9 3, 1 1 9 1 -11 9 7.
Huang, C. Y., Liao, H. Y., & Chang, S. H. (1998). Leak, G. K., Sc Parsons, C. J . (2001). The suscepti
Social desirability and the clinical self-report inven bility of three attachment style measures to socially
tory: Methodological reconsideration. J o u r n a l o f desirable responding. S o c ia l B e h a v io r a n d P ers o n
C lin ic a l P sy ch olog y , 5 4 , 5 1 7 - 5 2 8 . ality, 2 9, 2 1 - 3 0 .
Ivanoff, A., & Jang, S. J. (1991). The role of hopeless Linehan, M . M., & Nielsen, S. L. (1981). Assessment
ness and social desirability in predicting suicidal be of suicide ideation and parasuicide: Hopelessness
havior: A study of prison inmates. J o u r n a l o f C o n and social desirability. J o u r n a l o f C o n su ltin g a n d
su ltin g a n d C lin ic a l P sy ch olog y , 59, 3 9 4 - 3 9 9 . C lin ica l P sy ch olog y , 4 9, 7 7 3 - 7 7 5 .
Jackson, D. N. (1970). A sequential system for person Linehan, M . M ., & Nielsen, S. L. (1983). Social desir
ality scale development. In C. D. Spielberger (Ed.), ability: Its relevance to the measurement of hope
C u rren t to p ic s in c lin ica l a n d co m m u n ity p s y c h o lo lessness and suicidal behavior. J o u r n a l o f C o n su lt
g y (Vol. 2 , pp. 6 1 - 9 6 ) . New York: Academic Press. in g a n d C lin ica l P sy ch olog y , 51, 141-143.
Jackson, D. N. (1971). The dynamics of structured Loving, T. J., Sc Agnew, C. R. (2001). Socially de
personality tests: 1971. P sy c h o lo g ic a l R e v iew , 78, sirable responding in close relationships: A dual
229-248. component approach and measure. J o u r n a l o f S o
Jackson, D. N. (1973). Structured personality assess c ia l a n d P ers o n a l R e la tio n s h ip s, 18, 5 5 1 - 5 7 3 .
ment. In B. B. Wolman (Ed.), H a n d b o o k o f g en e ra l Maltby, J ., & Day, L. (200 0 ). Rom antic acts as a corre
30. S ocial D esirability 453
late of social desirability, neuroticism, and extraver Paulhus, D. L. (2002). Socially desirable responding:
sion. J o u r n a l o f P sy c h o lo g y : In terd iscip lin a ry a n d The evolution of a construct. In H. I. Braun, D.
A p p lied , 13 4 , 4 6 2 - 4 6 4 . N. Jackson, & D. E. Wiley (Eds.), T h e ro le o f c o n
Marissen, M . A. E., Franken, I. H. A., Blanken, P., stru cts in p s y c h o lo g ic a l a n d e d u c a t io n a l m e a s u r e
van den Brink, W., & Hendriks, V. M. (2005). The m en t (pp. 4 9 - 6 9 ) . Mah w ah, NJ: Erlbaum.
relation between social desirability and different Paulhus, D. E., Harms, P. D., Bruce, N., & Lysy, D.
measures of heroin craving. J o u r n a l o f A d d ictiv e C. (2003). The over-claiming technique: Measuring
D isea ses, 2 4 , 9 1 - 1 0 3 . self-enhancement independent of ability. J o u r n a l o f
Meehl, P. E. (1945). The dynamics of “structured” P erson ality a n d S o c ia l P sy ch olog y , 8 4, 8 9 0 - 9 0 4 .
personality assessment. J o u r n a l o f C lin ica l P sy ch o l Paulhus, D. L., & John, O. P. (1998). Egoistic and
og y , /, 2 9 6 - 3 0 3 . moralistic biases in self-perception: The interplay of
Meleddu, M ., & Guicciardi, M . (1998). Self-knowledge self-deceptive styles with basic traits and motives.
and social desirability of personality traits. E u r o J o u r n a l o f P erson ality , 6 6 , 1 0 2 5 - 1 0 6 0 .
p ea n J o u r n a l o f P erson ality , 12, 1 5 1 - 1 6 8 . Perlini, A. H., Bertolissi, S., & Lind, D. L. (1999). The
Mendonca, J. D., Holden, R. R., Maz m anian, D., &c effects of women’s age and physical appearance on
Dolan, J. (1983). The influence of response style on evaluations of attractiveness and social desirability.
the Beck Hopelessness Scale. C a n a d ia n jo u r n a l o f J o u r n a l o f S o c ia l P sy ch olog y , 139, 3 4 3 - 3 5 4 .
B e h a v io u r a l S c ien c e, 15, 2 3 7 - 2 4 7 . Perlini, A. H., Marcello, A., Hansen, S. D., & Pudney,
Mesmer-Magnus, J ., Viswesvaran, C., Deshpande, S., W. (2001). The effects of male age and physical ap
& Joseph, J. ( 2 00 6 ). Social desirability: The role of pearance on evaluations of attractiveness, social de
over-claiming, self-esteem, and emotional intelli sirability and resourcefulness. S o c ia l B e h a v io r a n d
gence. P sy c h o lo g y S c ien ce, 4 8 , 3 3 6 - 3 5 6 . P erson ality , 2 9 , 2 7 7 - 2 8 7 .
Meston, C. M., Heiman, J. R., Trapnell, P. D., & Paul Piedmont, R. L., M cCrae, R. R., Riemann, R „ &c
hus, D. L. (1998). Socially desirable responding and Angleitner, A. ( 2 0 00 ). On the invalidity of valid
sexualitv self-reports. J o u r n a l o f S ex R esea rch , 35, ity scales: Evidence from self-reports and observer
14 8 -15 7. ratings in volunteer samples. J o u r n a l o f P erson ality
Middleton, K. L., & Jones, J. L. (20 0 0). Socially desir a n d S o c ia l P sy ch o lo g y , 78, 5 8 2 - 5 9 3 .
able response sets: The impact of country culture. Rauch, W. A., Schweizer, K., &c Moosbrugger, H.
P sy c h o lo g y a n d M arketin g , 17, 14 9-1 6 3 . (20 0 7). Method effects due to social desirability as
Miotto, P., De Coppi, M., Frezza, M ., Rossi, M ., & a parsimonious explanation of the deviation from
Preti, A. (2002). Social desirability and eating dis iinidimensionalitv in I.OT-R scores. P erson ality
orders: A community study of an Italian school- a n d In d iv id u a l D iffe r e n c e s , 4 2 , 15 97 -16 0 7.
aged sample. A cta P sy ch iatrica S c a n d in a v ic a , 105, Riketta, M. (2004). Does social desirability inflate the
372-377. correlation between self-esteem and anxiety? P sy
Morisky, D. E., Ang, A., & Sneed, C. D. (2 002). c h o lo g ic a l R e p o r ts , 94, 1 2 3 2 - 1 2 3 4 .
Validating the effects of social desirability on self- Riketta, M . (20 05 ). Gender and socially desirable re
reported condom use behavior among commercial sponding as moderators of the correlation between
sex workers. A ID S E d u c a tio n a n d P rev en tion , 14, implicit and explicit self-esteem. C u rren t R e se a rch
351-360. in S o c ia l P sy ch olog y , 11, 1 4 - 2 8 .
Motl, R. W., McAuley, E., & DiStefano, C. (2 005). Rogers, G. M ., Reinecke, M. A., & Setzer, N . J . (2004).
Is social desirability associated with self-reported Childhood attachment experience and adulthood
physical activity? P rev en tiv e M ed icin e: An I n te r n a cognitive vulnerability: Testing state dependence
tio n a l J o u r n a l D e v o te d to P ractice a n d T h eo r y , 4 0 , and social desirability hypotheses. J o u r n a l o f C o g
735-739. n itive P sy ch o th era p y , 18, 7 9 - 9 6 .
Nevid, J. S. (1983). Hopelessness, social desirability, Sheridan, C. L „ Mulhern, M . A., & M artin , D. (1999).
and construct validity. J o u r n a l o f C o n su ltin g a n d The role of social desirability, negative affectivity,
C lin ic a l P sy ch olog y , 51, 1 3 9 - 1 4 0 . and female reproductive system symptoms in differ
O ’Grady, K. E. (1988). T he M arlo w e-C row n e and ences in reporting symptoms by men and women.
Edwards social desirability scales: A psychometric P sy ch o lo g ica l R e p o rts , 8 5 , 5 4 - 6 2 .
perspective. M u ltiv a riate B e h a v io r a l R esea rch , 2 3 , Sloan, J. J ., Ill, Bodapati, M . R., & Tucker, T. A.
87-101. (20 0 4). Respondent misreporting of drug use in
Ones, D. S., Viswesvaran, C., & Reiss, A. D. (1996). self-reports: Social desirability and other correlates.
Role of socially desirable responding in personal J o u r n a l o f D ru g Issu es, 3 4 , 2 6 9 - 2 9 2 .
ity testing for personnel selection: The red herring. Smith, D., Driver, S., Lafterty, M ., Burrell, C., &
J o u r n a l o f A p p lied P sy ch olog y , 81, 6 6 0 - 6 7 9 . Devonport, T. (2001). Social desirability bias and
Paulhus, D. L. (1984). Two-component models of so direction modified Competitive State Anxiety In
cially desirable responding. J o u r n a l o f P erson ality ventory— 2. P ercep tu a l a n d M o to r S kills, 95, 9 4 5 -
a n d S o c ia l P sy ch olog y , 4 6 , 5 9 8 - 6 0 9 . 952.
Paulhus, D. L. (1991). Measurement and control of re Steinmetz, H. L. (1932). Measuring ability to fake o c
sponse bias. In J. P. Robinson, P. R. Shaver, & L. cupational interest. J o u r n a l o f A p p lied P sy ch olog y ,
S. Wrightsman (Eds.), M easu re o f p e r s o n a lity a n d 16, 1 2 3 - 1 3 0 .
s o c ia l p s y c h o lo g ic a l a ttitu d es (Vol. 1, pp. 17-59). Strosahl, K. D., Linehan, M. M ., & Chiles, J. A.
San Diego, CA: Academic Press. (1984). Will the real social desirability scale please
Paulhus, D. L. (1998). P au lhu s D e c e p tio n S cales stand up? Hopelessness, depression, social desir
( PDS): T h e B a la n c e d In v en to ry o f D es ira b le R e- ability, and the prediction of suicidal behavior.
sp o n d in g — 7 user's m a n u al. North Tonawanda, J o u r n a l o f C o n su ltin g a n d C lin ica l P sy ch o lo g y , 5 2,
NY: Multi-Health Systems. 449-457.
454 V. M O T IV A T IO N A L D IS P O S IT IO N S
Tan, J. A., & Hall, R. J . (20 05 ). The effects of social E d u c a tio n a l a n d P sy c h o lo g ic a l M ea su rem en t, 2 4,
desirability bias on applied measures of goal orien 551-562.
tation. P erso n a lity a n d In d iv id u a l D iffe r e n c e s , 3 8 , Willebrand, M ., Wikehult, B. R., & Ekselius, L. (2005).
1891-1902. Social desirability, psychological symptoms, and
Tilgner, L., Wertheim, E. H., &c Paxton, S. J . (2004). perceived health in burn injured patients. J o u r n a l o f
Effect of social desirability on adolescent girls' re N er v o u s a n d M en ta l D ise a se , 19 3, 8 2 0 - 8 2 4 .
sponses to an eating disorders prevention program. Williams, K., Paulhus, D. L., & Harms, P. (20 01 , June).
In te r n a tio n a l J o u r n a l o f E atin g D iso r d er s, .35, T h e o v e r-c la im in g q u e s tio n n a ir e : In v u ln e ra b le to
211-216. fa k in g a n d w arn in g a b o u t fo ils . Paper presented at
Watson, K., Baranowski, T., Thompson, D., Jago, the annual convention of the American Psychologi
R., Baranowski, J ., & Klesges, L.. M . (20 0 6). In cal Society, Toronto, On tario, Canada.
novative application of a multidimensional item Zettle, R. D., & Houghton, L. L. (1998). The relation
response model in assessing the influence of social ship between mathematics anxiety and social de
desirability on the pseudo-relationship between sirability as a function of gender. C o lle g e S tu d en t
self-efficacy and behavior. H ea lth E d u c a tio n R e J o u r n a l, 3 2 , 8 1 - 8 6 .
sea rc h , 2 1 , 8 5 - 9 7 . Zickar, M. J . , Gibby, R. E., & Robie, C. (20 04 ). Un
White, L. A., Young, M. C., & Rumsey, M. G. (2001). covering faking samples in applicant, incumbent,
ABLE implementation issue and related research. and experimental data sets: An application of
In J. P. Campbell & D. J. Knapp (Eds.), E x p lo rin g mixed-model item response theory. O rg a n iz a tio n a l
th e lim its o f p e r s o n n e l se le c tio n a n d c la s sific a tio n R e se a rch M e th o d s , 7, 1 6 8 - 1 9 0 .
(pp. 5 2 5 - 5 5 8 ) . Mahwah, NJ: Erlbaum. Zickar, M . J., & Robie, C. (1999). Modeling faking
Wiggins, J. S. (1964). Convergences among stylistic good on personality items: An item-level analysis.
response measures from objective personality tests. J o u r n a l o f A p p lied P sy ch olog y , 8 4 , 5 5 1 - 5 6 3 .
CHAPTER 31
Sensation Seeking
M a r v in Z u c k e r m a n
455
456 V. M O T IV A T IO N A L D IS P O S IT IO N S
tated factor was similar, but rotation identi analyses. One of the factors that consistently
fied four other factors: Thrill and Adventure emerged in a five-factor analysis was de
Seeking (TAS), Experience Seeking (ES), scribed as Impulsive Unsocialized Sensation
Disinhibition (Dis), and Boredom Suscepti Seeking (Zuckerman, 1994a).
bility (BS). The general scale and new scales Analyses of items within all of the scales
for the subfactors were included in the SSS was done to develop a five-factor test of
Form IV (Zuckerman, 1971). The subscales personality, the Zuckerman-Kuhlm an Per
are briefly described: sonality Questionnaire (ZK PQ ; Zucker
TAS describes the seeking of arousal man, 1994a, 2002b , 2 0 0 8 ; Zuckerman et
through extreme sports involving unusual al., 1993). One of the replicable factors was
sensations and risks, such as skydiving, called Impulsive Sensation Seeking (ImpSS),
scuba diving, or flying. M ost of the items because it included both impulsivity, in the
are expressed as intentions rather than ac form of spontaneous reactions without plan
tual experiences, for example, “I would like ning, and general sensation-seeking items
to. . . . ” reflecting the need for excitement and nov
ES is the seeking of novel experiences elty without the specification of particular
through the mind and senses, travel, music, activities. The 19-item scale may be scored
art, and people. It also represents social non for the two subfactors, impulsivity and sen
conformity and a desire to associate with sation seeking, as well as the total.
unconventional people. Aluja and colleagues (2006) developed a
Dis is the seeking of intense experiences in shortened (50-item) cross-cultural form of
parties, social drinking, and sex. Although the ZK PQ with English, Spanish, French,
some items describe behavioral preferences, and German versions. The ImpSS, like the
others are general attitudes, for example, “I other four subscales, contains only 10 items,
like to have new and exciting experiences but alpha reliability coefficients range from
even if they are a little unconventional or il .72 to .74.
legal.” Arnett (1994) developed a short form of
BS items involve an intolerance for repeti the SSS with items selected to represent two
tive experience and boring people and rest qualities of stimuli attractive to sensation
lessness when exposed to such conditions. seekers: novelty and intensity. There are 10
An example of a forced-choice item is: “A. items for each. Although the internal reli
The worst social sin is to be a bore; B. The abilities are relatively low, the scale has been
worst social sin is to be rude.” correlated with a variety of risky behaviors.
Form V contained subscales of 10 items, Hoyle, Stephenson, Palmgreen, Lorch, and
each selected to enhance increased conver Donohew (2002) devised a Brief Sensation-
gent validity within scales and discriminant Seeking Scale appropriate for young and
validity between scales. The substitution of older adolescents using two items from each
a Total score, sum of the four subscales, re of the four subscales of the SSS-V, avoiding
placed the general scale in Forms II and IV mention of alcohol or drugs. Only a total
(Zuckerman, Eysenck, & Eysenck, 1978). score is used. This scale has been used in
The four subfactors of sensation seeking drug prevention campaigns, as described in
have been generally replicated in studies of a subsequent section.
translated scales in many different countries Cloninger (1987b) is the only other per
(Zuckerman, 1994b). sonality theorist to include sensation seeking
Forms II, IV, and V of the SSS were devel as one of the major dimensions of personal
oped in isolation from other personality fac ity. He devised a scale called Novelty Seek
tors. Beginning in the late 1980s we began ing (NS). In its most recent form it includes
a series of factor analyses with the goal of subscales for exploratory excitability, impul
providing a trait classification for P sychobi siveness, extravagance, and disorderliness
ology o f Personality (Zuckerman, 1991). All (Cloninger, Svrakic, & Przybeck, 1993). The
of the subscales of SSS-V, as well as other shorter form of the NS scale correlates very
scales for sensation seeking, impulsivity, so highly with the ImpSS scale (Zuckerman &
ciability, socialization, activity, neuroticism, Cloninger, 1996) and highly with the SSS-V
and anxiety, were included in the factor Total score.
31. Sensation Seeking 457
Russo and colleagues (1993) constructed creased riskiness. The result is a cross of the
an SSS form appropriate for children ages approach-avoidance gradient, with avoid
9 -1 4 . Three subscales were developed from ance predominating earlier along the risk
an item factor analysis: Thrill and Adven continuum in low sensation seekers.
ture Seeking, Social Disinhibition, and Drug Volunteering for dangerous military or ci
and Alcohol Attitudes. The latter scale is a vilian assignments is also more characteristic
measure of a specific attitude rather than a of high sensation seekers. Israeli soldiers who
personality factor. volunteered for combat units were higher in
sensation seeking than other recruits. Many
of them engaged in scuba diving in their free
Risky Behavior time (Hobfoll, Rom, & Segal, 1989). Trait
anxiety was unrelated to these voluntary
Volunteering and R isk A ppraisal
choices. Among those who actually engaged
In the 1960s we were doing experiments in combat, those who received medals for
in sensory deprivation and hypnosis. We bravery were higher sensation seekers than
noticed that our paid volunteers for both other soldiers (Neria, Solomon, Ginzburg,
types of experiments looked like high sensa & Dekel, 2000).
tion seekers. We confirmed this impression,
asking for volunteers from students at three
Sociability, Relationships, and S ex
universities. The volunteers for both experi
ments scored higher on the SSS-II general In the “alternative five” model, Impulsive
scale than nonvolunteers. Subsequent stud Sensation Seeking (ImpSS) and Sociability
ies by others showed that volunteering by are independent factors (Zuckerman, 2002b).
high sensation seekers depended on the type At the trait level, sensation seeking (SSS-V)
of experiment. Any study offering a chance is only weakly correlated with Eysenck’s E
at an unusual or novel type of experience scale and ImpSS with the N EO Extraversion
(e.g., sensory deprivation, hypnosis, drugs, scale (Zuckerman, et al., 1993). However,
gambling, sensitivity training, brain wave in a situation of close confinement with a
control, training in transcendental medita stranger, low sensation seekers showed much
tion, and encounter groups) attracted high more stress and distress than high sensation
sensation seekers. However, sensation seek seekers (Zuckerman, Perskv, Link, & Basu,
ing was unrelated to volunteering for more 1968). High sensation seekers engage in self
mundane experiments in learning or social disclosure with both casual and close friends
psychology. Subsequent research showed (Franken, Gibson, & Mohan, 1990). In an
that risk appraisal and expectations of ex interview, high sensation seekers show more
periencing anxiety versus expectations of spontaneous interactive behavior: fast reac
experiencing positive arousal influenced the tions, eye gaze, posture, vocalization, smiles,
approach or avoidance behavior related to and laughter (Cappella & Green, 1984).
volunteering. Comparing the SSS with Hendrick and
Risk appraisal across a variety of activi Hendrick’s (1986) scales for love styles, the
ties, even those never experienced, is nega Total SSS was positively related to the Ludus
tively related to sensation seeking (Zucker style (playful, less committed love) and neg
man, 1979, 1994b, 20 0 7 ). High sensation atively related to the Pragma style (rational
seekers tend to rate riskiness lower than appraisal of mate’s potential for a long-term
low sensation seekers. But even in experi relationship). The SSS Dis and BS subscales
ments in which risk appraisal is the same in had the highest correlations with Ludus.
high and low sensation seekers, the anxiety Thornquist, Zuckerman, and Exline
gradient rises more steeply and the antici (1991) studied the relations between the SSS
pation of positive affect (elation) decreases and relationships among unmarried het
with appraised riskiness of the experiment erosexual partners in current relationships.
in low sensation seekers, whereas in high They found moderate positive correlations
sensation seekers the slope of the anxiety between SSS scores of partners, a finding
gradient is less steep, and the anticipated similar to those found for married couples,
positive emotion does not decline with in indicating a role for the sensation-seeking
458 V. M O T IV A T IO N A L D IS P O S IT IO N S
trait in assortative mating. In both partners and effective treatm ents for the prevalent
Total SSS was negatively related to relation sexually transm itted diseases. This com
ship satisfaction, liking, and loving and placency changed in the 1980s with the
positively related to consideration of alter AIDs epidemic.
natives outside of the relationship. The most Hoyle, Fejfar, and M iller (2000) reviewed
satisfactory relationships were between two all studies relating personality traits to sex
low sensation seekers, and the least satisfac ual risk taking through 1999. O f all tests
tory were between two high sensation seek used in these studies, the SSS had the highest
ers. When one partner is low and the other correlations with risky sex, including num
is high on sensation seeking, there is usually ber of partners, unprotected sex (not using
conflict, not only about sex but in general condoms), and high-risk sex with strang
preferences for life activities and choice of ers. Zuckerman (2007) reviewed subsequent
friends. studies from 2 0 0 0 to 2 0 0 4 and found that
Nonverbal interactions between members the SSS and other versions of the scale pre
of a pair were also studied. Women high in dicted risky sexual behavior in large-scale
sensation seeking looked more at their part studies of young adolescent, college, and
ners and spoke more to them. Mutual gaze community populations. In adolescents, sen
correlated with the woman’s SSS when she sation seeking and impulsivity were related
was speaking but did not when the male was to the use of alcohol or drugs before having
speaking. Women high in sensation seeking sex.
tend to command attention from their mates Zuckerman (2007) also reviewed studies
when they are speaking but do not necessar of risky sexual behavior among gay men.
ily give it when their mates are speaking. Many of these studies were conducted by
The negative relation between sensation Kalichman and his colleagues (Kalichman,
seeking and relationship satisfaction is also Heckman, & Kelly, 1996), who developed
found for married or cohabiting couples scales for sexual sensation seeking and non-
(Schroth, 1991). Divorced men have higher sexual sensation seeking. M ost studies show
scores on the SSS than both married and sin that both the specific and general types of
gle men, and divorced women have higher sensation seeking are related to number of
scores than married women (Zuckerman &c sexual partners, to anal sex without using
Neeb, 1980). condoms, and to using alcohol and drugs be
In the general population there are high fore and during sex. The direct effect of sen
correlations between SSS scores of spouses, sation seeking accounts for 80% of the asso
indicating a high degree of assortative mat ciation with unprotected anal sex among gay
ing not typically found for other personal men, whereas alcohol and drug use account
ity traits. However, the correlations between for only 20% (Kalichman et al., 1996). Sen
couples entering couple therapy are lower sation seeking accounted for the relationship
than those of happily married couples (Fich- between substance use and high-risk sexual
er, Zuckerman, & Neeb, 1981; Ficher, Zuck behavior in gay men.
erman, & Steinberg, 1988). Low sensation-
seeking scores in the male partners relative
Sm oking (Tobacco), D rin k in g , and D rugs
to the female’s scores were related to sexual
dissatisfaction in both partners. The reverse Several studies we conducted in a college
was not necessarily true. Donaldson (1989) population during the 1970s and 1980s
reported similar findings among married showed that smokers were higher in sensa
and intimate college students. tion seeking than nonsmokers. Although the
Studies relating the SSS to sexual a t prevalence of smoking had diminished in
titudes and behavior in college students males (but not in females) during the 1980s,
conducted during the 1970s showed that the relationship between sensation seeking
sensation seeking was related to sexually and smoking was still a significant one. An
permissive attitudes and variety of sexu other study conducted at a nearby univer
al experience and partners (Zuckerm an, sity in 1993 showed the same differences
1994b , 2 0 0 7 ). There seemed to be less risk between smoking prevalence in participants
related to impulsive and promiscuous sex high and low in sensation seeking (Kuman,
in the 1 970s because of birth control pills Pekala, & Cummings, 1993).
31. Sensation Seeking 459
Between 1995 and 2003 six large-sample lant and depressant drugs, including alcohol,
studies of middle and high school students to suit their moods and situational needs.
showed relationships between various The SSS-Total and ImpSS are predictors of
sensation-seeking scales, including ImpSS, negative outcome in the treatment of cocaine
and smoking prevalence and prediction (see abuse (Ball, 1995; Patkar et al., 2004).
Zuckerman, 2007, Table 4.1). Zuckerman A series of laboratory and community
and Kuhlman (2000) found that smoking, studies showed the effectiveness of sensation-
drinking, drugs, and risky sex were all in seeking theory in designing communications
tercorrelated and correlated with the ZKPQ for the prevention and reduction of mari
ImpSS. juana use in adolescent populations (Dono-
Early studies of drinking in college students hew, Bardo, & Zimmerman, 2004). The
found relationships between heavy drinking laboratory studies showed that messages
and sensation seeking. These findings have with high-stimulation characteristics (novel,
been supported in more recent studies of complex, intense, exciting, fast paced) were
high school and college students and the gen more effective than low-stimulation messag
eral population in several countries. es in increasing intent to call a drug hotline,
Most heavy drinking by college students particularly in high sensation seekers using
occurs in social situations in which social drugs. A campaign using television ads with
disinhibition is a primary motive for both high-stimulation values reduced overall
sexes. Sexual disinhibition is another mo marijuana usage in two counties following
tive for men (Beck, Thombs, Mahoney, & 4-month campaigns (Palmgreen, Donohew,
Fingar, 1995). Katz, Fromme, and D ’Amico Lorch, Hoyle, & Stephenson, 2001).
(2000) found that positive social and sexu
al expectancies were positively related and
E ffect o f Social Predictors o f Sensation
that risk expectancy was negatively related
Seeking on the Trait
to sensation seeking. However, the relation
ship between sensation seeking and heavy Does sensation seeking increase the likeli
drinking was not mediated by expectancies. hood of drug use, or does drug use increase
Sensation or novelty seeking is high in alco sensation seeking? The causal relationships
holics, particularly Cloninger’s (1987a) Type between social-psychological variables and
2 alcoholic, characterized by early age of sensation seeking are best investigated in
onset, antisocial personality, and low harm longitudinal studies. Stacy, Newcomb, and
avoidance. Bentler (1991) studied such predictors in a
Marijuana and other drug use is even 9-year longitudinal study of sensation seek
more highly related to sensation seeking ing from adolescence to adulthood. In addi
than is drinking in high school and col tion to adolescent sensation seeking itself as
lege populations. These findings, beginning a predictor of adult levels of the trait, other
in the 1970s, have persisted into this cen predictors involved social conformity, emo
tury (e.g., Wagner, 2001). Jaffe and Archer tional distress, drug use, social support, and
(1987) compared the SSS and other person peer deviance. Only the sensation-seeking
ality scales in prediction power for the use of factor in adolescence directly predicted the
drugs in a college population. The SSS was same factor in adulthood using structural
the most powerful predictor of the use of 7 equation modeling. However, some social
of the 10 classes of drug and polydrug use. predictors, particularly social support, pre
Substance abusers were low in anxiety sen dicted specific subscales of sensation seek
sitivity. Within the drug-using community, ing. Lower social support during adoles
polydrug use is more highly related to sensa cence predicted high ES and BS in adulthood
tion seeking than the use of any one specific beyond the prediction from the adolescent
type of drug, although there is some pref levels of these subtraits. O f course, social
erence for stimulants and psychedelic drugs support may represent the familial reactions
among polydrug abusers. The qualities of to earlier manifestations of the nonconform
intensity and novelty are provided by these ing behavior involved in ES. Family support
drugs, in addition to the disinhibition also may moderate the effects of adolescent sen
produced by alcohol and depressant drugs. sation seeking and its persistence or nonper
Many polydrug users alternate or mix stimu sistence into adulthood.
460 V. M O T IV A T IO N A L D IS P O S IT IO N S
working in emergency rooms, and rape cri Sum m ary: R isk Taking
sis counselors had higher SSS scores than
Sensation seeking was defined as the willing
control groups in less risky or stressful jobs.
ness to take risks for the sake of the reward
Other vocational groups, such as naval fliers
of arousing stimulation. The acceptance of
and naval divers, scored higher on TAS but
physical risk by high sensation seekers is
actually lower than controls on ES and Dis,
shown in their driving behavior and their
perhaps as a function of the lower social de
sirability of the latter. engagement in risky sports and in risky vo
Within some groups some members have cations and activities within vocations. Such
behavior in prosocial vocations is often ad
higher scores related to their risky behavior.
mirable, but the pursuit of sensation in crim
For instance, Israeli soldiers who were dec
inal antisocial activities is the darker side of
orated for bravery in the 1973 war scored
higher on a short form of the SSS than others the trait.
who fought in the war (Neria et al., 2000). Risky sensation seeking is the outcome of
Within a group of police patrolmen scoring a conflict between the anticipation of plea
average on the SSS and a general risk scale, sure from a risky activity and the anticipa
those engaging in high-speed chases scored tion of anxiety related to the perceived risk.
higher on both of these scales (Homant, High sensation seekers have a stronger an
Kennedy, & Howton, 1994). ticipation of reward and a weaker anticipa
tion of anxiety or punishment and therefore
are more likely to engage in these activities,
A ntisocial, C rim inal, whereas low sensation seekers are more like
and D elinq uent Behavior ly to avoid them.
In the study by Goma-i-Freixanet (1995),
violent criminals scored significantly higher
on the Total and all of the SSS subscales Preferences in A rt, M edia, and Music
than participants in risky sports, risky vo
cational, and control groups. In the female Sensation seeking is not related only to risky
groups the criminals scored higher than the behavior. Entertainment preferences reflect
risky vocational and control groups on the the qualities of stimulation in the defini
SSS Total, ES, and Dis scales, but not the fe tion of the trait: novelty, intensity, complex
male risky sports group (Goma-i-Freixanet, ity, and variety (Zuckerman, 1994b, 2006).
2001). Imprisoned male delinquents scored High sensation seekers show a stronger ori
higher than controls on ES and BS but not enting reflex, a physiological (skin conduc
TAS or Dis scales (Romero, Luengo, & So- tance, heart rate) response to novel stimuli
bral, 2001). However, other studies have on their first presentations but not on rep
shown that Dis in early adolescence predicts etition. They are particularly responsive to
delinquent behavior in later adolescence stimuli that are interesting to them and in
(Newcomb & M cGee, 1991; W hite, Labou- tense. Their cortical evoked response is aug
vie, &C Bates, 1985). mented by high-intensity stimulation, where
Horvath and Zuckerman (1993) found that as evoked responses of low sensation seekers
SSS Total was highly correlated with self- tend to be reduced by intense stimuli.
reports (made anonymously) of delinquent High sensation seekers like novel and
and criminal behavior in college students. complex designs, whereas lows like simple
Sensation seeking and criminal behavior also and symmetrical ones. High sensation seek
correlated with risk appraisals for criminal ers like paintings with high tension and
behavior, but risk appraisal did not medi semiabstract or abstract styles and Pop art.
ate the relationship between sensation seek They also like erotic and violent themes in
ing and criminal behavior. Instead sensation art or photographs. Low sensation seekers
seeking mediated the relationship between like paintings with low tension and realistic
risk appraisal and criminal behavior. It is styles and dislike violent or morbid themes
probable that engaging in some kind of crim in art.
inal behavior without getting caught lowers Similarly, high sensation seekers enjoy
the perceived riskiness of the behavior. and go to movies with horror and explicit
462 V. M O T IV A T IO N A L D IS P O S IT IO N S
sexual themes. In television the high sensa been largely at the psychobiological level
tion seekers like violent action-adventure and involve comparative models (Zucker
programs, whereas the lows like game shows man, 1 9 8 4 ,1 9 9 1 ,1 9 9 4 b ,1995, 1996, 2 0 0 3 ,
and news programs. Given a chance to select 2 0 0 5 , 2007). A full account of this theory
programs, high sensation seekers tend to do and research can be found in these refer
more channel switching, showing a need for ences.
variety and a boredom susceptibility. Genetic twin studies of sensation seek
Sensation seeking correlates positively ing including separated twins have shown a
with a liking for rock music, particularly high degree of heritability for the SSS and
hard rock, and negatively with a liking for most of its subscales (Zuckerman, 2002a).
religious and bland soundtrack music. How A specific gene for the dopamine 4 recep
ever, those who score high on the ES sub tor (D R D 4) was found to be associated
scale like a broader variety of music, includ with novelty seeking, a scale highly cor
ing folk and classical, in addition to hard related with ImpSS (Ebstein et al., 1996).
and soft rock. Intensity of sound, particu Replications have been inconsistent, but a
larly the drums and loud bass of the ampli meta-analysis shows a small but significant
fied guitars, dissonance, and “edginess” of effect comparing the long allele form with
lyrics, are qualities of music appreciated by the shorter form (Schinka, Letsch, 8c Craw
high sensation seekers. ford, 2002). W hat is interesting is that this
form of the gene is also associated with the
strength of the orienting reflex response to
Psychobiology o f Sensation Seeking novel stimuli in infants, with heroin and
alcohol abuse and gambling in adults and
The explanation for differences in impul with attention-deficit hyperactivity disorder
sive sensation seeking is a psychobiological in children. Other genes are also involved,
one involving genetics, psychophysiology, additively and interactively, in the personal
psychopharmacology, and neuropsychology ity trait and forms of psychopathology. This
(Zuckerman, 1994b, 1995, 2 0 0 5 ). Accord may explain why associations are weak and
ing to this model, the attraction to novel sometimes difficult to replicate.
and intense sensations and activities is a Insights into the neurochemical basis of
function of genetic predispositions toward sensation seeking have been derived from
impulsive approach based on dopaminergic comparisons of selectively bred strains of
activity in the reward pathways in the lim exploratory and novelty-reactive rats and
bic brain, dysregulated by low levels of the those who are more neophobic (Dellu, Pi
enzyme monoamine oxidase type B. The azza, Mayo, LeM oal, 8c Simon, 1996; Siegel
hormone testosterone also potentiates the 8c Driscoll, 1996). The exploratory strains
approach mechanism. A weak behavioral have more dopamine in the nucleus accum-
inhibitory mechanism is a function of in bens and are more dopaminergic reactive to
sensitivity of serotonergic inhibitory sys stimulant drugs or stress. The nucleus ac-
tems. Weak arousal or fearlessness in the cumbens is a reward nucleus in the medial
face of risk is a function of attenuated no forebrain bundle hypothesized to be more
radrenergic reactivity. Sensation seeking is reactive in human sensation seekers. More
affected by the balance between the three inhibited rats respond to novelty and stress
behavioral mechanisms (approach, inhibi with increases in serotonin, an inhibitory
tion, and arousal) and the biological sys neurotransmitter acting antagonistically to
tems underlying them. This might be called dopamine.
the “three-monoamine theory,” as contrast Studies of brain neurochemistry in hu
ed with theories only involving one mono mans rely largely on metabolites of the neu
amine for impulsivity and sensation seeking rotransmitters in cerebrospinal fluid (CSF)
(e.g., Cloninger, 1987b). or on blood or hormonal indices of reactivity
As we have seen, cognitive phenomena of the neurotransmitter to stimulants. Basal
such as expectations and risk appraisal are levels of metabolites have shown little rela
involved in the relationships of the trait with tionship to sensation and novelty seeking.
some of the behavioral phenomena, but not However, responses to neurochemical ago
always as mediators. My explanations have nists show a greater serotonergic reactivity
31. Sensation Seeking 463
in low sensation seekers and blunted reactiv factor measure of personality in cocaine abusers.
ity in high sensation seekers (see Table 1.4 P sy c h o lo g ic a l A ssessm en t, 7, 1 4 8 - 1 5 4 .
Beck, K. H., Thombs, D. I.., Mahoney, C. A., &C Fingar,
in Zuckerman, 20 0 7 ). The findings on rela K. M . (1995). Social context and sensation seeking:
tionships with dopaminergic reactivity have Gender differences in college student drinking m o
been mixed. Recent brain imaging studies, tivations. I n te r n a tio n a l J o u r n a l o f th e A d d ictio n s,
however, have shown greater reactivity of the 3 0 , I 101- 1115.
Boileau, I., Dagher, A., Leyton, M ., Gunn, R. N.,
dopaminergic system in the ventral striatum Baker, G. B., Diksic, M., et al. ( 2 0 06 ). Modeling
or nucleus accumbens and greater dopamin sensitization to stimulants in humans. A rch iv es o f
ergic reactivity and sensitization to amphet G e n e r a l P sy ch iatry , 63, 1 3 8 6 - 1 3 9 5 .
amine in high novelty seekers (Boileau et al., Burns, P. C., & Wilde, G. S. (1995). Risk taking in
male taxi drivers: Relationships among personality,
2 0 0 6 ; Leyton et al., 2002). These tentative
observational data and driver records. P erson ality
results support the three-monoamine model a n d In d iv id u a l D iffer en c es , 18, 2 6 7 - 2 7 8 .
of sensation seeking (Zuckerman, 1995) and Cappella, J. N., & Green, J . O. (1984). The effects of
explain the attraction of stimulant drugs for distance and individual differences in arousability
high sensation seekers. on nonverbal involvement: A test of discrepancy
arousal theory, jo u r n a l o f N o n v er b a l B eh a v io r, 8,
259-286.
Cloninger, C. R. (1987a). Neurogenic adaptive mecha
Conclusions nisms in alcoholism. S cien ce, 2 3 6 , 4 1 0 - 4 1 6 .
Cloninger, C. R. (1987b). A systematic method for
clinical description and classification of personal
Sensation seeking was a construct designed
ity variants. A rch iv es o f G e n e ra l P sychiatry, 4 4 ,
to predict a narrow range of phenomena in an 573-588.
experimental situation (sensory deprivation). Cloninger, C. R ., Svrakic, D. M . , & Przybeck, T. R.
Unexpectedly, it was found to be applicable (1993). A psychobiological model of temperament
to a wide variety of behavioral preferences, and character. A rch iv es o f G e n e r a l P sychiatry, SO,
975-990.
and the scales used to measure it showed a Dellu, F., Piazza, P. V., Mayo, W., L eM oal, M ., &
broad construct validity. The construct in Simon, H. (1996). Novelty-seeking in rats: Biobe-
volves a tendency to approach and engage havioural characteristics and possible relationship
in novel and arousing activities and weak with the sensation seeking trait in man. N e u r o p s y
c h o b io lo g y , 3 4 , 1 3 6 - 1 4 5 .
inhibition or avoidance associated with the Donaldson, S. (1989). Similarity in sensation seeking,
perceived riskiness of these activities. The sexual satisfaction, and contentment in relationship
balance between approach and avoidance in in heterosexual couples. P sy c h o lo g ic a l R e p o rts , 6 4 ,
these areas of voluntary risk taking is based 405-406.
on genetic and biological determinants and Donohew, L., Bardo, M. T., & Zim merman, R. S.
(2 004). Personality and risky behavior: Co mmun i
their interactions with environmental oppor cation and prevention. In R. M. Stelmack (Ed.), O n
tunities and influences. The trait of sensation th e p s y c h o b io lo g y o f p e r s o n a lity : E ssays in h o n o r
seeking in humans has been connected with o f M arvin Z u c k e rm a n (pp. 2 2 3 - 2 3 5 ) . Oxford, UK:
explorativeness and novelty seeking in other Elsevier.
Ebstein, R. P., Novick, O., Umansky, R., Priel, B.,
species through shared biological markers
Osher, Y., Blaine, D., et al. (1996). Dopamine D 4
suggesting evolutionary origins. Sensation receptor (D 4D R ) exon III polymorphism associated
seeking at some optimal level ensured gene with the human personality trait of novelty seeking.
survival: Too little and you starved or failed N atu re G e n e tic s , 12, 7 8 - 8 0 .
to mate; too much and you met a premature Ficher, 1. V., Zuckerman, M ., & Neeb, M. (1981).
Marital compatibility in sensation seeking trait as
death. a factor in marital adjustment. J o u r n a l o f S ex a n d
M a r ita l T h era p y , 7, 6 0 - 6 9 .
Ficher, I. V., Zuckerman, M ., & Steinberg, M. (1988).
R eferen ces Sensation seeking congruence in couples as a deter
minant of marital adjustment: A partial replication
Aluja, A., Rossier, J., Garcia, L. F., Angleitner, A., and extension. J o u r n a l o f C lin ica l P sy ch olog y , 4 4 ,
Kuhlman, M ., &c Zuckerman, M. (200 6). A cross- 803-809.
cultural shortened form of the Z K P Q (Z K PQ -5 0-cc) Franken, R. E., Gibson, K. J ., & Mohan, P. (1990).
adapted to English, French, Germ an, and Spanish Sensation seeking and disclosure to close and casual
languages. P erson ality a n d In d iv id u a l D iffer en c es , friends. P erson ality a n d In d iv id u a l D iffe r e n c e s , 11,
41, 6 1 9 - 6 2 8 . 829-832.
Arnett, J. (1994). Sensation seeking: A new concep Goma-i-Freixanet, M . (1995). Prosocial and antisocial
tualization and a new scale. P erson ality a n d In d i aspects of personality. P erson ality a n d In d iv id u a l
v id u a l D iffer en c es , 26, 2 8 9 - 2 9 6 . D iffer en c es , 19, 1 2 5 - 1 3 4 .
Ball, S. A. (1995). T he validity of the alternative five- Goma-i-Freixanet, M. (2001). Prosocial and antiso
464 V. M O T IV A T IO N A L D IS P O S IT IO N S
cial aspects of personality in women: A replication erans. P erso n a lity a n d In d iv id u a l D iffe r e n c e s , 29,
study. P erso n a lity a n d In d iv id u a l D iffe r e n c e s , 3 0 , 921-932.
1 4 0 1 -1 4 1 1 . Newcomb, M . D., 8c M cGee , L. (1991). Influence
Goma-i-Freixanet, M . (200 4). Sensation seeking and of sensation seeking on general deviance and spe
participation in physical risk sports. In R. M . Stel- cific problem behaviors from adolescence to young
mack (Ed.), O n th e p s y c h o b io lo g y o f p e r so n a lity : adulthood. J o u r n a l o f P erson ality a n d S o c ia l P sy
E ssa y s in h o n o r o f M arvin Z u c k e r m a n (pp. 1 8 5 - c h o lo g y , 61, 6 1 4 - 6 2 8 .
201). Oxford, UK: Elsevier. Palmgreen, P., Donohew, L ., Lorch, E. P., Hoyle, R.
Heino, A. (1996). R isk ta k in g in c a r d riv in g : P erce p H., & Stephenson, M . T. (2001). Television c am
tion s, in d iv id u a l d iffe r e n c e s a n d e ffe c ts o f sa fety paigns and adolescent marijuana use: Tests of
in cen tiv es. Unpublished doctoral dissertation, Uni sensation-seeking targeting. A m er ica n J o u r n a l o f
versity of Groningen. P u blic H ea lth , 91, 2 9 2 - 2 9 6 .
Hendrick, C., & Hendrick, S. S. (1986). A theory and Patkar, A. A., Murray, H. W., Mannelli, P., Gottheil,
method of love. J o u r n a l o f P erson ality a n d S o cia l E., Weinstein, S. P., 8c Vergare, M . J. (20 04 ). Pre
P sy c h o lo g y , SO, 3 9 2 - 4 0 2 . treatment measures of impulsivity, aggression, and
Hobfoll, S. E., Rom, T., & Segal, B. (1989). Sensa sensation seeking are associated with treatment out
tion seeking, anxiety, and risk-taking in the Israeli come for African-American cocaine dependent pa
context. In S. Ebstein (Fid.), D rugs a n d a lc o h o l u se: tients. J o u r n a l o f A d d ictiv e D ise a se s, 2 3 , 1 0 9 - 1 2 2 .
Issu es a n d fa c ts (pp. 5 3 - 5 9 ) . New York: Plenum Romero, E., Luengo, A., & Sobral, J . (2001). Personal
Press. ity and antisocial behavior: Study of temperamental
Homant, R. J ., Kennedy, D. B., 8c Howton, J. D. dimensions. P erso n a lity a n d In d iv id u a l D iffe r e n c
(1994). Risk taking and police pursuit. J o u r n a l o f es, 3 1, 3 2 9 - 3 4 8 .
S o c ia l P sy ch olog y , 1 3 4 , 2 1 3 - 2 2 1 . Russo, M. F., Stokes, G. S., Lahey, B. B., Christ, M.
Horvath, P., & Zuckerman, M . (1993). Sensation seek A. G., McBurnett, K., Loeber, R., et al. (1993). A
ing, risk appraisal, and risky behavior. P erson ality sensation-seeking scale in children: Further refine
a n d In d iv id u a l D iffe r e n c e s , 14, 4 1 - 5 1 . ment and psychometric development. J o u r n a l o f
Hoyle, R., Fejfar, M. C., & Miller, J. D. ( 2 0 00 ). Per P s y c h o p a th o lo g y a n d B e h a v io r a l A ssessm en t, 15,
sonality and sexual risk taking: A quantitative re 69-86.
view. J o u r n a l o f P erson ality , 6 8 , 1 2 0 3 - 1 2 3 1 . Schinka, J . A., I.etsch, E. A., 8c Crawford, F. C.
Hoyle, R., Stephenson, M . T., Palmgreen, P., Lorch, E. (2002). D R D 4 and novelty seeking: Results of
P., 8c Donohew, R. L. (2 002). Reliability and valid meta-analyses. A m er ica n J o u r n a l o f M e d ic a l G e
ity of a brief measure of sensation seeking. P ers o n n etics, 114, 6 4 3 - 6 4 8 .
a lity a n d In d iv id u a l D iffer en c es , 3 2 , 4 0 1 - 4 1 4 . Schroth, M . L. (1991). Dyadic adjustment and
Ja ck , S. J., &c Ronan, K. R. (1998). Sensation seek sensation-seeking compatibility. P erson ality a n d
ing among high- and low-risk sports participants. In d iv id u a l D iffe r e n c e s , 1 2 , 4 6 7 - 4 7 1 .
P erson ality a n d In d iv id u a l D iffe r e n c e s , 2 5 , 1 0 6 3 - Siegel, J . , 8c Driscoll, P. (1996). Recent developments
1083. in an animal model of visual evoked potential aug
Ja ffe, L. T., 8c Archer, R. P. (1987). The prediction menting/reducing and sensation-seeking behavior.
of drug use among college students from M M P I , N e u r o p s y c h o b io lo g y , 3 4 , 1 3 0 - 1 3 5 .
M C M I , and sensation-seeking scales. J o u r n a l o f Stacy, A. W., Newcomb, M. D., 8c Bentler, P. M.
P erson ality , 5 1 , 2 4 3 - 2 5 3 . (1991). Social psychological influences on sensation
Jo n ah, B. A. (1 997). Sensation seeking and risky driv seeking from adolescence to adulthood. P erson ality
ing: A review and synthesis of the literature. A c c i a n d S o c ia l P sy ch o lo g y B u lletin , 17, 7 0 1 - 7 0 8 .
d e n t A n aly sis a n d P rev en tio n , 2 9, 6 5 1 - 6 6 5 . Thornquist, M . H., Zuckerman, M ., 8c Exline, R. V.
Kalichman, S. C., Heckman, T., & Kelly, J. A. (1996). (1991). Loving, liking, looking and sensation seek
Sensation seeking as an explanation for the associa ing in unmarried college couples. P erson ality a n d
tion between substance use and HIV-related risky In d iv id u a l D iffe r e n c e s , 12, 1 2 8 3 - 1 2 9 2 .
sexual behavior. A rch iv es o f S ex u a l B eh a v io r, 2 5 , Wagner, M. K. (2001). Behavioral characteristics re
141-154. lated to substance abuse and risk-taking, sensation
Katz, E. C., Fromme, K., 8c D ’Amico, E. J. (20 0 0). Ef seeking, and self-reinforcement. A d d ic tiv e B e h a v
fects of outcome expectancies and personality on iors, 2 6 , 1 1 5 - 1 2 0 .
young adults’ illicit drug use, heavy drinking and Whissel, R. W., 8c Bigelow, B. J. (2003). The speeding
risky sexual behavior. C o g n itiv e T h era p y a n d R e attitude scale and the role of sensation seeking in
sea rc h , 2 4 , 1 - 2 2 . profiling young drivers at risk. R is k A n alysis, 2 3 ,
Kuman, V. K., Pekala, R. J ., & Cummings, J. (1993). 811-820.
Sensation seeking, drug use, and reported paranor White, W. R ., Labouvie, E. W., 8c Bates, M. E. (1985).
mal beliefs and experiences. P erson ality a n d In d i T he relationship between sensation seeking and de
v id u a l D iffer en c es , 14, 6 8 5 - 6 9 1 . linquency: A longitudinal analysis. J o u r n a l o f R e
Leyton, M ., Boileau, I., Benkelfat, C., Diksic, M., sea rch in C r im e a n d D elin q u en cy , 2 2 , 1 9 7 -2 11 .
Baker, G., 8c Dagher, A. (2002). Amphetamine Zuckerman, M . (1969). Theoretical formulations: I. In
induced increases in extracellular dopamine, drug J. P. Zubek (Ed.) S en so ry d e p r iv a tio n : f i f t e e n y ears
wanting, and novelty seeking. N e u r o p s y c h o p h a r o f resea rc h (pp. 4 0 7 - 4 3 2 ) . New York: Appleton-
m a c o lo g y , 27, 1 0 2 7 - 1 0 3 5 . Century-Crofts.
Neria, Y., Solomon, Z ., Ginzburg, K., 8c Dekel, R. Zuckerman, M. (1971). Dimensions of sensation seek
(200 0). Sensation seeking, wartime performance, ing. J o u r n a l o f C o n su ltin g a n d C lin ica l P sy ch olog y ,
and long-term adjustment among Israeli war vet 36, 45-52.
31. Sensation Seeking 465
Rejection Sensitivity
R a in e r R o m e r o - C a n y a s
Va n e s s a T . A n d e r s o n
K a v it a S. R e d d y
G e r a l d in e D o w n e y
466
32. R eje ctio n S ensitivity 467
goal, and a protective system such as rejec nificant other. This assumption is reflected
tion sensitivity may develop to serve it. We in the operationalization of rejection sensi
view rejection sensitivity as a defensively tivity in the Rejection Sensitivity Question
motivated system that develops through re naire (RSQ ; Downey &c Feldman, 1996). The
jection experiences to defend people against RSQ presents a series of interpersonal situ
rejection while maintaining connection with ations, identified through extensive qualita
the source of the threat. The adaptive value tive pilot work, in which people make a re
of rejection sensitivity is its ability to trig quest of someone who is important to them.
ger quick defensive responses under threat In addition, researchers have developed and
conditions. However, the system becomes validated measures of rejection sensitivity
maladaptive if activated in situations that re tailored for particular populations, resulting
quire reflective, tactical behavior, when the in measures that reflect specific types of situ
threat is minimal, or when efforts to prevent ations in which rejection concerns are likely
rejection undermine other personal goals. to be activated. These include measures of
Several strands of evidence suggest that sensitivity to rejection for personal reasons
the rejection-sensitivity dynamic operates tailored to college students, community
as a defensive, affectively based system that adults, middle-school children, and incar
evolved to guide rapid and intense responses cerated women and measures of sensitivity
to threats of danger (Davis, 1992; LeDoux, to rejection because of status characteristics
1996). When rejection is the danger, activa including race (Chan & Mendoza-Denton,
tion of the rejection-sensitivity system ori 2 0 0 8 ; Mendoza-Denton, Downey, Purdie,
ents and prepares the individual to detect Davis, & Pietrzak, 2002), gender (London,
signs of social danger and to be ready to act Downey, Rattan, & Tyson, 20 0 6 ), physical
to avert the danger, escape, or strike out in appearance (Park, 2007a), and sexual orien
self-defense. This helps to explain why the tation (Pachankis, Goldfried, & Ram rattan,
behavior of highly rejection-sensitive people 2008).
can include a mixture of accommodation, A sample situation from the personal
withdrawal, and intense hostility and draws RSQ states: “You approach a close friend
attention to the need to account for the spe to talk after doing something that seriously
cific contexts in which each type of behavior upset him/her.” Respondents indicate their
will emerge. expectations of rejection (e.g., by rating the
We devote this chapter to an overview of degree to which “1 would expect that he/she
the impact of rejection sensitivity on peo would want to talk with me to try to work
ple’s behavior, an overview that also pres things out”) and their concern or anxiety
ents evidence for the view that the rejection- about the outcome (e.g., “How concerned
sensitivity dynamic operates as a defensive or anxious would you be over whether or
motivational system. We summarize work not your friend would want to talk with
on sensitivity to interpersonal rejection gen you?”). The level of anxious expectations
erally, as well as work on sensitivity to rejec in each situation is calculated by multiply
tion that is based on specific characteristics ing the degree of concern by the level of ex
or social identities. We begin with a descrip pectation of rejection (which, incidentally,
tion of how the dynamic is measured in re did not covary in a large validation sample;
search on rejection sensitivity. Downey & Feldman, 1996). This operation
alization captures the view that rejection
sensitivity is a “hot cognition” (Metcalfe
M easurem ent & Mischel, 1999) that is activated in situ
ations of threat. Highly rejection-sensitive
Given the dynamic conceptualization of re individuals do not merely expect rejection
jection sensitivity, individual differences in (as, e.g., telephone solicitors do) but also feel
rejection sensitivity should be most evident threatened by the possibility of rejection (as
in situations in which rejection by important telephone solicitors apparently do not). In
others is a possibility. Anxious expectations contrast, people low in rejection sensitivity
of rejection are at the core of rejection sen may tend to expect acceptance and/or to be
sitivity and are particularly likely to be acti less concerned about the possibility of rejec
vated when the person is dependent on a sig tion. RSQ scores are calculated by averaging
32. R e je ctio n S ensitivity 469
the computed RS levels across the situations tuned to signs of threat (Downey, Mougios,
that constitute the measure. RSQ scores are et al., 2004).
approximately normally distributed and In social interactions, highly rejection-
show a stable one-factor structure, good in sensitive people are vigilant for cues of the
ternal and test-retest reliability (Downey & relevant threat (rejection) in the same way
Feldman, 1996), and discriminant validity that people with severe phobias are vigilant
in samples of college students (see Downey for cues of their particular fear objects. For
& Feldman, 1996), adolescents (Downey, example, in the presence of rejection-themed
Lebolt, Rincon, & Freitas, 1998), and adults art, highly rejection-sensitive people show
(Downey, Berenson, & Kang, 2006). increased physiological reactivity associated
The RSQ does not tap into a general sen with a vigilant state, as measured using a
sitivity to negative events but, rather, mea startle probe paradigm (Downey, Mougios,
sures specifically fears and expectations of et al., 2004). However, rejection-sensitive
personal rejection by significant others. For people do not show this heightened response
example, the RSQ is distinct in terms of its to art that has other negative themes or to
predictive validity from measures of sen positive- or acceptance-themed art.
sitivity to race- or gender-based rejection Although they are vigilant to cues of re
(Chan & Mendoza-Denton, 2 0 0 8 ; London jection, rejection-sensitive people also show
et al., 2 0 0 6 ; Mendoza-Denton et al., 2002). a bias to ignore information about potential
Rejection-sensitivity scores are also not re rejection. When the option to avoid rejec
dundant with constructs with which RS tion— or situations that are likely to result
might be expected to overlap, such as in in rejection— exists, the highly rejection-
troversion, neuroticism, self-esteem, general sensitive person is likely to seize it. This
attachment style, depression, social anxiety, avoidance has emerged in work exploring
and social avoidance (Downey & Feldman, attentional biases in the perception of threat
1996). It is also not significantly associated ening faces using a visual probe task (Beren
with self-monitoring or with the perspective- son & Downey, 2008b ; Mogg, Mathews,
taking or empathic-concern dimensions of & Eysenck, 1992). Rejection sensitivity pre
Davis’s empathy measure (Romero-Canyas dicted disrupted attention away from threat
& Downey, 2008). Rejection sensitivity is, stimuli (angry faces) but was unrelated to
however, associated with the subscale of Da reactions to neutral stimuli. This resembles
vis’s empathy measure that assesses the ten the pattern that abused children with post-
dency to find other people’s distress upset traumatic stress disorder (PTSD) show (e.g.,
ting. Finally, the RSQ is weakly negatively Pine et al., 2005) and reflects motivated
associated with narcissism. strategies for regulating threat. When de
tecting and reducing threat are priorities, at
tentional vigilance for opportunities to avoid
The Perception o f and Im m ediate the threat sources should be the preferred re
R eaction to R ejection Cues sponses. This strategy may develop as a re
sult of the competition between the motiva
As noted, the network of cognitive-affective tion to prevent rejection and the motivation
units that are theorized to constitute a per to seek acceptance that characterize highly
sonality dynamic such as rejection sensitivity rejection-sensitive people. When seeking to
include encoding and perceptual biases that remain close to the source of the potential
moderate behavior. These biases are acti threat, turning a blind eye to indications of
vated in situations of social threat (Downey, threat keeps that threat at bay, at least at the
Irwin, Ramsay, & Ayduk, 2 0 0 4 ; Downey, perceptual and subjective levels. However,
Mougios, Ayduk, London, & Shoda, 2004) in day-to-day life, people face situations in
and influence the processes of detection, which indications that they are not liked or
interpretation, and reaction to signals of accepted cannot be avoided.
potential rejection. However, rejection sen Once confronted with cues of rejection
sitivity is unrelated to biases in people’s per and with no option to avoid them, highly
ception of signs of acceptance or positive rejection-sensitive people see those cues as
affect. In other words, rejection sensitivity more threatening and negative compared
functions as a defensive system singularly at with other people, showing a bias to see
470 V. M O T IV A T IO N A L D IS P O S IT IO N S
and aggression (Leary, Twenge, & Quinlivan, as well as indirect efforts to sabotage other
2006). For highly rejection-sensitive people, singers in the audition (DiBenigno, Romero-
this link is particularly strong, noticeable Canyas, & Downey, 2007).
even among children. Rejection-sensitive The rejection-hostility link is also evident
middle-school children react more strongly in highly rejection-sensitive people’s reac
than less sensitive children to a staged rejec tions to distant, powerful others. Rejection-
tion from a class peer. Furthermore, over the sensitivity scores predict people’s withdrawal
course of the school year, they have more of support from politicians who people feel
hostility-related behavioral problems in the betrayed them (Romero-Canyas & Downey,
classroom than less rejection-sensitive chil 2003) and expressed hostility toward and
dren (Downey, Lebolt, et al., 1998). distancing from God among religious people
The link between rejection and hostility who are facing personal difficulties (Ander
thoughts is strong and automatic among son, Romero-Canyas, & Downey, 2008).
people who are rejection sensitive. In se Clearly, behaving in a hostile way toward
quential priming-pronunciation paradigms, others is likely to elicit rejection. Dating cou
rejection-related words facilitate pronuncia ples that include a highly rejection-sensitive
tion of hostility-related words among people person are almost three times more likely
who are high in rejection sensitivity but not than couples without a highly rejection-
among people who are low in rejection sensi sensitive person to have separated within a
tivity (Ayduk, Downey, Testa, Yen, & Shoda, year of the time their rejection sensitivity was
1999; Romero-Canyas, Downey, Berenson, measured (Downey, Freitas, et al., 1998).
Ayduk, & Kang, in press), suggesting an au The processes underlying these outcomes
tomatic link between rejection and hostile have been explored in laboratory work. In
thoughts. These hostile thoughts translate to one study (Downey, Freitas, et al., 1998),
more hostile behavior for highly rejection- independent raters coded videotapes of
sensitive people. In one study (Ayduk et al., couples engaged in discussing a relationship
1999), women expected to meet with a man issue. Women’s rejection-sensitivity scores
after exchanging biographical information predicted greater partner-reported anger, as
with him. When the man then refused to well as more behavior coded as angry by in
meet the participant, rejection-sensitivity dependent raters. Women high in rejection
scores predicted a more negative evaluation sensitivity were more hostile, defensive, and
of him. negative than women low in rejection sensi
In diary studies of couples (Ayduk et al., tivity. This hostility had an impact on their
1999; Downey, Freitas, et al., 1998), rejec partners’ affect such that the coders’ ratings
tion sensitivity predicted a greater probabil of negative behavior accounted for 54% of
ity of conflict following a day when highly the effect of women’s rejection sensitivity on
rejection-sensitive people felt more rejected, their partner’s negative affect.
suggesting that highly rejection-sensitive In sum, the negative response that highly
women react to rejection with some hostil rejection-sensitive people show to cues of
ity toward their partners. Similarly, rejec rejection engenders correspondingly nega
tion sensitivity predicts greater relation tive responses from others. However, highly
ship violence by male college students who rejection-sensitive people are also motivated
were highly invested in their relationships to avoid rejection and to seek acceptance,
(Downey, Feldman, & Ayduk, 20 0 0 ). High even when faced with very clear cues indi
ly rejection-sensitive people are also more cating that rejection is probable.
likely to aggress against strangers who re
ject them, as captured in a study using the
hot-sauce paradigm. In this study, highly R ejection Sensitivity and Efforts
rejection-sensitive people fed their reject to Secure A cceptance
ers a condiment that they knew the reject
ers found particularly aversive: hot sauce Rejection-sensitive people are highly mo
(Ayduk, Gyurak, & Luerssen, 2007). Simi tivated to prevent rejection. In the course
larly, in a study of singers in auditions, re of social interactions, highly rejection-
jection sensitivity predicted aggression and sensitive people will engage in impression-
derogation of judges following rejection, management strategies to prevent rejection
472 V. M O T IV A T IO N A L D IS P O S IT IO N S
from loved ones or new interaction partners. ticipants reported their willingness to carry
When faced with strong cues that rejection out a series of tedious tasks for the group
is impending, as when they are turned away (e.g., cooking dinner for the group, archiving
from a social group or rejected by a prospec past messages exchanged by group members)
tive dating partner, highly rejection-sensitive and how much money they would donate to
people will make efforts to win back the per a group meeting. After rejection, but not
son who has rejected them. after acceptance, men’s rejection-sensitivity
Efforts to prevent rejection from signifi scores predicted larger monetary contribu
cant others are evident in early work on re tions to the group and a greater willingness
jection sensitivity. In a study of adolescent to perform the tedious tasks.
girls in romantic relationships (Purdie & The rejection-avoidance behaviors that
Downey, 2 0 0 0 ), rejection sensitivity corre highly rejection-sensitive people adopt may
lated positively with greater willingness on not always yield the desired result. People
the part of girls to “do anything” to keep who fear rejection and are asked to sacrifice
their boyfriends, even if that meant doing personal interests for their partners are more
something they thought was wrong. Among likely to end their relationships than people
low-income women at risk for contracting who had no such concerns (Impett, Gable,
HIV, higher rejection sensitivity predicted &C Peplau, 2005). Highly rejection-sensitive
the likelihood of engaging in risky sexual people’s apparently ingratiating behavior
behavior to prevent their sexual partners after rejection may elicit mistrust and sus
from leaving them (Berenson & Downey, picion from the social targets whom they
2008a). These self-silencing strategies also pursue, leading to eventual rejection. In ad
emerge in novel relationships in situations dition, a highly rejection-sensitive person’s
in which people are given information that willingness to transform him- or herself into
suggests that rejection is possible. In a study a different person to gain acceptance may
of self-presentation strategies, when in have long-term costs leading to unstable re
troducing themselves to a group of highly lationships and troubled affect.
artistic peers who were described as being
unathletic, rejection-sensitive participants
presented themselves as less athletic using The Link between R ejection
the same rating scale they had used less Sensitivity and Health Problem s
than an hour earlier to indicate high levels
of athleticism to the experimenter. Sim ilar The findings we have reviewed thus far sug
ly, rejection sensitivity predicted changes in gest that people who are rejection sensitive
self-ratings of political conservatism among react strongly to rejection and that their
college students joining a group of peers in overreactions often have negative conse
a university campus where politically con quences for them. In combination with other
servative students feel stigmatized and alien psychological dynamics, rejection sensitiv
ated (Romero-Canyas, Downey, Pelayo, & ity predicts poor mental health. Rejection-
Bashan, 20 0 4 ). Highly conservative, highly sensitive children tend to be socially avoidant
rejection-sensitive men who were randomly and lonely (London, Downey, Bonica, &
assigned to interact with a very liberal group Paltin, 20 0 7 ). The same pattern is evident
decreased their self-ratings of conservatism among college students, for whom rejection
at the time of their public presentation to the sensitivity also predicts a smaller number of
group. Highly conservative, highly rejection- close friends, a smaller number of significant
sensitive men placed in a conservative group dating relationships, and longer periods of
did not change their scores. time before entering relationships (Beren
People high in rejection sensitivity are son, Kang, & Downey, 2 0 0 8 ; Downey et
also more likely than people low in rejection al., 2000).
sensitivity to make efforts to regain accep Rejection sensitivity also predicts strong
tance from the people who have expressed internalization of rejection experiences, pre
no interest in them and have rejected them dicting depression following interpersonal
(Romero-Canyas et al., 200 8 ). In a series of losses associated with rejection (Ayduk,
studies, after being rejected or accepted by a Downey, & Kim, 2001). Rejection-sensitiv-
novel group of peers or dating partners, par ity scores obtained 2 weeks before the start
32. R e je ctio n Sensitivity 473
of the school year predicted more depressive that people with BPD avoided angry faces
symptoms at the end of the school year for and showed less attention to happy faces,
participants who had experienced a partner- whereas people with APD show a bias to
initiated breakup during the 6 months pre ward angry faces. Hence, patients with
ceding the end of the school year. Rejection BPD behaved like highly rejection-sensitive
sensitivity was unrelated to depressive symp people. In the paradigm used in the studies
toms among women who had initiated the of overestimation of negativity, people with
breakups or who had not experienced any BPD perceived significantly more negativity
breakups. For male college students who and slightly less positivity in faces and were
viewed themselves as conservative, rejec more certain about their interpretations
tion sensitivity predicted a higher number of than control participants. People with BPD
depressive symptoms, as well as a lowered or APD also showed a lower threshold for
sense of belonging at their liberal university detecting angry faces in angry-afraid mor
(Romero-Canyas, Downey, &C Cavanaugh, phed faces. Finally, when asked to imagine
2003). These correlations are not evident that close others might be losing interest in
among highly rejection-sensitive, liberal men them, people with BPD reported a higher
who did not expect peer rejection because of likelihood of losing control of their tempers,
their beliefs. gratifying impulsive urges including harm to
themselves, and a lower likelihood of talk
ing to close others to improve the relation
T h e Association o f Rejection Sensitivity
ships. In response to the same scenario, peo
and Personality Disorders
ple with APD reported imagining that they
Sensitivity to rejection, conceptualized as would withdraw and feel worthless.
a tendency to overreact to rejection, is one
of the diagnostic criteria for some forms of
E ating and Body D ysm orphic Disorders
psychopathology, such as depression and
borderline personality disorder (American Being sensitive to rejection based on physi
Psychiatric Association, 1994). There are cal appearance or attractiveness has specific
clear parallels in the perceptual processes, physical and mental health implications.
behaviors, and outcomes between rejection Atlas (2004) found that appearance-related
sensitivity and borderline personality disor sensitivity (measures of feelings of attrac
der (BPD) and avoidant personality disorder tiveness and investment in one’s appear
(APD). BPD is characterized by impulsivity, ance) together with rejection sensitivity pre
instability in mood, and self-directed harm dicted a drive for thinness and symptoms
or injury. APD is characterized by social in of bulimia. Park (2007a) and her colleagues
hibition, feelings of inadequacy, and avoid developed an appearance-based rejection-
ance of social interaction. At the core of both sensitivity measure to capture the dynamic
disorders is a fear of rejection or abandon of anxiously expecting rejection because of
ment that leads to volatile relationships and one’s appearance or physical attractiveness.
low self-concept clarity. People with BPD or High appearance-based rejection sensitivity
APD have significantly higher levels of rejec is strongly associated with body dysmorphic
tion sensitivity than healthy controls, and the disorder and predicts intent to undergo cos
relationship between BPD and high rejection metic surgery for social rather than personal
sensitivity holds even when controlling for reasons (Park, DiRaddo, &c H arwin, 2007).
depression (Berenson, 2008). This suggests People high in appearance-based rejection
that rejection sensitivity might play a role in sensitivity experience distress when inter
the processing of social information among acting with others in situations in which
people who are diagnosed as having BPD or they believe appearance is important (Park,
APD. 2007b).
Consistent with this hypothesis, Berenson As the work on appearance-based rejection
(2008) found that people with APD and es sensitivity illustrates, people can be rejection
pecially those with BPD selectively attend sensitive with regard to particular aspects
ed to, detected, interpreted, and reacted to of themselves. Specific concerns, situations,
social threat cues in ways that reflect high or aspects of the person can be the basis
rejection sensitivity. Berenson also showed for concerns about rejection. For example,
474 V. M O T IV A T IO N A L D IS P O S IT IO N S
torically black university, where the threat nantly white environments, such as students
of race-based rejection was highly unlikely. in predominantly white high schools. In An
Anderson and Downey (2005) replicated derson and colleagues’ study (2008), African
this finding in a study of race-based rejec American and Hispanic students’ scores on
tion sensitivity among high school students. the RSQ-race predicted a preference for col
Significantly, positive experiences with peo leges with more students of color, whether
ple from the threatening group attenuate the these were fictitious colleges or real colleg
effects of race-based rejection sensitivity on es to which they desired to apply. In focus
African Americans. Hence, having quality group discussions, students offered such ex
friendships with white students increased planations as, “I just can’t repeat this experi
sense of belonging among highly race- ence. There needs to be enough Black people
based rejection-sensitive African American in my college for me to be com fortable” (An
students at a predominantly white college derson, 2005). O f course, institutional racial
(Mendoza-Denton et al., 20 0 6 ). composition is not the only factor considered
The discomfort experienced by students in the college choice of minority students.
high in race-based rejection sensitivity in Although factors such as proximity to home
predominantly white environments may and availability of financial aid all influence
motivate them to avoid or at least minimize the choice of where students decide to attend
further discomfort by avoiding people who college, the racial makeup of a given college
represent the rejecting authorities. Among may be very important in the college transi
African American students at predominantly tion of students high in RS-race, as evidenced
white universities and among African Amer by the adverse outcomes for students high in
ican and Hispanic students at predominant RS-race seen in Mendoza-Denton and col
ly white and Asian high schools, race-based leagues (2002).
rejection sensitivity predicted less academic African Americans and Hispanics in ma
help seeking and use of available resources, jority white settings may be more at risk for
such as attending professors’ and teaching compromised health if they score high in
assistants’ office hours and review sessions race-based rejection sensitivity due to the
(Anderson et al., 2 0 0 8 ; Mendoza-Denton et higher probability of perceiving race-based
al., 2002). In an experimental study, Lon rejection. Pietrzak (2004) found that stu
don, Downey, and Dweck (2008) found that dents high in race-based rejection sensitiv
after receiving feedback on essays, African ity reported more somatic symptoms, such
American participants high in race-based as stomachache and pounding heart, after
rejection sensitivity who believed that the reading a vignette about a negative racial
evaluating professor was aware of their race incident or after recalling a recent person
were less willing to meet the professor to dis al negative racial experience. Furthermore,
cuss improving their essays relative to peers these students are also more likely to respond
low in RS-race. In challenging academic to such experiences with self-silencing or
settings in which seeking help is crucial in strong emotional outbursts (e.g., crying, yell
managing difficult coursework, avoiding ing), reactions that are likely to further their
uncomfortable but potentially beneficial sit isolation (Pietrzak, 2 0 0 4 ; Velilla, Mendoza-
uations may translate into academic under Denton, London, & Downey, 2001).
achievement. Indeed, Mendoza-Denton and
colleagues (2002) found that students high
R ace-B ased Rejection Sensitivity
in race-based rejection sensitivity exhibited
among A sia n A m ericans
a decline in grade point average over their 4
years at a competitive predominantly white Because Asian American stereotypes and dis
university. crimination experiences differ from those of
The desire to avoid the discomfort of po other ethnic minorities in the United States
tential rejection makes the prospect of at (Oyserman & Sakamoto, 1997), a scale spe
tending a predominantly white college unat cific to the stigmatizing and discriminating
tractive to high school students who are high experiences of Asian Americans was devel
in race-based rejection sensitivity. This is oped (rejection sensitivity— Asian; Chan &
especially true for students high in RS-race Mendoza-Denton, 2008). Asian Americans
who have extensive exposure to predomi as a group have lower self-esteem than Af
476 V. M O T IV A T IO N A L D IS P O S IT IO N S
Y. (1999). Does rejection elicit hostility in rejection p e r fo r m in g artists? Poster presented at the annual
sensitive women? S o c ia l C o g n itio n , 17, 2 4 5 - 2 7 1 . meeting of the Society for Personality and Social
Ayduk, O., Gyurak, A., & Luerssen, A. ( 2 0 07 ). Indi Psychology, Memphis, T N .
vidual differences in the rejection-aggression link Downey, G., Berenson, K. R., 8c Kang, J. (2006).
in the hot sauce paradigm: The case of rejection sen T h e A du lt R e je c tio n S en sitiv ity Q u es tio n n a ir e (A -
sitivity. J o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch olog y , R S Q ). Unpublished questionnaire, Columbia Uni
44, 775-782. versity.
Ayduk, O., Mendoza-Denton, R., Mischel, W., Downey, G., Feldman, S., 8c Ayduk, O. (200 0 ). Rejec
Downey, G., Peake, P. L. K., & Rodriguez, M. tion sensitivity and male violence in romantic rela
(20 0 0). Regulating the interpersonal self: Strategic tionships. P ers o n a l R e la tio n s h ip s, 7, 4 5 - 6 1 .
self-regulation for coping with rejection sensitivity. Downey, G., 8c Feldman, S. I. (1996). Implications
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o lo g y , 79, of rejection sensitivity for intimate relationships.
776-792. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 70,
Ayduk, O. N., Zayas, V., Downey, G., Cole, A. B., 1327-1343.
Shoda, Y., 8c Mischel, W. (200 8 ). Rejection sen Downey, G., Freitas, A. L., Michaelis, B., 8C Khouri,
sitivity and executive function: Jo int predictors of H. (1998). The self-fulfilling prophecy in close rela
borderline personality features. J o u r n a l o f R e se a rch tionships: Rejection sensitivity and rejection by ro
in P erson ality , 4 2 , 1 5 1 - 1 6 8 . mantic partners. J o u r n a l o f P erson ality a n d S o cia l
Barash, D. P. (1 977). S o c io b io lo g y a n d b eh a v io r . O x P sy ch olog y , 75, 5 4 5 - 5 6 0 .
ford, UK: Elsevier North-Holland. Downey, G., Irwin, L., Ramsay, M ., 8c Ayduk, 0 .
Baumeister, R. F., DeWall, C. N., Ciarocco, N. J ., 8C (200 4). Rejection sensitivity and girls’ aggression.
Twenge, J. M . (2005). Social exclusion impairs self In M. M. Moretti, C. L. Odgers, &c M. A. Jackson
regulation. J o u r n a l o f P erson ality a n d S o c ia l P sy (Eds.), G irls a n d a g g r es sio n : C o n trib u tin g fa c t o r s
c h o lo g y , 8 8 , 5 8 9 - 6 0 4 . a n d in terv en tio n p r in c ip le s (pp. 7 - 2 5 ) . New York:
Baumeister, R.F., 8c Leary, M. R. (1995). The need to Kluwer Academic/Plenum Press.
belong: Desire for interpersonal attachments as a Downey, G., Lebolt, A., Rincon, C., & Freitas, A.
fundamental human motivation. P s y c h o lo g ic a l B u l L. (1998). Rejection sensitivity and children’s in
letin , 117, 4 9 7 - 5 2 9 . terpersonal difficulties. C h ild D e v e lo p m e n t, 69,
Baumeister, R. F., Twenge, J . M ., &c Nuss, C. K. (2002). 1074-1091.
Effects of social exclusion on cognitive processes: Downey, G., Mougios, V., Ayduk, O., London, B. E.,
Anticipated aloneness reduces intelligent thought. 8c Shoda, Y. (200 4 ). Rejection sensitivity and the
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 83, defensive motivational system: Insights from the
8 1 7 -8 27 . startle response to rejection cues. P sy c h o lo g ic a l S ci
Berenson, K. R. (200 8 ). R e je c tio n sen sitiv ity a s an a n en c e , 15, 6 6 8 - 6 7 3 .
t e c e d e n t o f b o r d e r lin e p e r s o n a lity d iso rd er. M anu Feldman, S., 8c Downey, G. (1994). Rejection sensi
script in preparation. tivity as a mediator of the impact of childhood
Berenson, K. R ., 8c Downey, G. (2 008a). R e jec tio n exposure to family violence on adult attachment
sen sitiv ity a n d in c re a se d risky s e x u a l b e h a v io r f o r behavior. D e v e lo p m e n t a n d P s y c h o p a th o lo g y , 6,
w o m e n a t risk f o r con tra c tin g sex u a lly tra n sm itted 231-247.
illn esses. Unpublished manuscript, Columbia Uni Horney, K. (1937). T h e n eu r o tic p e r s o n a lity o f o u r
versity. tim e. New York: Norton.
Berenson, K. R., & Downey, G. (2008b). R e je c Impett, E. A., Gable, S. L., 8c Peplau, L. A. (2 005).
tio n sen sitiv ity p re d ic ts a tte n tio n b ia s a w a y fro m Giving up and giving in: The costs and benefits of
t h re a te n in g fa c e s . Unpublished manuscript, Colum daily sacrifice in intimate relationships. J o u r n a l o f
bia University. P erson ality a n d S o c ia l P sy ch olog y , 8 9, 3 2 7 - 3 4 4 .
Berenson, K. R ., Kang, J ., 8c Downey, G. (2008). Kross, E., Egner, T., Ochsner, K., Hirsch, J ., 8c
R e jec tio n sen sitiv ity a n d in tim a te r e la tio n sh ip in Downey, G. ( 2 0 0 7). Neural dynamics of rejection
th e tran sition to y ou n g a d u lt h o o d . Manuscript in sensitivity. J o u r n a l o f C o g n itiv e N e u r o sc ie n c e, 19,
preparation. 945-956.
Bourgeois, K. S., 8c Leary, M . R. (2001). Coping with Leary, M. R., Kowalski, R. M ., Smith, L., 8c Phillips,
rejection: Derogating those who choose us last. M o S. (2003). Teasing, rejection, and violence: Case
tiv ation a n d E m o tio n , 2 5 , 101—111. studies of the school shootings. A g g ressiv e B e h a v
Burklund, L. J . , Eisenberger, N. I., 8c Lieberman, M. ior, 2 9, 2 0 2 - 2 1 4 .
D. (20 07 ). The face of rejection: Rejection sensitiv Leary, M . R., Twenge, J. M., 8c Quinlivan, E. (2006).
ity moderates dorsal anterior cingulated activity to Interpersonal rejection as a determinant of anger
disapproving facial expressions. S o c ia l N e u r o s c i and aggression. P erson ality a n d S o c ia l P sy ch o lo g y
en c e , 2 , 2 3 8 - 2 5 3 . R ev iew , 10, 1 1 1 - 1 3 2 .
Chan, W., 8c Mendoza-Denton, R. (200 8 ). Statu s- LeDoux, J. E. (1996). T h e e m o tio n a l b ra in : T h e m y s
b a s e d re jec tio n sen sitiv ity a m o n g A sian A m er ic a n s: ter io u s u n d erp in n in g s o f e m o tio n a l life. New York:
Im p lic a tio n s f o r p s y c h o lo g ic a l distress. Manuscript Simon 8c Schuster.
submitted for publication. London, B., Downey, G., Bonica, C., 8c Paltin, I.
Davis, M. (1992). The role of the amygdala in fear (2 0 07 ). Social causes and consequences of rejection
and anxiety. A n n u al R e v ie w o f N e u r o s c ie n c e , 15, sensitivity. J o u r n a l o f R e se a rc h o n A d o le s c e n c e , 17,
353-375. 481-506.
DiBenigno, J ., Romero-Canyas, R., 8c Downey, G. London, B., Downey, G., 8c Dweck, C. (2008). T h e
(2 00 7, January). D o re jectio n sen sitiv ity a n d p e r s t u d e n t ’s d ile m m a : C o p in g resp o n s es to s te r e o ty p e
fo r m a n c e p r e d ic t e x t r e m e r e a c tio n s to re jec tio n fo r th rea t. Manuscript in preparation.
32. R eje ctio n S ensitivity 479
London, B., Downey, G., Rattan, A., & Tyson, D. A p p e a r a n c e -b a s e d re jectio n sen sitiv ity a n d clin i
(2 0 06 ). G e n d e r re jec tio n sen sitiv ity : T h eo r y , v a li c a l o u t c o m e s : E ffe c ts o n b o d y d y sm o rp h ic d is o r
d a tio n , a n d im p lic a tio n s f o r th e p s y c h o s o c ia l w e ll d er a n d d es ire fo r c o s m e t ic su rgery. Manuscript in
b ein g a n d a c h ie v e m e n t o u t c o m e s o f w o m e n . Un preparation.
published manuscript. Pietrzak, J. (2 004). Race-based rejection sensitivity
Maslow, A. (1987). M otiv atio n a n d p e r s o n a lity (3rd and ethnic identity: Interactive effects on institu
ed.). New York: Harper & Row. tional affiliation and well-being. D is s er ta tio n A b
McClelland, D. C. (1987). H u m an m o tiv a tio n . C am stracts In te rn a tio n a l, 6 4 , 6 3 7 9 - 6 4 5 3 .
bridge, UK: Cambridge University Press. Pine, D. S., Mogg, K., Bradley, B. P., Montgomery, L.,
Mendoza-Denton, R., Downey, G., Purdie, V., Davis, M onk, C. S., McClure, E., et al. (2005). Attention
A., & Pietrzak, J . (2002). Sensitivity to status-based bias to threat in maltreated children: Implications
rejection: Implications for African American stu for vulnerability to stress-related psychopathology.
dents’ college experience. J o u r n a l o f P erson ality A m erica n J o u r n a l o f P sychiatry, 162, 2 9 1 - 2 9 6 .
a n d S o c ia l P sy ch olog y , 8 3, 8 9 6 - 9 1 8 . Purdie, V., Sc Downey, G. (20 00 ). Rejection sensitivity
Mendoza-Denton, R., Page-Gould, E., & Pietrzak, J. and adolescent girls’ vulnerability to relationship-
(2 0 06 ). Mechanisms for coping with status-based centered difficulties. C h ild M a ltrea tm en t, 5, 3 3 8 -
rejection expectations. In S. Levin & C. van Laar 349.
(Eds.), S tig m a a n d g ro u p in eq u a lity : S o c ia l p s y c h o Romero-Canyas, R., Sc Downey, G. (2 0 0 3 , February).
lo g ic a l p er sp e c tiv e s (pp. 151-16 9). Mahwah, NJ: S en sitivity to re jectio n b y g ro u p s (rejectio n s e n si
Erlbaum. tiv ity -G ) a s a p r e d ic to r o f p o lit ic a l b eh a v io r . Poster
Mendoza-Denton, R., Pietrzak, J . , & Downey, G. presented at the conference of the Society of Person
(2 0 08 ). Distinguishing institutional identification ality and Social Psychology, Los Angeles.
from academic goal pursuit: Interactive effects of Romero-Canyas, R., Sc Downey, G. (20 08 ). R e je c
ethnic identification and race-based rejection sen tion sen sitiv ity p r e d ic ts o v ere stim a tin g th e n eg ativ e
sitivity. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l m o o d o f o th er s a n d p re d ic ts co n ta g io n o f n eg ativ e
og y , 95, 1 0 8 0 - 1 0 9 4 . a ffe c t . Manuscript in preparation.
Metcalfe, J ., Sc Mischel, W. (1999). A hot/cool system Romero-Canyas, R ., Downey, G., Berenson, K.,
analysis of delay of gratification: Dynamics of will Ayduk, O., Sc Kang, J . (in press). Rejection sensitiv
power. P sy c h o lo g ic a l R e v iew , 1 0 6 , 3 - 1 9 . ity and the rejection-hostility link in romantic rela
Mischel, W., Sc Shoda, Y. (1995). A cognitive-affective tionships. J o u r n a l o f P erson ality.
system theory of personality: Reconceptualizing sit Romero-Canyas, R., Downey, G., & Cavanaugh, T. J.
uations, dispositions, dynamics, and invariance in (2003). F eelin gs o f a lien a tio n a m o n g c o lle g e m en :
personality structure. P s y c h o lo g ic a l R e v iew , 102, T h e im p a c t o f rejectio n sen sitiv ity a n d p o litic a l
246-268. b e lie fs . Unpublished manuscript, Columbia Univer
Mogg, K., Mathews, A., & Eysenck, M. (1992). At sity.
tentional bias to threat in clinical anxiety states. Romero-Canyas, R., Downey, G., Pelayo, R., &
C o g n itio n a n d E m o tio n , 6, 1 4 9 -1 59 . Bashan, U. ( 2 0 0 4 , February). T h e th re a t o f rejectio n
Olsson, A., Carmona, S., Downey, G., &C Ochsner, K. triggers s o c ia l a c c o m m o d a t io n in re jectio n sen sitiv e
N. (20 08 ). P erceiv in g th rea t a n d lea rn in g to fe a r : A m en . Poster presented at the annual meeting of the
p ro c e s s in g a c c o u n t o f re jec tio n sen sitivity. M a n u Society for Personality and Social Psychology, Aus
script in preparation. tin, Texas.
Oyserman, D., 8c Sakamoto, I. (1997). Being Asian Romero-Canyas, R., Downey, G., Reedy, K. S., Rodri
American: Identity, cultural constructs, and stereo guez, S., Cavanaugh, T., & Pelayo, R. (2 008). P ay
type perception. J o u r n a l o f A p p lied B e h a v io r a l S ci in g to b elo n g : W ho tries. T h e lin k b e tw e e n rejectio n
en c e , 3 3 , 4 3 5 - 4 5 3 . sen sitiv ity a n d in g ratiation a f t e r re jectio n . M a n u
Pachankis, J. E., Goldfried, M. R., Sc Ramrattan, M. script submitted for publication.
E. (20 08 ). Extension o f the rejection sensitivity co n Twenge, J. M ., Baumeister, R. F., Tice, D. M., &
struct to the interpersonal functioning of gay men. Stucke, T. S. (2001). If you can’t join them, beat
J o u r n a l o f C o n su ltin g a n d C lin ica l P sy ch olog y , 76, them: Effects of social exclusion on aggressive be
306-317. havior. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l
Park, L. (2007a). Appearance-based rejection sensi og y , 81, 1 0 5 8 - 1 0 6 9 .
tivity: Implications for mental and physical health, Velilla, E., Mendoza-Denton, R., London, B., &
affect, and motivation. P erson ality a n d S o c ia l P sy Downey, G. (20 0 1 , July). R a c e b a s e d re jec tio n s e n
c h o lo g y B u lletin , 3 3 , 4 9 0 - 5 0 4 . sitiv ity in a p r e - c o lle g e s a m p le o f stu d en ts. Paper
Park, L. (2007b). In te r p e r s o n a l d y n a m ics o f a p p e a r presented at the Leadership Alliance Research Sym
a n c e - b a s e d re jec tio n sen sitivity. Manuscript in posium, Atlanta, GA.
preparation. Williams, K. D. (2001). O stra cism : T h e p o w e r o f s i
Park, L. E., DiRaddo, A., & Har win, M. J. (2 007). len ce . New York: Guilford Press.
CHAPTER 33
■ • A 4 « « A *
Psychological Defensiveness
Repression, Blunting, and Defensive Pessimism
J u l ie K . N o r e m
sychologists from many theoretical per and performance consequences for those
P spectives across clinical, personality,
social, cognitive, and physiological psychol
who use them and how the relative costs and
benefits of a strategy vary as a function of
ogy have been interested in understanding specific tasks and contexts.
defensive processes. As a result, numerous
constructs and operationalizations describe
perceptual and information-processing Individual Differences in
styles and self- and affect-regulation strate Physiological Inhibition and Arousal
gies that involve defensive components. Fur
thermore, individual differences in defensive Theories concerning individual differences
processes overlap with work on coping, nar in the relative strength of biologically based
cissism, self-esteem, self-deception, sensa aspects of the nervous system have a long
tion seeking, introversion and extraversion, history in philosophy, medicine, and psy
the behavioral activation-inhibition system, chology. Many of these conceptions focus on
neuroticism, anxiety, rejection sensitivity, physiological phenomena related to inhibi
and social desirability, as well as numerous tion and arousal, responses to environmen
clinical constructs and categories. tal stimuli, and reactions to reinforcement
Comprehensive review of the ways in which contingencies (Traue & Pennebaker, 1993).
defensiveness potentially plays a role in of Pavlov (1927), for example, conceived of
all of these phenomena would be beyond the “strong” and “weak” nervous systems, and
scope of an entire volume, much less a single subsequent researchers elaborated his defini
chapter, so this chapter focuses on constructs tions to characterize individual differences
that concern how people typically process in the strength of response to reinforcement
information in situations in which there is a contingencies (Hull, 1950; Spence, 1936).
potential anxiety about specific outcomes. I Biologically based models typically assert
briefly review three individual-difference di that the goal of the nervous system is to
mensions (augmenting-reducing, repression- regulate arousal in reaction to stimuli and
sensitization, and monitoring-blunting) re that efforts to moderate arousal involve in
lated to processing sensory, affective, and hibiting motoric action, including emotional
cognitive stimuli. Then I review research expression. Arousal differences stem from
on ways in which the strategies of defensive organic nervous system differences that cre
pessimism and strategic optimism influence ate differences in perception and physiologi
cognitive, affective, motivational, social, cal experience.
480
33. P sy ch o lo g ical D efensiveness 481
Study of inhibition and arousal, as they ciation with current usage of repression and
relate to repression, emotional expressive defensiveness , past uses do not correspond
ness, and perception, led to several attempts to present understandings in a straightfor
to develop individual-difference models that ward way.
would capture physiological and psycho
logical patterns of responding. Theorists
developed views of individual responses to A ugm enters and Reducers
reward and punishment that are related to
the biologically based construct of inhibi Other models of reaction to stimuli also sug
tion, including theories of introversion- gest connections to defensiveness. Petrie pos
extraversion and the behavioral inhibition ited two perceptual-cognitive styles called
system and behavioral approach system (Ey augmenting and reducing (A -R). At one
senck, 1967; Gray, 1972). extreme she described “augmenters,” who
Buck (1976) argued that the biologically magnify or intensify effects of sensory input,
based disposition of introversion is associ and at the other extreme “reducers,” who at
ated with higher electrodermal and other tempt to decrease or attenuate the effect of
physiological reactivity, which, under so sensory input (Petrie, 1967). These styles re
cialization, leads to the inhibition of ex flect temperamental differences in the mod
pression. Early work found disjunctions ulation of stimulation. She operationalized
between physiological responses (e.g., skin these styles with the kinesthetic aftereffect
conductance and heart rate) on the one hand measure, in which individuals typically re
and verbal reports and emotional expres port that a standard block is larger after they
siveness/responsiveness on the other hand have handled a smaller block and smaller
(Buck, 1976). Research on this disjunction after they have handled a larger block. She
is important to current understanding of found individual differences in the size of the
defensiveness as exhibited in defensive self kinesthetic aftereffect: Those who are more
esteem, self-deception, and narcissism. Pen extreme in their reports of how large the test
nebaker and his colleagues, for example, block is after the smaller block (in the aug
have done extensive work showing that in menting condition) also tend to be less ex
tentional thought suppression has a number treme in their reports of how small the test
of generally negative physiological conse block is after the larger block (in the reduc
quences (Pennebaker & Chew, 1985; Petrie, ing condition); they are labeled “augment
Booth, &c Pennebaker, 1998). ers.” Others augment less and reduce more
Nevertheless, this work takes us in a differ and are called “reducers.” Petrie (1967) also
ent direction than defensiveness per se in that identified two other groups: moderates, who
its focus is on intentional lying or suppres showed little aftereffect in either condition,
sion of troublesome emotions and thoughts. and the “stimulus-governed,” who exagger
Suppression that is initially intentional, of ated in both directions. Most research on
course, may develop into a characteristic or augmenting-reducing has focused on the
habitual suppression that occurs automati former two groups.
cally outside of awareness, which is one way Congruent with a focus on a sensory mea
of understanding repression-sensitization, sure, much of the research on these styles
as discussed later. Yet the subdued emotion has investigated reports of experience with
al responses measured by self- and observer other sensory stimuli; in particular, there
reports of introversion do not correspond in has been considerable research on the expe
any simple way to the verbal-physiological rience of pain as a function of A -R . In gen
disjunction noted either when people are eral, augmenters tend to be less tolerant of
trying to lie or when they automatically sup pain than reducers. Interestingly, they also
press. Indeed, suppression, or lying to oneself report greater pain relief from analgesics
that occurs outside of awareness, appears (Petrie, 1967) and from hypnosis (Morgan,
to be more strongly associated with extra Lezard, Prytulak, & Hilgard, 1970) than
verted characteristics than introverted ones reducers do. Petrie’s kinesthetic aftereffect
(Paulhus & John, 1998). Although use of the method correlates with questionnaire mea
term inhibition in this early work may sug sures of need for sensory stimulation, inter
gest to contemporary readers a close asso est in contact sports, delinquency, and toler
482 V. M O T IV A T IO N A L D IS P O S IT IO N S
ance of cold and pain (Herzog, W illiams, & fense mechanisms push threatening thoughts
Weintraub, 1985). and emotions out of conscious awareness
Sales (1971) found that reducers seek out and transform them so that they are no lon
and enjoy interesting and intense stimulus ger recognizable (Freud, 1946). Ample re
situations, whereas augmenters seek out and search on specific psychoanalytic defensive
enjoy quieter and duller stimulus situations; mechanisms demonstrates that people use
he argued that these styles represent differ distinct cognitive transformations when they
ent levels of “need for stimulation.” Further face psychological threat (Baumeister, Dale,
results connect A -R to levels of cortical at & Sommer, 1998; Cramer, 1995). There are
tenuation of incoming stimuli: Research has also defense mechanism inventories derived
shown systematic differences between aug from psychodynamic assumptions about
menters and reducers in evoked potentials, types of defense mechanisms (for a review,
reaction times, and physiological reactivity see Davidson & MacGregor, 1998).
(Schwerdtfeger & Baltissen, 2002). These There are commonalities between theo
differences are not specific to emotional ries that argue that biologically based
stimuli (Schwerdtfeger, 2003), which would individual differences in inhibitory and
tend to differentiate A -R from psychody arousal-related neural systems underlie in
namic accounts of defensiveness. dividual differences in physiological and
Individual differences in physiological psychological reactions to environmental
sensitivity could include or predispose one to stimuli, and psychodynamic propositions
emotional sensitivity that increases subjec that individuals are motivated to protect
tive need for defensive processing. Psycho themselves from— that is, keep out of con
analytic models that focus on defensiveness scious awareness— threatening or anxiety-
in response to sexual or aggressive content producing information and thoughts. Both
would have difficulty incorporating such predict individual differences in levels and
predispositions. Neuroticism as a broad di awareness of anxiety, potential disjunctions
mension of personality includes aspects of between physiological reactivity and verbal
heightened sensitivity and defensiveness, but behavior, and differences in characteris
not in ways that account for disjunctions tic modes of responding to threatening or
between physiological and conscious self- anxiety-producing stimuli.
reported experience. In contrast, models of
defensiveness that focus on threats to self-
concept might more easily include this kind The “ New L o o k ”
of sensitivity. Overall, although descriptions and R epression-Sensitization
of augmenters and reducers seem on the sur
face as if they would correspond to certain During the post-World War II years, several
conceptions of psychological defensiveness, psychologists developed new theories of per
empirical work suggests that they are bet ception, cognition, and personality that inte
ter understood as part of the constellation grated psychodynamic ideas about defensive
of characteristics that make up sensation processing with a functionalist approach to
seeking (see Zuckerman, Chapter 31, this perception (Bruner & Postman, 1947). The
volume) and extraversion (Bruneau, Roux, central idea of this “New Look” approach
Perse, & Lelord, 1984; Eysenck, 1973; see was that motivation influences our percep
W ilt & Revelle, Chapter 3, this volume). tion and processing of stimuli.
From this perspective grew a construct
related to individual differences in defen
Psychoanalytic and siveness: repression-sensitization (Eriksen,
Social-C ognitive Perspectives 1966), operationalized by the Repression-
Sensitization (R -S) Scale (Byrne, 1961).
Psychoanalytic theory is arguably the most Repression-Sensitization was initially
influential source of ideas about defensive characterized as a bipolar dimension, the
ness. Freud (1914) famously argued that extremes of which represented characteris
human beings are motivated to defend them tic and rigid defenses against threatening,
selves against the intrusion of psychological aversive, or anxiety-provoking stimuli (Bo
ly painful ideas and affects. An arsenal of de nanno & Singer, 1995). “Repressors” avoid,
33. P sy ch o lo g ical D efensiveness 483
deny, or minimize aversive information and 1995; Lorig, Singer, Bonanno, & Davis,
anxiety. “Sensitizers,” in contrast, seek out 1994; Mitchell, 1998; Rohrmann, Netter,
as much information as they can about Hennig, & Hodapp, 2003).
threatening situations and expend consider Repressors tend to encode emotions less
able effort worrying and ruminating about complexly (Hansen & Hansen, 1988) and
them. show greater memory failures for both nega
Byrne’s (1961) early scale correlated too tive and positive emotional events (Davis &
highly with anxiety measures, produced con Schwartz, 1987). This pattern led Tesser and
flicting results, and has largely been replaced his colleagues (Mendolia, M oore, & Tesser,
with new operationalizations. Weinberger, 1996) to reason that repressors are hyper
Schwartz, and Davidson (1979) developed sensitive to all emotional events but mo
the most popular measure of R -S , arguing tivated to repress only when they appraise
that the repressor end of the R -S scale con an emotional event as threatening to their
founded repressive individuals with those self-evaluations; their evidence supports
who were genuinely low in anxiety, whereas this more specific hypothesis. Subsequent
the sensitizer end confounded true sensitiz research findings converge with these. For
ers with those who were unsuccessful in example, repressors show electroencepha-
their defenses (as opposed to nondefensive) lographic (EEG) activity associated with
and thus unable to repress their anxiety. anxiety and an absence of cognitive activ
Confounding of true person variance with ity when faced with the recall of personally
defensive self-reports is an ongoing issue in threatening information (Lorig et al., 1994).
research on defensiveness. In addition, sensitizers show stronger emo
Weinberger and colleagues (1979) attempt tional reactions to potentially threatening
ed to disentangle these confounds by simul ambiguous stimuli and better memory for
taneously measuring trait anxiety with the such stimuli than repressors (Hock & Kroh-
M anifest Anxiety Scale (Taylor, 1953) and ne, 2004). Furthermore, repressors avoid
tendencies to respond in socially desirable disturbing self-relevant information when
ways with the M arlow e-Crow ne Social D e possible, and when they are unable to do so,
sirability Scale (Crowne & Marlowe, 1960) they rationalize and refute that information
to create four groups of individuals. Those (Baumeister & Cairns, 1992).
low in self-reported trait anxiety and low in
social desirability tendencies are nonanxious
and nondefensive. Those low in self-reported M onitoring and Blunting
trait anxiety but high in social desirability
are repressors. Those high in self-reported Congruent with the rise of social-cognitive
trait anxiety but low in social desirability approaches to personality during the 1980s,
are sensitizers, whereas those high in both Miller (1987) used information-processing
anxiety and social desirability are defensive terms to describe coping strategies that
and highly anxious individuals. Interest resemble R -S and A -R . Monitoring and
ingly, until recently, most research using this Blunting (M -B ) refer to sets of strategies,
method has concentrated only on the first hypothesized to be at least partially indepen
three of these groups. dent, for coping with the arousal or anxi
Research on R -S (often described as re ety generated by threatening events (Miller,
search on repressive coping) converges some 1987). Miller developed a self-report mea
what with earlier research on inhibition- sure called the M iller Behavioral Style Scale
expression and augmenting-reducing, in (MBSS) that includes four hypothetical and
that repressors typically show disjunction uncontrollable stressful scenes, followed by
between self-reported and autonomic indi eight statements describing characteristi
cators of stress (Weinstein, Averill, Opton, cally monitoring or blunting reactions. High
& Lazarus, 1968). Repressors’ self-reports monitors seek out and attend to information
suggest less stress than their autonomic re about their performance more than high
sponses indicate, whereas sensitizers’ self- blunters, whereas high blunters try to dis
reports suggest more stress than is reflected tract themselves from, deny, or reinterpret
in their autonomic responses (Bonanno, information about stressors. For example,
Davis, & Singer, 1991; Bonanno & Singer, when facing an unpleasant medical proce
484 V. M O T IV A T IO N A L D IS P O S IT IO N S
dure, monitors were interested in more in the M BSS Blunting scale loaded on a single
formation and felt less anxious after receiv factor, and all correlated significantly with
ing it (Miller & M anagan, 1983). measures of both anxiety and social desir
As with the original measure of repression- ability (Turvey & Salovey, 1993). Turvey
sensitization, one critique of the M BSS is and Salovey (1993) concluded that, based on
that it may confound nonanxious respon psychometric characteristics (high internal
dents with repressing or blunting respon consistency and normal distribution of re
dents, and, indeed, the Blunting subscale of sponses) and relative ease of administration
the M BSS typically has lower reliability and (only 22 items), the WAI— Defensiveness
less predictive validity than the M onitor scale was the most practically useful. As
ing subscale. The M BSS work has also been they noted, however, although their analysis
criticized for lack of predictive specificity supports the inference that these measures
in that some studies show that blunting re converge on one construct, it does not ad
lates to the outcomes measured, some stud dress the fundamental question of how this
ies show that monitoring predicts outcomes, construct is best understood.
and some studies use the difference score All self-report measures of individual
from the two subscales to predict outcomes. differences in defensiveness, when used
Finally, because the M BSS presents only by themselves, leave open the question of
objectively uncontrollable situations, critics whether systematic variations in response
have argued that it may not adequately as assess repression or denial of threatening
sess systematic variation in response to re negative content or the relative absence of
alistic situations in which control is possible that content. Interpretations of the defensive
(Krohne, 1996). nature of self-reports rest on assumptions
about human experience that are not shared
by all theoretical perspectives and that may
W hich Measure? not apply equally to all individuals. Valid
measure of defensive processes may require
Despite different theoretical origins, mea using two different methodologies (e.g.,
sures of R -S and M -B both relate to de self-report and physiological measures or
fensive tendencies and to some similar out self-report and observer reports) because
comes. Each measure also has weaknesses, neither alone can demonstrate the specifi
and there are questions about construct va cally defensive nature of processes that can
lidity, discriminant validity, and general util appear identical to nondefensive processes
ity. In response to some of these questions, when only one measure is used (Davidson &
Weinberger and Schwartz (1990) developed M acGregor, 1998).
a measure of repression tendencies as part
of the Weinberger Adjustment Inventory
(WAI) that includes three factors measured Defensive Pessim ism
by 10 subscales, two of which are designed and Strategic O ptim ism
to measure repressive tendencies: a Denial of
Distress subscale that measures respondents’ Research on the cognitive strategies of defen
claims not to be experiencing negative affect sive pessimism and strategic optimism devel
and a Repressive Defensiveness subscale that oped in the context of the self-enhancement
measures suppression of self-serving behav and self-protection research that informed
ior (Weinberger & Schwartz, 1990). These Taylor and Brown’s (1988) conclusion that
subscales show average correlations of be self-enhancement was necessary for posi
tween .4 and .5. tive adaptation and avoidance of depression.
In an analysis comparing several measures Along with other critiques of this conclusion
of repression-related constructs, including (see Kwan, John, Denny, Bond, & Robins,
the WAI, the M M B S, the Byrne Repression- 2 0 0 4 , for recent arguments), Norem and
Sensitization scale, Weinberger’s Repres Cantor (1986a) argued that different indi
sive Coping Scale, Sackeim and Gur’s Self- viduals will face similar situations with dif
Deception Questionnaire (Sackeim & Gur, ferent specific goals and that the importance
1979), and Paulhus’s Self-Deception Ques of and approaches used for self-protection
tionnaire (Paulhus & Reid, 1991), all except will vary across individuals (Norem & Can
33. P sy ch o lo g ical Defensiveness 485
tor, 1986b). They describe those differences back, whereas defensive pessimists’ attribu
in terms of the strategies individuals use. tions did not vary as a function of perfor
Strategies describe coherent patterns of mance feedback (Norem & Cantor, 1986a).
emotions, thoughts, motivations, and be The initial Optimism-Pessimism Pre
havior as they unfold during the process of screening Questionnaire (OPPQ; Norem &
pursuing goals (Cantor, Norem, Niedenthal, Cantor, 1986a) was a nine-item face-valid
Langston, & Brower, 1987; Norem, 1989). self-report measure. This measure has since
Although the steps of a particular strategy been revised to create the 17-item Defensive
can be described without reference to the Pessimism Questionnaire (DPQ; Norem,
goals or characteristics of the individual 2001). Questions are worded to reflect the
using that strategy, strategy coherence fol particular domain (e.g., academic, social,
lows from an individual’s understanding of recreational) under study, and strategy use
what he or she is trying to do in a given con shows average cross-situational correlations
text. That understanding, in turn, is influ between .30 and .50. The revised version
enced by prior experiences, self-knowledge, of the scale correlates at r = .65 with the
and other aspects of personality. original OPPQ and has both higher reli
Defensive pessimism describes a strategy ability (average Cronbach’s alpha = .78) and
used by anxious individuals who face the a cleaner factor structure. The DPQ has a
challenge of managing their anxiety to pre 3-year test-retest reliability of r = .55 among
vent it from interfering with achieving suc college women and a 2-month test-retest re
cess. Defensive pessimism involves expecting liability of r = .68 among male and female
negative outcomes prior to a performance, college students.
task, or specific situation and reflecting— Questions on the DPQ load satisfactorily
in concrete and vivid detail— on how those on one major unrotated factor. Oblique rota
negative outcomes could happen. Expect tion produces two correlated factors labeled
ing bad outcomes is hypothesized to protect Reflectivity and Pessimism. M ost research
one’s self-concept if those outcomes occur using the scales relies on a single score com
(hence the pessimism is defensive); therefore, puted by summing the reflectivity and pes
those using defensive pessimism would have simism items (after reverse scoring). Those
little need to make self-serving attributions scoring in the upper third of the distribution
to protect themselves. Moreover, thinking are categorized as defensive pessimists, and
about how negative outcomes might occur those in the bottom third are categorized as
requires a focus away from anxious feelings strategic optimists in prescreening, although
and toward task-relevant thoughts, which fa one can use continuous scores from the mea
cilitates planning of specific actions to avoid sure.
negative outcomes (Showers, 1988). Research indicates that pessimistic expec
M ost research on defensive pessimism tations play a crucial role in the defensive
has contrasted it with a strategic optimism. pessimism strategy. For example, experi
Those who use strategic optimism are typi mental manipulations that raise defensive
cally not aware of being anxious prior to pessimists’ expectations lead to lower sub
a performance or other self-relevant situa sequent performance (Norem & Cantor,
tion; they feel in control and are optimistic 1986b). Similarly, manipulations designed to
and confident of achieving good outcomes. get participants to feel that future successes
They set high expectations and actively dis were subjectively closer than future failures
tract themselves from thinking about pos lead to poorer performance for defensive
sible outcomes. When negative outcomes pessimists (Sanna, Chang, Carter, & Small,
do occur, they protect their self-concepts 20 0 6). M artin and his colleagues, however,
by attributing those outcomes to bad luck have data suggesting that pessimism and re
or other external factors beyond their con flectivity play different roles over time with
trol. One of the first experimental studies of respect to other variables (M artin, Marsh,
strategic optimism and defensive pessimism & Debus, 2 001a, 2001b ; M artin, Marsh,
demonstrated that strategic optimists vary W illiamson, & Debus, 2003). Exam in
their attributions in self-protective or self ing how expectations and reflectivity work
enhancing ways according to whether they separately and together will be important in
received, respectively, failure or success feed ongoing research.
486 V. M O T IV A T IO N A L D IS P O S IT IO N S
The DPQ shows small to moderate nega Research supports the conclusion that
tive correlations with Extraversion and defensive pessimists’ preparatory negative
small to moderate positive correlations with thinking is importantly different from the
Neuroticism, small positive correlations negative rumination of depressives. Showers
with Conscientiousness, small negative cor and her colleagues (Showers, 1992; Show
relations with Agreeableness, and no consis ers & Reuben, 1990) found that defensive
tent correlation with Openness. It correlates pessimists did not use the avoidant coping
negatively and moderately with disposition methods found among those with depres
al optimism, as measured by the Revised sive disorders and that defensive pessimists
Life Orientation Test (Scheier, Carver, & did not persist in feeling anxious or in ru
Bridges, 1994). Defensive pessimism does minating after stressful events. The focus on
not bear much relationship to attributional the future rather than the past, along with
or explanatory pessimism. The DPQ has a the defensive pessimists’ ability to think in
small positive correlation with the Internal concrete terms that readily suggest specific
subscale of the Attributional Style Ques action, explains why defensive pessimists do
tionnaire (ASQ) and no correlation with not seem to be at risk for depression, even
the Stable or Global ASQ subscales (Peter though their perspective can be quite nega
son, 1991). Correlations between the DPQ tive (Hosogoshi & Kodama, 2 0 0 6 ; Norem,
and R -S are typically positive and small 2 0 0 6 ; Tomaya, 2005).
to moderate; those between the DPQ and Understanding how and why defensive
Monitoring are small and positive and those pessimists’ negativity is helpful rather than
between the DPQ and Blunting are small debilitating requires understanding the po
and negative (Norem, 2001). The DPQ cor tential problems created by anxiety. When
relates moderately negatively with both the we are anxious, the dominant response is
Self-Deception and Impression Management avoidance of or flight from whatever makes
subscales of the Balanced Inventory of De us anxious. Yet often that which makes us
sirable Responding (Paulhus & Reid, 1991). anxious (e.g., an attractive potential roman
tic partner) is also something we are moti
vated to approach. Anxious individuals thus
Defensive Pessim ism , Negative need a strategy to help them control their
T hinking, and Negative A ffect urge to flee and allow them instead to act ef
fectively to achieve their goals. One strategy,
People who use defensive pessimism reliably of course, is to suppress anxiety. Especially
report higher trait anxiety and lower self in the short term, that can be an effective
esteem, score higher in negative affect, have strategy, though certainly not one without
more negative expectations for their per costs and not one available to all individu
formances, report more goal conflict, and als. In performance situations, even if the
generate more negative potential outcomes felt emotion of anxiety is repressed, the au
and plans than those who use strategic opti tonomic correlates of that anxiety may inter
mism, despite comparable past performance fere with performance. Suppression requires
(Cantor et al., 1987; Norem & Illingworth, effort, takes attentional capacity away from
1993, 2 0 0 4 ; Sanna, 1998). the task at hand, and may interfere with
There is also evidence that defensive pes performance as a result of ironic processes
simists are simultaneously motivated by the (Wegner, 1989).
desire to avoid failure and the motivation An alternative strategy is self-handi-
to achieve success. They focus on specific capping, by which individuals preemptively
performance-oriented goals that include provide themselves with a less incriminating
both avoiding failure and doing well, and attribution for failure. For example, a social
they have a higher ratio of negative to positive ly anxious individual may down a few stiff
self-knowledge than strategic optimists (El drinks before approaching a romantic pros
liot & Church, 2 0 0 3 ; Yamawaki, Tschanz, pect (Jones & Berglas, 1999), thereby both
& Feick, 2004). Perhaps as a result, they are dulling felt anxiety and providing a conve
likely to feel conflicted, particularly in situ nient explanation for rejection. This strategy
ations in which they value the success that involves considerable potential costs, howev
might be obtained. er, because most people do not become more
33. P sy ch o lo g ical D efensiveness 487
socially adept and attractive to others when Trying to “think positively” or “just
they are intoxicated. Indeed, repeated reli relax” (both frequent admonitions directed
ance on the handicap is likely to lead to de toward defensive pessimists) also interferes
teriorating outcomes over time: “Obnoxious with performance for defensive pessimists.
drunk” is hardly a less incriminating social In an experiment designed to investigate the
reputation than “socially awkward.” effect of visualization techniques on perfor
Defensive pessimism, in contrast to self- mance, defensive pessimists who listened to
handicapping, helps anxious individuals mastery or relaxation imagery recordings
focus on thoughts relevant to successful per performed significantly more poorly than
formance of the task at hand. An anxious those who listened to guided imagery re
student using defensive pessimism would cordings of what might go wrong, whereas
think about the possibility of failure on an strategic optimists performed best in the re
upcoming test by focusing on the process of laxation imagery condition and significantly
preparing for and taking the test. She might worse in the other two conditions (Spencer
first imagine reading the exam and not rec & Norem, 1996).
ognizing important terms or phrases and This pattern occurs across a variety of
then rehearse with note cards twice a day studies and samples. “Prefactual” and
for the week before the test. Defensive pes “counterfactual” thinking are mental simu
simism involves breaking down large goals lations of possible events prior to their oc
into smaller pieces that resemble “imple currence or after the fact and can involve
mentation intentions,” which, according to either upward mental simulation (thinking
literature on goal pursuit, provide a clear of outcomes better than those expected or
guide for translating abstract motivations experienced) or downward mental simula
into action (Gollwitzer, 1999); this should tion (thinking of outcomes worse than those
ultimately reduce anxiety. imagined or experienced). Overall, defen
Research supports the hypothesis that de sive pessimists prefer to engage in prefactual
fensive pessimism is a self-regulation strat thinking, whereas strategic optimists pre
egy that helps manage anxiety (Norem, fer to engage in downward counterfactual
2008). Defensive pessimists who were dis thinking when performance is disappoint
tracted prior to a performance task in a ing (Sanna, 1996). Each group suffers per
laboratory study scored lower on that task formance decrements when using the other
and felt less in control and more anxious groups’ preferred simulation.
than when they worked through possible “Cheering up” is ineffective in helping de
outcomes before the task; anxiety mediated fensive pessimists to control anxiety. Even
these results (Norem & Illingworth, 1993, though defensive pessimists can be put into
Study 1). A conceptual replication of this more positive moods, their performance suf
study using experience-sampling method fers as a result (Norem & Illingworth, 2 0 0 4 ;
ology found that defensive pessimists who Sanna, 1998). Sanna (1998) showed that
were prompted to think through what they good moods interfere with prefactual mental
were doing as they worked on “real life” simulation for defensive pessimists. He ar
goals over the course of several weeks re gues that they use negative affect as a cue to
ported more progress compared with de work harder, which typically leads to better
fensive pessimists whose strategy was not performance. In contrast, strategic optimists
reinforced (1993, Study 2). In both stud rely on “mood repair” through downward
ies, strategic optimists showed the opposite counterfactuals (i.e., imagining that things
pattern, and there were no main effects of had come out even worse than they did)
strategy on performance outcomes. In other when experiencing negative outcomes.
words, defensive pessimists who used their
strategy performed significantly better than
defensive pessimists who did not, and both Defensive Pessim ism and Adaptation
defensive pessimists and strategic optimists
performed best (and equivalently well) under There are few differences in outcome be
conditions that matched their strategies and tween defensive pessimists and strategic
significantly worse in conditions that inter optimists when each group is using its pre
fered with their strategies. ferred strategy, and each group is vulnerable
488 V. M O T IV A T IO N A L D IS P O S IT IO N S
constructs, but currently we know relatively Rationale, reliability, and validity. J o u r n a l o f P er
so n a lity , 2 9 , 3 3 4 - 3 4 9 .
little about differential influences of the self
Cantor, N., Norem, J . K., Niedenthal, P. M., Langston,
across different constructs. We also need C. A., & Brower, A. (1987). Life tasks, self-concept
more systematic research on the contribu ideals, and cognitive strategies in a life transition.
tions of conscious and nonconscious pro jo u r n a l o f P erson ality a n d S o c ia l P sy ch olog y ,
cessing to the development and operation of 53(6), 1 1 7 8 -1 19 1 .
Chang, E. C. (1996). Cultural differences in opti
these individual differences. mism, pessimism, and coping: Predictors of subse
The research necessary to pursue inte quent adjustment in Asian American and Caucasian
gration across constructs will be challeng American college students. J o u r n a l o f C o u n selin g
ing, because we will need to measure sev P sy ch olog y , 43(1), 1 1 3 - 1 2 3 .
Chang, W. C., & Sivam, R.-W. (200 4 ). Constant vigi
eral constructs and to use different kinds of
lance: Heritage values and defensive pessimism in
operationalizations within the same study coping with severe acute respiratory syndrome in
(e.g., physiological and self-report methods, Singapore. A sian J o u r n a l o f S o c ia l P sy ch olog y , 7,
self- and observer reports, independent as 35-53.
Cramer, P. (1995). Identity, narcissism, and defense
sessments of adaptation) if we are to make
mechanisms in late adolescence. J o u r n a l o f R e
real progress. Nevertheless, we need not be se a r c h in P erson ality , 29(3), 3 4 1 - 3 6 1 .
defensive about how far we have come, nor Crowne, D. P., & Marlowe, D. (1960). A new scale
anxious about prospects for future research. of social desirability independent of psychopathol
ogy. jo u r n a l o f C o n su ltin g P sy ch olog y , 6 6 , 5 4 7 -
The appropriate methods are available, and
555.
there is a deep, broad, and fascinating foun Davidson, K., & MacGregor, M. W. (1998). A critical
dation on which to build. appraisal of self-report defensive mechanism mea
sures. jo u r n a l o f P erson ality , 6 6 , 9 6 5 - 9 9 2 .
Davis, P. J ., & Schwartz, G. E. (1987). Repression and
the inaccessibility of affective memories. J o u r n a l o f
References P erso n a lity a n d S o c ia l P sy ch olog y , 5 2 , 1 5 5 - 1 6 2 .
Elliot, A. J ., & Church, M . A. (2003). A motiva
Baumeister, R. F., &C Cairns, K. B. (1992). Repression
tional analysis of defensive pessimism and self-
and self-presentation: When audiences interfere
handicapping. J o u r n a l o f P erso n a lity a n d S o c ia l
with self-deceptive strategies, j o u r n a l o f P ers o n a l
P sy ch o lo g y , 73(3), 3 6 9 - 3 9 6 .
ity a n d S o c ia l P sy ch olog y , 6 2 , 8 5 1 - 8 6 2 .
Eriksen, C. W. (1966). Defense against ego-threat in
Baumeister, R. F., Dale, K., & Sommer, K. L. (1998). memory and perception. J o u r n a l o f A b n o r m a l a n d
Freudian defense mechanisms and empirical find S o c ia l P sy ch olog y , 47, 2 3 0 - 2 3 5 .
ings in modern social psychology: Reaction form a Eronen, S., Nurmi, J. E., & Salmela Aro, K. (1998).
tion, projection, displacement, undoing, isolation,
Optimistic, defensive-pessimistic, impulsive and
sublimation, and denial, j o u r n a l o f P erson ality , 6 6,
self-handicapping strategies in university environ
1081-1124.
ments. L e a r n in g a n d In stru c tio n , 8(2), 159-177.
Bonanno, G. A., Davis, P. J., & Singer, J. L. (1991). Eysenck, H. J. (1967). T h e b io lo g ic a l b a sis o f p e r s o n
The repressor personality and avoidant information ality. Springfield, II.: Thomas.
processing: A dichotic listening study, j o u r n a l o f Eysenck, H. J. (1973). O n ex tr a v er sio n . New York:
R e se a rc h in P erson ality , 25(4), 3 8 6 - 4 0 1 . Wiley.
Bonanno , G. A., & Singer, J. L. (1995). Repressive Freud, A. (1946). T h e e g o a n d m ech a n is m s o f d e fe n c e .
personality style: theoretical and methodologi Oxf or d, UK: International Universities Press.
cal implications for health and pathology. In J. L. Freud, S. (1914). P sy c h o p a th o lo g y o f e v e ry d a y life (A.
Singer (Ed.), R e p re ss io n a n d d is s o c ia tio n : I m p lic a A. Brill, Trans.). Oxford, UK: Macmillan.
tio n s f o r p e r s o n a lity th eo r y , p sy c h o p a th o lo g y , a n d Garcia, T. (1995). T he role of motivational strategies
h ea lth (pp. 4 3 5 - 4 7 0 ) . Chicago: University of Chi in self-regulated learning. In P. R. Pintrich (Ed.),
cago Press. N e w d ire c tio n s fo r tea c h in g a n d lea rn in g (Vol. 63,
Brower, A. M ., & Ketterhagen, A. (20 04 ). Is there an pp. 2 9 - 4 2 ) . San Francisco: Jossey-Bass.
inherent mismatch between how black and white Gollwitzer, P. M. (1999). Implementation intentions:
students expect to succeed in college and what their Strong effects of simple plans. A m erica n P s y c h o lo
college expects from them? J o u r n a l o f S o c ia l Issu es, g ist, 5 4 , 4 9 3 - 5 0 3 .
60(1), 9 5 - 1 1 6 . Gray, J . A. (1972). The psychophysiological nature of
Bruneau, N., Roux, S., Perse, J., &c Lelord, G. (1984). introversion-extraversion: A modification of Ey
Frontal evoked responses, stimulus intensity con senck’s theory. In V. D. Nebylitsyn & J . A. Gray
trol, and the extraversion dimension. A n n als o f the (Eds.), B io lo g ic a l b a sis o f in d iv id u a l b e h a v io r
N ew Y ork A c a d e m y o f S cien ces, 4 2 5 , 5 4 6 - 5 5 0 . (pp. 3 7 2 - 3 9 9 ) . London: Academic Press.
Bruner, J. S., & Postman, L. (1947). Emotional selec Hansen, C. H., & Hansen, R. D. (1988). Repression
tivity in perception and reaction. J o u r n a l o f P ers o n of emotional tagged memories: The architecture of
a lity , 16, 6 9 - 7 7 . less complex emotions. J o u r n a l o f P erson ality a n d
Buck, R. (1976). H u m an m o tiv a tio n a n d em o tio n . S o c ia l P sy ch olog y , 8 8 , 8 1 1 - 8 1 8 .
New York: Wiley. Herzog, T. R „ Williams, D. M ., & Weintraub, D. J.
Byrne, D. (1961). The Repression-Sensitization Scale: (1985). Meanwhile, back at personality ranch: The
33. P sy ch o lo g ical Defensiveness 491
augmenters and reducers ride again, j o u r n a l o f P er Mitchell, D. C. (1998). Repression and relief: Mood
so n a lity a n d S o c ia l P sy ch olog y , 4 8 (5 ), 1 3 4 2 - 1 3 5 2 . and cardiovascular changes following threat, think
Higgins, E. T. (2005). Value from regulatory fit. C u r ing about threat, and threat removal for repressors
ren t D irec tio n s in P sy c h o lo g ic a l S cien ce, 14, 2 0 9 - and nonrepressors. D iss er ta tio n A b stra cts In te r n a
213. tio n a l, 5S(9-B), 5180. (University Microfilms No.
Hock, M ., & Krohne, H. W. (20 0 4). Coping with 1 9 9 8 - 9 5 0 0 6 -0 8 1 )
threat and memory for ambiguous information: Morgan, A. H., Lezard, F., Prytulak, S., Sc Hilgard, F^.
Testing the repressive discontinuity hypothesis. R. (1970). Augmenters, reducers, and their reaction
E m o tio n , 4, 6 5 - 8 6 . to cold pressor pain in waking and suggested hyp
Hosogoshi, H., 8c Kodama, M . ( 2 0 06 ). Examination notic analgesia. J o u r n a l o f P erson ality a n d S o c ia l
of psychological well-being and subjective well P sy ch olog y , /6(1), 5 - 1 1 .
being in defensive pessimists. J a p a n e s e J o u r n a l o f Norem, J. K. (1989). Cognitive strategies as personal
P sy ch olog y , 77(2), 141 -1 48 . ity: Effectiveness, specificity, flexibility and change.
Hull, C. L. (1950). Simple qualitative discrimination In D. M. Buss 8c N. Cantor (Eds.), P erso n a lity p s y
learning. P s y c h o lo g ic a l R e v iew , 57, 3 0 3 - 3 1 3 . c h o lo g y : R ecen t tren d s a n d em e rg in g d ire ctio n s
Jones, E. F., 8c Berglas, S. (1999). Control of the at (pp. 4 5 - 6 0 ) . New York: Springer-Verlag.
tributions about the self through self-handicapping Norem, J. K. (2001). Defensive pessimism, optimism,
strategies: The appeal of alcohol and the role of un and pessimism. In E. C. Chang (Ed.), O p tim ism a n d
derachievement. In R. F. Baumeister (Ed.), T h e s e lf p es sim is m : Im p lic a tio n s fo r th eo ry , resea rch a n d
in s o c ia l p s y c h o lo g y (pp. 4 3 0 - 4 3 5 ) . Philadelphia: p r a c tic e (pp. 7 7 - 1 0 0 ) . Washington, DC: American
Psychology Press/Taylor & Francis. Psychological Association.
Kelly, J . A., St. l.awrence, J. S., Brasfield, T. L., Lemke, Norem, J. K. ( 2 0 06 ). Defensive pessimism: Positive
A., Amidei, T., Roffman, R. E., et al. (1990). Psy past, anxious present, and pessimistic future. In I,.
chological factors that predict AIDS high-risk ver Sanna 8c E. C. Chang (F.ds.), Ju d g m e n ts o v e r tim e:
sus AIDS precautionary behavior. J o u r n a l o f C o n T h e in te rp la y o f th ou g h ts, feelin g s, a n d b e h a v io r s
su ltin g a n d C lin ic a l P sy ch olog y , 58(1), 1 17-120. (pp. 3 4 - 4 6 ) . Oxford, UK: Oxfor d University Press.
Krohne, J. W. (1996). Individual differences in coping. Norem, J. K. (200 7). Defensive pessimism as a positive
In M. Zeidner & N. S. Endler (Eds.), H a n d b o o k o f self-critical tool. In E. C. Chang (Ed.), S elf-criticism
c o p in g : T h eo ry , research , a p p lic a tio n s (pp. 3 8 1 — a n d s e lf-e n h a n c e m e n t: T h eo ry , research , a n d clin i
4 09 ). Oxford, UK: Wiley. ca l im p lic a tio n s (pp. 8 9 - 1 0 4 ) . Washington, DC:
Kwan, V. S. Y., John, O. P., Denny, D. A., Bond, M. American Psychological Association Press.
H., & Robins, R. W. (20 0 4). Reconceptualizing Norem, J. K. (2008). Defensive pessimism, anxiety
individual differences in self-enhancement bias: and the complexity of self-regulation. S o c ia l a n d
An interpersonal approach. P sy c h o lo g ic a l R ev iew , P erson ality C o m p a ss , 2, 1 2 1 - 1 3 4 .
111(1), 9 4 - 1 1 0 . Norem, J. K., Sc Andreas Burdzovic, J . A. (20 07 ). Un
Lorig, T. S., Singer, J. L., Bonanno, G. A., & Davis, P. derstanding journeys: Individual growth analysis as
(1994). Repressor personality styles and E EC pat a tool for studying individual differences in change
terns associated with affective memory and thought over time. In A. D. Ong Sc M. v. Dulmen (Eds.),
suppression. C o g n itio n a n d P erson ality , 14(3), H a n d b o o k o f m e th o d s in p o s itiv e p sy c h o lo g y
203-210. (pp. 4 7 7 - 4 8 6 ) . London: Oxford University Press.
Martin , A. J ., Marsh, H. W., Sc Debus, R. L. (2001a). A Norem, J. K., Sc Cantor, N. (1986a). Anticipatory and
quadripolar need for achievement representation of post hoc cushioning strategies: Optimism and de
self-handicapping and defensive pessimism. A m er i fensive pessimism in “risky” situations. C o g n itiv e
can E d u c a tio n a l R e se a rc h J o u r n a l, 3 8 , 5 8 3 - 6 1 0 . T h era p y a n d R esea rch , 10(3), 3 4 7 - 3 6 2 .
Martin, A. J., Marsh, H. W., Sc Debus, R. I.. (2001b). Norem, J. K., Sc Cantor, N. (1986b). Defensive pessi
Self-handicapping and defensive pessimism: E x mism: Harnessing anxiety as motivation. J o u r n a l o f
ploring a model of predictors and outcomes from a P erson ality a n d S o c ia l P sy ch olog y , 51, 1 2 0 8 - 1 2 1 7 .
self-protection perspective. J o u r n a l o f E d u c a tio n a l Norem, J. K., 8c Chang, E. C. (2001). A very full glass:
P sy ch olog y , 9,i(l), 8 7 - 1 0 2 . Adding complexity to our thinking about the im
Martin , A. J ., Marsh, H. W., Williamson, A., & plications and applications of optimism and pessi
Debus, R. I.. (2003). Self-handicapping, defensive mism research. In FL. C. Chang (Ed.), O p tim ism a n d
pessimism, and goal orientation: A qualitative study p es sim is m : Im p lic a tio n s fo r th eo r y , resea rch a n d
of university students, lo u r n a l o f E d u c a tio n a l Psy p r a c tic e (pp. 3 4 7 - 3 6 7 ) . Washington, DC: American
c h o lo g y , 95(3), 6 1 7 - 6 2 8 . Psychological Association.
Mendolia, M., Moore, J ., & Tesser, A. (1996). Dispo Norem, J . K., 8c Illingworth, K. S. S. (1993). Strategy-
sitional and situational determinants of repression. dependent effects of reflecting on self and tasks:
J o u r n a l o f P erson ality a n d S o cia l P sy ch olog y , 70, Some implications of optimism and defensive pes
856-867. simism. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l
Miller, S. M. (1987). Monitoring and blunting: Valida ogy, 65(4), 8 2 2 - 8 3 5 .
tion of a questionnaire to assess styles of informa Norem, J. K., Sc Illingworth, K. S. S. (20 0 4). Mood
tion seeking under threat. J o u r n a l o f P erson ality and performance among defensive pessimists and
a n d S o c ia l P sy ch olog y , 52(2), 3 4 5 - 3 5 3 . strategic optimists. J o u r n a l o f R e se a rch in P erso n
Miller, S. M ., Sc Man agan, C. E. (1983). The interact ality, 3 8 , 3 5 1 - 3 6 6 .
ing effects of information and coping style in adapt Paulhus, D. L., 8c Joh n, O. P. (1998). Egoistic and
ing to gynecologic stress: Should the doctor tell all? moralistic biases in self-perception: The interplay of
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 45, self-deceptive styles with basic traits and motives.
223-236. J o u r n a l o f P erson ality , 66(6), 1 0 2 5 - 1 0 6 0 .
492 V. M O T IV A T IO N A L D IS P O S IT IO N S
Paulhus, D. L., & Reid, D. B. (1991). Enhancement Schwerdtfeger, A., 8c Baltissen, R. (2 002). Augmenting-
and denial in socially desirable responding. J o u r n a l reducing paradox lost?: A test of Davis et al.’s (1983)
o f P erso n a lity a n d S o c ia l P sy ch olog y , 6 0 , 3 0 7 —317. hypothesis. P erson ality a n d In d iv id u a l D iffe r e n c e s ,
Pavlov, I. P. (19 27). C o n d it io n e d re flex e s. London: 32, 257-271.
Oxford University Press. Showers, C. J. (1988). The effects of how and why
Pennebaker, J . W., 8c Chew, C. H. (1985). Deception, thinking on perceptions of future negative events.
EDA and inhibition of behavior. J o u r n a l o f P ers o n C o g n itiv e T h era p y a n d R e se a rc h , 1 2, 2 2 5 - 2 4 0 .
ality a n d S o c ia l P sy ch olog y , 4 9 , 1 4 2 7 -1 43 3. Showers, C. J . (19 92). The motivational and emotional
Perry, S. (20 07 , January). M a k in g le m o n a d e ? D e fe n consequences of considering positive or negative
siv e c o p in g sty le m o d e r a te s th e e ffe c t o f s te r e o ty p e possibilities for an upcoming event. J o u r n a l o f P er
th re a t o n w o m en 's m a th test p e r fo r m a n c e . Paper so n a lity a n d S o c ia l P sy ch olog y , 6 3(3), 4 7 4 - 4 8 4 .
presented at the annual meeting of the Society for Showers, C. J ., 8c Reuben, C. (1 990). Distinguishing
Social and Personality Psychology, Memphis, T N . defensive pessimism from depression: Negative ex
Peterson, C. (1991). T he meaning and measurement pectations and positive coping mechanisms. C o g n i
of explanatory style. P sy c h o lo g ic a l In q u iry , 2(1), tiv e T h era p y a n d R e se a rc h , 14(4), 3 8 5 - 3 9 9 .
1- 10 . Spence, K. W. (1936). The nature of discrimination
Petrie, A. (19 67). In d iv id u a lity in p a in a n d su fferin g . learning in animals. P sy c h o lo g ic a l R e v iew , 4 3,
Oxford, UK: University of Chicago Press. 427-449.
Petrie, K. J ., Booth, R. J., & Pennebaker, J. W. (1998). Spencer, S. M ., 8c Norem, J. K. (1996). Reflection
The immunological effects of thought suppression. and distraction: Defensive pessimism, strategic op
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 75(5), timism, and performance. P erson ality a n d S o c ia l
1264-1272. P sy ch o lo g y B u lletin , 2 2(4), 3 5 4 - 3 6 5 .
Rohrman n, S., Netter, P., Hennig, J . , 8c Hodapp, V. Taylor, J . A. (1953). A personality scale of manifest
(2003). Repression-sensitization, gender, and dis anxiety. J o u r n a l o f A b n o r m a l a n d S o c ia l P sy ch o l
crepancies in psychobiological reactions to ex am i og y , 4 8 , 2 8 5 - 2 9 0 .
nation stress. A n x iety , Stress a n d C o p in g : An In te r Taylor, S. E., 8c Brown, J . D. (1988). Illusion and well
n a tio n a l J o u r n a l, 16(3), 3 2 1 - 3 2 9 . being: A social psychological perspective on mental
Sackeim, H. A., & Gur, R. C. (1979). Self-deception, health. P sy c h o lo g ic a l B u lletin , 103(2), 1 9 3 - 2 1 0 .
other-deception, and self-reported psychopatholo Tomaya, M. (2 005). Influence of cognitive strategies
gy. J o u r n a l o f C o n su ltin g a n d C lin ica l P sy ch olog y , on test coping strategies and academic achieve
47(1), 2 1 3 - 2 1 5 . ment: Defensive pessimism and strategic optimism.
Sales, S. M. (1971). Need for stimulation as a factor in J a p a n e s e J o u r n a l o f E d u c a tio n a l P sy ch olog y , 53(2),
social behavior. J o u r n a l o f P erson ality a n d S o c ia l 220-229.
P sy ch olog y , 19(1), 1 2 4 - 1 3 4 . Traue, H. C., 8c Pennebaker, J. W. (1993). Inhibition
Sanna, L. J . (199 6). Defensive pessimism, optimism, and arousal. In H. C. Traue 8c J. W. Pennebaker
and simulating alternatives: Some ups and downs (Eds.), E m o tio n in h ib itio n a n d h ea lth (pp. 10-31).
of prefactual and counterfactual thinking. J o u r n a l Seattle, WA: Hogrefe 8c Huber.
o f P erso n a lity a n d S o c ia l P sy ch olog y , 71(5 ), 1 0 2 0 - Turvey, C., & Salovey, P. (1993). Measures of repres
10 36. sion: Converging on the same construct? I m a g in a
Sanna, L . J . (1998). Defensive pessimismand optimism: tion , C o g n itio n a n d P erson ality , 13(4), 2 7 9 - 2 8 9 .
The bitter-sweet influence of mood on performance Wegner, D. M . (1989). W hite b e a r s a n d o t h e r u n
and prefactual and counterfactual thinking. C o g n i w a n te d th ou g h ts: S u p p ression , o b s e s s io n , a n d th e
tio n a n d E m o tio n , 12(5), 6 3 5 - 6 6 5 . p s y c h o lo g y o f m e n ta l c o n tro l. New York: Penguin
Sanna, L. J . , Chang, E. C., Carter, S. E., 8c Small, E. Books.
M . ( 2 0 06 ). The future is now: Prospective temporal Weinberger, D. A., 8c Schwartz, G. E. (1990). Distress
self-appraisals among defensive pessimists and opti and restraint as superordinate dimensions of self-
mists. P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , reported adjustment: A typological perspective.
3 2 (6 ), 7 2 7 - 7 3 9 . J o u r n a l o f P erson ality , 5 8 , 3 8 1 - 4 1 7 .
Scheier, M. F., Carver, C. S., 8c Bridges, M . W. (1994). Weinberger, D. A., Schwartz, G. E., 8c Davidson, R. J.
Distinguishing optimism from neuroticism (and trait (1979). Low-anxious, high-anxious, and repressive
anxiety, self-mastery, and self-esteem): A reevalua coping styles: Psychometric patterns and behavioral
tion of the Life Orientation Test. J o u r n a l o f P ers o n and physiological responses to stress. J o u r n a l o f
ality a n d S o c ia l P sy ch olog y , 67(6), 1 0 6 3 - 1 0 7 8 . A b n o r m a l P sy ch olog y , 8 8 , 3 6 9 - 3 8 0 .
Schoneman, S. W. (2002). The role of the cognitive Weinstein, J ., Averill, J . R ., Opton, E. M. J . , 8c L a
coping strategy of defensive pessimism within the zarus, R. S. (1968). Defensive style and discrepancy
social-evaluative continuum. D iss er ta tio n A bstra cts between self-report and physiological indexes of
In te rn a tio n a l, 6 3 , 3 0 2 4 . (University Microfilms No. stress. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l
2002-95024-319) ogy, 10(4), 4 0 6 - 4 1 3 .
Schwerdtfeger, A. (2003). Using affective pictures in Yamawaki, N., Tschanz, B. T., 8c Feick, D. I.. (2004).
stead of white noise: Still different response patterns Defensive pessimism, self-esteem instability, and
for Petrie-style augmenters and reducers. P ers o n a l goal striving. C o g n itio n a n d E m o tio n , 18(2), 2 3 3 -
ity a n d In d iv id u a l D iffe r e n c e s , 3 4 (2 ), 2 5 3 - 2 6 2 . 249.
PART VI
Se l f - R elated D is p o s it io n s
CHAPTER 34
■ * a • A *m
A ll a n F e n ig s t e in
Self-Consciousness
S elf-consciousness refers to a relatively
stable individual difference in the ten
dency to direct attention and thought toward
as a Personality C haracteristic
495
496 V I. S E L F -R E L A T E D D IS P O S IT IO N S
nal conditions— or whether it is the result of Public self-consciousness has been found
a low threshold for external stimuli that draw to relate to variables that suggest an aware
attention to the self. The most relevant em ness of oneself as an object of attention from
pirical evidence on this question is research others, such as social anxiety (Fenigstein, et
examining the interaction between self- al., 1975; Leary, 1983), shyness (Schlenker
consciousness measures and experimental &C Weigold, 1990), sociability (Turner et
manipulations of self-awareness; the results, al., 1978), and self-monitoring, a tendency
however, are mixed. Consistent with the for toward managing one’s public impressions
mer possibility, some research has found a (e.g., Briggs, Cheek, & Buss, 1980; Fenig
ceiling effect: Self-awareness manipulations stein, 1979; Turner et al., 1978). That pri
have a stronger effect on persons low in self- vate and public self-consciousness relate to
consciousness than on those who are already separate sets of variables-— one concerned
at a high level of self-consciousness (e.g., with inner thought and the other dealing
Carver & Scheier, 1978). However, other re with social issues— offers further evidence
search has found that self-awareness manip that they are largely orthogonal dimensions;
ulations have the greatest impact on persons the presence of one neither precludes nor ne
high, rather than low, in self-consciousness, cessitates the other. Some studies, however,
consistent with the idea that people who have questioned that independence and sug
are high in self-consciousness are especially gested a conceptual overlap between the fac
sensitive to self-attention-inducing stimuli tors. For example, Trapnell and Campbell
(e.g., Brockner, 1979). Still other research (1999) found that the dimensions of private
has found that the combined effects of self- and public self-consciousness were both re
consciousness and self-awareness are addi lated to the Big Five factor of neuroticism
tive, suggesting that the disposition and the (see later in the chapter for a fuller discussion
manipulation are essentially two indepen of the implications of self-consciousness for
dent ways of varying the same psychologi various forms of psychopathology). It is im
cal state (e.g., Scheier, 1976). In sum, the portant to consider, however, that these rela
issue of whether self-consciousness is being tionships may be mediated by fundamentally
“pushed” from within the person or being different pathways to disordered behavior,
“pulled” from the outside has not yet been in which case it would still be appropriate to
resolved. regard the two self-consciousness factors as
relatively independent.
Carver and Glass (1976) provided evi
Relation o f Self-Consciousness dence of discriminant validity for the self-
Factors to O ther Personality Variables consciousness factors. As expected, neither
private nor public self-consciousness related
Research has suggested that the private and to measures of intelligence, need for achieve
public self-consciousness factors are rela ment, test anxiety, or impulsivity. Other
tively reliable dimensions (e.g., Fenigstein studies have shown little relation between
et al., 1975) that are largely independent of self-consciousness and measures of self
each other. Most studies have found a weak esteem (e.g., Brockner et al., 1983) or social
to moderate correlation between them, usu desirability (Turner et al., 1978). These find
ally below .30 (see Buss, 1980, and Carver ings are consistent with the idea that self-
& Scheier, 1981, for reviews; however, see consciousness is a relatively pure attentional
Wicklund & Gollwitzer, 1987, for an alter tendency; thought or attention may be direct
native view). ed toward aspects of the self that are either
Evidence of convergent validity has positive, neutral, or negative, but by itself,
been found for both factors. Private self- neither public nor private self-consciousness
consciousness is correlated with a number has a reliable relationship with any specific
of psychological measures to which it is value-laden, self-relevant content.
theoretically related, such as Openness to Since its development, the SCS has be
Experience (a Big Five factor), thoughtful come the primary means by which self-
ness, reflectiveness, and the use of visual consciousness, private or public, is measured.
imagery (e.g., Trapnell & Campbell, 1999; So far, there have been published translations
Turner, Scheier, Carver, & Ickes, 1978). of the SCS into at least 16 other languages
498 V I. S E L F -R E L A T E D D IS P O S IT IO N S
(Arabic, Chinese, Dutch, Estonian, French, ing attention toward the self as a social ob
German, Greek, Hebrew, Italian, Japanese, ject involves an external rather than a self
Persian, Polish, Portuguese, Spanish, Swed orientation in which the person’s main focus
ish, and Turkish). The scale has also shown is on what others think of him- or herself.
significant heuristic value, having appeared As Fenigstein (1987) pointed out, however,
in more than 1,3 0 0 published studies and this argument fails to recognize that for a
countless dissertations, but it should be said number of influential “self” theorists (e.g.,
that although SCS-related research is con Argyle, 1969; Mead, 1934), attention to or
tinuing presently, it has slowed down a great awareness of what others think of the per
deal since the 1980s. son, rather than detracting or distracting
from self-awareness, actually plays a criti
cal role in influencing how one views one
The Self-Attentional Properties o f self. In fact, Duval and Wicklund (1972), in
Private and Public Self-Consciousness their original formulation of self-awareness
theory, explicitly defined self-attention in
A question of fundamental importance to the terms of an external perspective, likening it
construct validity of the SCS is whether the to a process of figuratively coming outside
two types of self-consciousness both involve oneself to look back on oneself. In other
self-focused attention. Although the ques words, if public self-consciousness involves
tionnaire items for both subscales sought an “external” perspective on the self, using
to identify tendencies or characteristics that either the mind’s eye to view oneself from
were explicitly associated with attention to the “outside” (cf., Hass, 1984) or taking the
either private or public aspects of the self, a presumed viewpoint of another (e.g., C ool
more definitive response needs to go beyond ey, 1902), that perspective in no way com
face validity to independent sources of evi promises the idea that attention is being di
dence. Construct validity research concern rected toward oneself, even if that attention
ing the self-attentional properties of private is specifically focused on those aspects of the
self-consciousness has been clear and consis person that render him or her an object of
tent. Carver and Scheier (1978) showed that presumed interest to others.
private self-consciousness was associated A number of studies provide direct evidence
with self-focused responses on a sentence- for the idea that public self-consciousness
completion task. Other studies have dem involves attention toward the self. For ex
onstrated that private self-consciousness is ample, Franzoi and Brewer (1984), in a
related to easier access to information about naturalistic study of ongoing thought, found
the private self (e.g., Hull & Levy, 1979; that publicly self-conscious persons were es
Nasby, 1985; Turner, 1980). Finally, many pecially likely to think about themselves as
of the findings associated with private self- social objects who were being observed by
consciousness effectively parallel those as others. In addition, the trait of public self-
sociated with self-attention-inducing stimuli consciousness has been associated with ei
such as mirrors (e.g., Froming, Walker, & ther better memory for or memory biased
Lopyan, 1982; Scheier & Carver, 1977), in favor of information specific to the so
strongly suggesting that both the experi cially observable aspects of the person (e.g.,
mental and the dispositional findings could Nasby, 1989a; Turner, Gilliland, &C Klein,
be explained most parsimoniously in terms 1981). Finally, several studies have shown
of self-attention. Thus the idea that private that public self-consciousness is associated
self-consciousness measures attention to the with outcomes very similar to those induced
self has been generally accepted, even by by stimuli that heighten attention to the out
skeptics of the dispositional approach (e.g., wardly visible aspects of self, such as an ob
Gibbons, 1990; Wicklund & Gollwitzer, serving audience or a videotape camera (e.g.,
1987). Fenigstein, 1979, 1984; Froming & Carver,
Serious questions, however, have been 1981; Hass, 1984). Thus, both theoretically
raised regarding the self-attentional prop and empirically, strong and consistent evi
erties of public self-consciousness. Gibbons dence supports the self-attentional nature of
(1990), for example, has argued that direct public self-consciousness.
34. P riv a te and P ub lic S elf-C onsciousn ess 499
that are more consistent over time (e.g., tive affect. Hull and Young (1983) found
Nasby, 1989b) and that correspond more that following failure (and the experience of
closely with both peer-reported descrip negative affect), participants high, compared
tions (e.g., Bernstein & Davis, 1982; Fran- to low, in private self-consciousness engaged
zoi, 1983) and with participants’ subsequent in greater alcohol consumption, presumably
behavior (e.g., Scheier, Buss, & Buss, 1978; as a means of reducing their self-critical feel
Smith & Shaffer, 1986; Underwood & ings. Finally, in a test of the relation between
M oore, 1981). Overall, these findings sug private self-consciousness and the clarity of
gest that private self-consciousness is asso bodily experiences, Scheier, Carver, and
ciated with a heightened awareness of the Gibbons (1979) found that private self-
private aspects of oneself and an increased consciousness correlated with the ability to
tendency to act in accordance with those as resist bogus information on a taste test.
pects. Several studies suggest that heightened
In addition to its association with in awareness of private thoughts and feelings
creased awareness of self-relevant informa is associated not only with increased access
tion, private self-consciousness is associated to that information but also with an increase
with heightened awareness, along with in in the personal or subjective importance of
tensified experiences, of emotions and other those self-aspects. Cheek and Briggs (1982),
bodily sensations. (The possibility that for example, showed that persons high,
self-consciousness is related to increased re- in comparison with low, in private self-
sponsivity to emotional stimuli, rather than consciousness constructed their identities
increased aw areness of internal emotional out of the unique, idiosyncratic, and relative
states, has not been ruled out as an alterna ly private elements of their existence, rather
tive explanation; see, e.g., Hull, Slone, Me- than through any social connection with
teyer, & M atthews, 20 0 2 .) Some reviews others. Similarly, Fenigstein and Vanable
(e.g., Fejfar & Hoyle, 2 0 0 0 ; Gibbons, 1990) (1993) found that private self-consciousness
have emphasized the relation with negative was more closely associated with a private
affect (largely because of the preponderance self-esteem measure, evaluating how highly
of research on that relationship), but there is one valued oneself, than with a measure of
a general acceptance of the view that inten public self-esteem, assessing the extent to
sified affect, both positive and negative, is which persons believed they were valued by
related to private self-consciousness. Private others.
self-consciousness has been shown to cor Another indication of the relationship be
relate positively both with greater laughter tween private self-consciousness and the sub
in response to humorous stimuli (Porterfield jective importance of the self is found in stud
et al., 1988) and with stronger aggression in ies involving a conflict between situational
response to being angered (Scheier, 1976). standards and personal standards. When
Similarly, Scheier and Carver (1977) found faced with this conflict, persons high, com
that in response to both positive and nega pared with low, in private self-consciousness
tive emotional stimuli, participants high in tend to emphasize self-standards; they show
private self-consciousness reported more relatively little concern for social expecta
intense affective reactions than those low tions or conformity pressures but instead act
in private self-consciousness. In the area of in accordance with their own beliefs, val
sexuality, private self-consciousness is posi ues, or characteristics (e.g., Ellis & Holmes,
tively associated with responsivity to sexual 1982; Froming & Carver, 1981; Greenberg,
stimuli (Meston, 2 0 0 6 ). Outside the labo 1982; Scheier, 1980).
ratory, research showed that private self- As noted earlier, private self-consciousness
consciousness was related to both stress and increases the extent to which thoughts about
associated somatic symptoms among factory the private self are salient and accessible.
workers (Frone & M cFarlin, 1989). This increased availability has several inter
A novel application of the relationship be esting cognitive implications, each of which
tween self-consciousness and heightened neg effectively contributes to the subjective im
ative feelings argued that alcohol consump portance of the self. One effect of private self-
tion is a means of reducing self-awareness, consciousness, as posited in Hull and Levy’s
which would then minimize existing nega (1979) cognitive model of self-awareness, is
34. P riv a te and P ub lic S elf-C onsciousn ess 501
that conscious (Hull, Van Treuren, Ashford, had for them than by a personal sense of
Propsom, & Andrus, 1988), as well as non- self-esteem, compared with people low in
conscious (Hull et al., 20 0 2 ), information public self-consciousness (Fenigstein & Van-
is more likely to be processed or encoded able, 1993). In constructing their identities,
according to its self-relevance. In addition, they also emphasized those self-aspects that
by activating self-relevant thoughts, self- related to their public self, such as physical
consciousness may influence subsequent de characteristics and group affiliations, over
cisions, such as attributional judgments. A aspects of self that were more private in na
number of studies have found a relation be ture (Cheek & Briggs, 1982). In compari
tween heightened private self-consciousness son with privately self-conscious persons,
and the extent to which the self, relative to who are not especially concerned with the
external stimuli, is perceived as causally re opinions or desires of others, those who are
sponsible for events (e.g., Buss & Scheier, publicly self-conscious were more responsive
1976; Fejfar & Hoyle, 2 0 0 0 ; Fenigstein & to the perceived expectations of others, as
Carver, 1978). In summary, chronic attention evidenced by greater susceptibility to con
to the private self, in the form of disposition formity pressures (e.g., Froming & Carver,
al private self-consciousness, is associated 1981) and by a willingness to act in ways
with more accessible information regarding they thought others would approve, regard
private self-aspects, a higher subjective value less of their own private beliefs (e.g., Green
or importance of those self-aspects relative berg, 1982; Scheier, 1980).
to external influences, and an increase in the The acute awareness of the socially “pre
extent to which private self-oriented cogni sented” self on the part of publicly self-
tion influences subsequent thoughts or judg conscious persons is also related to a height
ments. ened concern with the impressions they
convey to others, particularly with respect
to physical appearance (e.g., Cash & La-
Research on Public Barge, 1996; Franzoi, Anderson, & From-
Self-Consciousness melt, 1990; Solomon & Schopler, 1982),
although these concerns may be especially
The psychological consequences of public pronounced in women. Miller and Cox
self-consciousness may be understood as the (1982) found that among women who were
result of a process of attending to aspects of preparing to have their pictures taken, high
the self that are public or observable to other er public self-consciousness was associated
people or that constitute one’s group identity. with applying a greater amount of makeup.
Attending to these “outer” aspects of oneself Furthermore, public self-consciousness is
should make one more cognizant of (and, significantly greater among women who
presumably, more knowledgeable about) are restrained eaters (Blanchard & Frost,
those self-aspects. Consistent with this argu 1983) or who have eating disorders (Strie-
ment, research has shown that persons high gel-Moore, Silberstein, & Rodin, 1993) than
in public self-consciousness, compared with among normal controls.
those who are low in it, had easier access to Publicly self-conscious people’s con
and better memory for information about cerns with how they are viewed by others
the externally displayed aspects of them extend beyond appearance. Shepherd and
selves (e.g., Agatstein & Buchanan, 1984; Arkin (1989) found that in preparing for a
Nasby, 1989a; Turner et al., 1981) and were difficult task, persons high in public self-
more aware of their appearance, their ges consciousness were more likely than those
tures, and the public impressions that they low in it to engage in self-handicapping, a
were conveying (e.g., Gallaher, 1992; Tobey strategy designed to protect impressions
&C Tunnell, 1981). of one’s competence in the eyes of others.
Evidence also suggests that chronic atten Another type of impression-management
tion to the public aspects of self increases strategy was found in a study of road rage:
the extent to which the socially presented, After being angered by another driver,
public self is valued as important. Persons drivers high, compared with low, in public
high in public self-consciousness were more self-consciousness were less likely to drive
affected by the esteem they thought others aggressively (Millar, 20 0 7 ). Finally, in a
502 V I. S E L F -R E L A T E D D IS P O S IT IO N S
sight into one’s own mental and emotional ing or escaping unpleasant or threatening
experiences, the research in support of that truths about themselves (e.g., Baumeister,
claim is inconsistent and equivocal (see Sil 1991; Gibbons, 1990). In addition, it has
via & Gendolla, 2001). been argued that self-focused attention is
Much of this research suggests a psycholog inextricably linked to greater rumination
ical equivalent to the Heisenberg uncertainty about self-critical thoughts, physical symp
principle: Observing a phenomenon changes toms, or dysphoric emotions, which in turn
it. Self-attention or self-consciousness often increases the psychological distress associ
seem to change the nature of whatever as ated with those experiences (e.g., Lyubomi
pect of self is being observed. When atten rsky & Nolen-Hoeksema, 1993; Wells &
tion or thoughtful scrutiny is directed to Matthews, 1994; Wood, Saltzberg, & Gold-
ward self-relevant psychological phenomena samt, 1990). This section examines three
such as ongoing affect (e.g., Scheier & Carv specific forms of psychological dysfunction
er, 1977), judgments regarding causality for that have been related to self-consciousness:
one’s behavior (Buss & Scheier, 1976), the anxiety, depression, and paranoia.
reasons underlying one’s attitudes or feelings
(e.g., W ilson, 1990), or the self as a social
Anxiety
target (e.g., Fenigstein, 1984), the effect of
“looking inward,” rather than being clarify Researchers have long suspected that height
ing or “insightful,” instead tends to alter or ened attention to the self promotes anxiety,
distort whatever aspect of the self is being and several studies support this notion (e.g.,
thought about. Although it has sometimes Izard, 1972; Sarason, 1972; Wine, 1982).
been assumed that self-consciousness fa Private self-focus has been found to increase
cilitates self-knowledge, research suggests awareness of and responsiveness to affective
some skepticism regarding the accuracy of stimuli generally (e.g., Scheier 8c Carver,
the self-insights gained through self-focused 1977) and, more specifically, to the nega
attention. tive affect of anxiety (e.g., Csikszentmih
alyi & Figurski, 1982; Gibbons, 1990). To
the extent that much of ordinary thought
Self-Consciousness deals with current concerns and unfinished
and Psychological Disorders projects (e.g., Singer, 1988), heightened at
tention to these thoughts, as in the case of
If self-consciousness undermines self-insight, self-conscious persons, is likely to provoke
then to the extent that accurate self-insight discomforting emotions such as anxiety,
is a prerequisite for mental health (as is the fear, sadness, and anger. In addition to this
prevailing wisdom in clinical psychology), “amplification” effect involving the intensi
self-consciousness may actually contribute fication of negative affect, self-consciousness
to psychological disorder. That possibility is may also contribute to anxiety by directly
explored in this section. increasing negative thinking (e.g., Beck &
Although relatively little research suggests Clark, 1988) or by fostering ruminative,
that self-consciousness has positive conse nonproductive coping responses in which
quences for mental health, there has been a people focus on their inadequacies (e.g.,
virtual explosion of research in recent years Beck & Fernery, 1985; Nolen-Hoeksema,
examining the relation between self-focused 2004) or by contributing to performance
attention and clinical disorders (e.g., Ingram, deficits (e.g., Carver, Peterson, Follansbee,
1990; Nolen-Hoeksema, 2 0 0 4 ; Pyszczyn- & Scheier, 1983; W ine, 1971).
ski, Greenberg, Hamilton, & N ix, 1991). A An important insight into the relation
number of different pathways by which self ship between self-attention and anxiety was
attention or self-scrutiny produces decre provided by Izard’s (1972) analysis of emo
ments in mood or self-esteem have been ex tion, which suggests that anxiety in a per
amined— by focusing attention on personal formance context almost always involves an
shortcomings (e.g., Brockner, 1979; Duval element of self-evaluation, along with the
&C Wicklund, 1972), by attuning the self to concern and apprehension that results from
inner contradictions (e.g., Scheier & Carver, self-evaluation. Self-consciousness has also
1980), and by preventing people from ignor been conceptualized in terms of a real or
34. P riv a te and P ub lic S elf-C onsciousn ess 505
tion toward oneself, the self-conscious per M illon, 1981; Shapiro, 1965) but instead are
son may come to believe that others are also inclined to regard negative events as evidence
directing attention toward him or her. That for another person’s malevolent intentions
is, by heightening the salience and subjective toward them. Given this dispositional at
importance of the self, self-consciousness tribution of hostile intent, minor slights be
may contribute to the distinctly paranoid come major insults, and, eventually, the ac
tendency to perceive oneself as the target of cumulation of such occurrences constitutes
others’ thoughts and actions. That misper evidence for a paranoid view of the world
ception, as, for example, when the laughter as a hostile and threatening place. Recogniz
of others is seen as self-directed, is common ing that perspective, other manifestations of
ly referred to as an idea (or delusion) of refer paranoia then become comprehensible: sus
ence and is regarded as one of the hallmark picion and guardedness, selective attention
characteristics of paranoid thinking (e.g., to signs of trickery or exploitation, misin
Cameron, 1943; M agaro, 1980; M illon, terpretation of apparently harmless events
1981; Swanson, Bohnert, & Smith, 1970). as malevolent, readiness to take offense and
Paranoia is characterized by self- counterattack, and blaming others for one’s
referential distortions in the way in which difficulties. As a result, hostilities become
others’ actions are perceived and interpret intensified, suspicions are confirmed, and
ed. The relation of paranoia to public self- enemies are found everywhere.
consciousness, which involves a heightened Another means by which self-focus may
awareness of those aspects of the person that relate to paranoia involves the inability of
are potentially observable by others, should, paranoid people to understand the motiva
then, not be surprising. As suggested earlier, tions and perspectives of others (e.g., M il
awareness of one’s own observability to oth lon, 1981) or to examine behavior from any
ers can easily induce a feeling of visibility viewpoint other than their own (e.g., Sha
or conspicuousness (e.g., Fenigstein &c Van- piro, 1965). Research has shown that public
able, 1992), resulting in the paranoid-like self-consciousness, perhaps by heightening
assumption that others are aware of one’s the salience of one’s own perspective, is like
presence and thus are likely to be acting with wise associated with a diminished capacity
the person in mind (e.g., Fenigstein, 1979). It to consider others’ viewpoints (e.g., Fenig
is not unreasonable at that point to engage stein &C Abrams, 1993). In this way, self-
in paranoid interpretations of others’ behav consciousness may contribute to the narrow
ior, transforming insignificant and irrelevant ness and rigidity of paranoid thought— the
events into ones that appear to have personal failure to examine events critically or in a
relevance so that, for example, the appear broader context, the extreme selectivity in
ance of a stranger on the street is taken as processing information, and the unwilling
an indication that one is being watched or ness to consider alternative perspectives.
plotted against (e.g., Fenigstein, 1984). Thus a good deal of evidence suggests that
Although the research on self-conscious public self-consciousness is associated with a
ness and paranoia has been largely con number of thought patterns— in particular,
cerned with the direct association between personalistic, self-referential inferences, and
self-consciousness and various forms of rigid egocentrism— that are similar to those
paranoid behavior, self-focus may also be observed in paranoid ideation, and these
related to some critical mediating mecha similarities may help explain the relationship
nisms that affect paranoia. In particular, to between self-consciousness and paranoia.
interpret events as if they were intentionally
directed toward the self is to engage in per-
sonalistic thinking (Fieider, 1958; Jones & Conclusion
Davis, 1965). This type of thinking, in which
another’s behavior is interpreted in terms of The purpose of this chapter has been to pro
their personal intent or dispositions, is both vide a broad overview of the origins and cur
characteristic of paranoia and related to rent understanding of the personality con
public self-consciousness (Fenigstein, 1984). struct of self-consciousness. Growing out of
Paranoid people, for example, rarely accept self-awareness theory, one of the most sig
the idea that bad things just happen (e.g., nificant discoveries associated with the scale
3 4. P riv a te and P u b lic S elf-C onsciousn ess 507
Public and private self-awareness: When personal Ingram, R. E. (1990). Self-focused attention in clinical
attitudes conflict with societal expectations. J o u r n a l disorders: Review and a conceptual model. P sy ch o
o f E x p e r im e n ta l S o c ia l P sy ch olog y , 18, 4 7 6 - 4 8 7 . lo g ic a l B ulletin , 107, 1 5 6 - 1 7 6 .
Frone, M . R., & McEarlin, D. B. (1989). Chronic o c Ingram, R. K., I.umry, A. E., Cruet, D., & Sieber, W.
cupational stressors, self-focused attention, and (1987). Attentional processes in depressive disor
well-being: Testing a cybernetic model. J o u r n a l o f ders. C o g n itiv e T h e o r y a n d R esea rch , 1 /, 3 5 1 - 3 6 0 .
A p p lied P sy c h o lo g y , 74, 8 7 6 - 8 8 3 . Izard, C. (1972). P attern s o f em o tio n : A n ew an aly sis
Gallaher, P. (1992). Individual differences in nonverbal o f a n x ie ty a n d d ep r es sio n . New York: Academic
behavior: Dimensions of style. J o u r n a l o f P ers o n a l Press.
ity a n d S o c ia l P sy ch olog y , 6 3 , 1 3 3 - 1 4 5 . James, W. (1890). T h e p rin cip le s o f p s y c h o lo g y (Vol.
Gibbons, F. X . (1990). Self-attention and behavior: A 1). New York: Holt.
review and theoretical update. In L. Berkowitz (Ed.), Jones, E. E., & Davis, K. E. (1965). From acts to dis
A d v a n c es in e x p e r im e n ta l s o c ia l p s y c h o lo g y (Vol. positions: The attribution process in person percep
2 3 , pp. 2 4 9 - 3 0 3 ) . New York: Academic Press. tion. In I.. Berkowitz (Ed.), A d v a n c es in e x p e r im e n
Gilovich, T., Medvec, V. H., & Savitzky, K. (200 0). ta l s o c ia l p s y c h o lo g y (Vol. 2). New York: Academic
The spotlight effect in social judgment: An egocen Press.
tric bias in estimates of the salience of one’s own Jung, C. G. (1957). T h e u n d isc o v e re d self. New York:
actions and appearance. J o u r n a l o f P erson ality a n d New American Library.
S o c ia l P sy ch olog y , 79, 2 1 1 - 2 2 2 . Kihlstrom, J. F., Cantor, J., Albright, J ., Chew, B.,
Goffman , E. (1959). T h e p r e s e n ta tio n o f s e l f in e v e r y Klein, S., & Niedenthal, P. ( i 988). Information pro
d a y life. Garden City, NY: Doubleday. cessing and the study of the self. In I.. Berkowitz
Greenberg, J. (1982). Self-image versus impression (F.d.), A d v a n c es in e x p e r im e n ta l s o c ia l p s y c h o lo g y
management in adherence to distributive justice (Vol. 21, pp. 1 4 5 -1 78 ). New York: Academic Press.
standards: The influence of self-awareness and self- Klein, S. B „ & Loftus, J. (1988). T he nature of self
consciousness. J o u r n a l o f P erson ality a n d S o cia l referent encoding: T he contributions of elaborative
P sy ch olog y , 4 4 , 5 - 1 9 . and organizational processes. J o u r n a l o f P ers o n a l
Greenberg, J ., & Pyszczynski, T. (1986). Persistent high ity a n d S o c ia l P sy ch olog y , 5 5, 5 - 1 1 .
self-focus after failure and low self-focus after suc Kraepelin, F^. (1915). P sy ch iatrie: E in L eh r h u c h [P sy
cess: T he depressive self-focusing style. J o u r n a l o f c h ia try : A t e x t b o o k ] (7th ed.). Leipzig, Germany:
P erso n a lity a n d S o c ia l P sy ch olog y , 50, 1 0 3 9 - 1 0 4 4 . Barth.
Greenwald, A. G. (1980). T he totalitarian ego: Fabri Leary, M . R. (1983). Social anxiousness: The construct
cation and revision of personal history. A m er ica n and its assessment. J o u r n a l o f P erson ality A ssess
P sy ch olog ist, 3 5 , 6 0 3 - 6 1 8 . m en t, 47, 6 6 - 7 5 .
Hass, R. G. (1984). Perspective taking and self- Leary, M. R. (1995). S elf-p re se n ta tio n : Im p ressio n
awareness: Drawing an E on your forehead. J o u r m a n a g e m e n t a n d in te r p e r s o n a l b eh a v io r . Madison,
n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 4 6 , 7 8 8 - W I: Brown & Benchmark.
798. Leary, M. R., & Kowalski, R. M. (1995). S o c ia l a n x i
Heider, F. (1958). T h e p s y c h o lo g y o f in te r p e r s o n a l re ety. New York: Guilford Press.
la tio n s. New York: Wiley. Leary, M. R., &c Meadows, S. (1991). Predictors, elici-
Heinemann, W. (1979). T he assessment of private and tors, and concomitants of social blushing. J o u r n a l o f
public self-consciousness: A German replication. P erson ality a n d S o c ia l P sy ch olog y , 6 0 , 2 5 4 - 2 6 2 .
E .uropean J o u r n a l o f P erson ality , 9, 3 3 1 - 3 3 7 . Lewinsohn, P. M., Hoberman, H., Teri, I.., & Haut-
Hope, D. A., & Heimberg, R. G. (1988). Public and zinger, M. (1985). An integrative theory of depres
private self-consciousness and social phobia. J o u r sion. In S. Reiss & R. Bootzin (Eds.), T h e o r e t ic a l is
n a l o f P erson ality A ssessm en t, 5 2 , 6 2 6 - 6 3 9 . su es in b e h a v io r th era p y (pp. 3 3 1 - 3 5 9 ). New York:
Hull, J. G., Levinson, R. W., Young, R. D., & Sher, Academic Press.
K. J. (1983). Self-awareness-reducing effects of alco Lyubomirsky, S., & Nolen-Hoeksema, S. (1993). Self-
hol consumption. J o u r n a l o f P erson ality a n d S o cia l perpetuating properties of dysphoric rumination.
P sy ch olog y , 4 4 , 4 6 1 - 4 7 3 . J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 6 5,
Hull, J. G., & Levy, A. S. (1979). The organizational 339-349.
functions of self: An alternative to the Duval and Magaro, P. A. (1980). C o g n itio n in s c h iz o p h r e n ia a n d
Wicklund model of self-awareness. J o u r n a l o f P er p a r a n o ia : T h e in te rp r eta tio n o f c o g n itiv e p r o c es se s.
s o n a lity a n d S o c ia l P sy ch olog y , 37, 7 5 6 - 7 6 8 . Hillsdale, NJ: Erlbaum.
Hull, J. G., Slone, L. B., Meteyer, K. B., &. M a t Markus, H. (1977). Self-schemata and processing in
thews, A. R. (2002). The nonconsciousness of self- formation about the self. J o u r n a l o f P erson ality a n d
consciousness. J o u r n a l o f P erson ality a n d S o cia l S o c ia l P sy ch olog y , 3 5 , 6 3 - 7 8 .
P sy ch olog y , 8 3 , 4 0 6 - 4 2 4 . Mead, G. PL (1934). M in d, self, a n d s o c iety . Chicago:
Hull, J . G., Van Treuren, R. R., Ashford, S. J., Prop- University of Chicago Press.
som. P., & Andrus, B. W. (1988). Self-consciousness Mes ton, C. M. (200 6). The effects of state and trait
and the processing of self-relevant information. self-focused attention on sexual arousal in sexually
fo u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 54, functional and dysfunctional women. B e h a v io r R e
'4 5 2 - 4 6 5 . s e a r ch a n d T h era p y , 4 4 , 5 1 5 - 5 3 2 .
Hull, J . G., & Young, R. D. (1983). Self-consciousness, Millar, M. (20 07 ). The influence of public self-
self-esteem, success-failure as determinants of alco consciousness and anger on aggressive driving.
hol consumption in male social drinkers. J o u r n a l o f P erson ality a n d In d iv id u a l D iffe r e n c e s , 4 3 , 2 1 1 6 -
P erson ality a n d S o c ia l P sy ch olog y , 4 4 , 1 0 9 7 -11 09 . 2126.
510 V I. S E L F -R E L A T E D D IS P O S IT IO N S
Miller, L. C., & Cox, C. L. (1982). For appearance’s Raichle, K. A., Christensen, A. J ., Ehlers, S., Moran,
sake: Public self-consciousness and makeup use. P. J ., Karnell, L., & Funk, G. (2001). Public and
P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , 8, private self-consciousness and smoking behavior
748-751. in head and neck cancer patients. A n n als o f B e h a v
Millon , T. H. (1981). D iso r d er s o f p er so n a lity . New io r a l M ed icin e, 2 3 , 1 2 0 - 1 2 4 .
York: Wiley. Rogers, T. B., Kuiper, N. A., & Kirker, W. S. (1977).
M ittal, B., & Balasubramanian, S. K. (1987). Testing Self-reference and the encoding of personal infor
the dimensionality of the Self-Consciousness Scales. mation. J o u r n a l o f P erson ality a n d S o c ia l P sy c h o l
J o u r n a l o f P erso n a lity A ssessm en t, 51, 5 3 - 6 8 . og y , 3 5 , 6 7 7 - 6 8 8 .
Mueller, J. FL (1982). Self-awareness and access to Sarason, I. G. (1972). Experimental approaches to test
material rated as self-descriptive or non-descriptive. anxiety: Attention and the uses of information. In
B u lletin o f th e P sy c h o n o m ic S o ciety , 19, 3 2 3 - 3 2 6 . C. D. Spielberger (Ed.), A n x iety : C u rren t tren d s in
Musson, R. E., & Alloy, L. B. (1988). Depression and t h e o r y a n d re se a r ch (Vol. 2 , pp. 3 8 1 - 4 0 3 ) . Orlando,
self-directed attention. In L. B. Alloy (Ed.), C o g n i FL: Academic Press.
tiv e p r o c e s s e s in d e p r e s s io n . New York: Guilford Scheier, M. F. (1976). Self-awareness, self-
Press. consciousness, and angry aggression. J o u r n a l o f
Nasby, W. (1985). Private self-consciousness, articula P erson ality , 4 4 , 6 2 7 - 6 4 4 .
tion of the self-schema, and recognition memory of Scheier, M. F. (1980). Effects of public and private self-
trait adjectives. J o u r n a l o f P erso n a lity a n d S o c ia l consciousness on the public expression of personal
P sy ch olog y , 4 9 , 7 0 4 —709. beliefs. J o u r n a l o f P erso n a lity a n d S o c ia l P sy c h o l
Nasby, W. (1989a). Private and public self-consciousness og y , 3 9, 5 1 4 - 5 2 1 .
and articulation of the self-schema. J o u r n a l o f P er Scheier, M . F., Buss, A. H ., & Buss, D. M . (1978). Self-
so n a lity a n d S o c ia l P sy ch olog y , 5 6 , 1 1 7 - 1 2 3 . consciousness, self-report of aggressiveness, and
Nasby, W. (1989b). Private self-consciousness, self- aggression. J o u r n a l o f R e se a rc h in P erson ality , 12,
awareness, and the reliability of self-reports. J o u r n a l 133-140.
o f P erso n a lity a n d S o c ia l P sy ch o lo g y , 5 6 , 9 5 0 - 9 5 7 . Scheier, M. F., & Carver, C. S. (19 77 ). Self-focused at
Nisbett, R. E., & Wilson, T. D. (1977). Telling more tention and the experience of emotion: Attraction,
than we can know: Verbal reports on mental pro repulsion, elation, and depression. J o u r n a l o f P er
cesses. P sy c h o lo g ic a l R e v iew , 8 4, 2 3 1 - 2 5 9 . so n a lity a n d S o c ia l P sy ch olog y , 3 5 , 6 2 5 - 6 3 6 .
Nolen-Hoeksema, S. (200 4). Response styles theory. Scheier, M. F., & Carver, C. S. (1980). Private and
In C. Papageorgiou & A. Wells (Eds.), D e p r e s public self-attention, resistance to change, and dis
siv e ru m in a tio n : N atu re, th eo r y , a n d trea tm en t sonance reduction. J o u r n a l o f P erso n a lity a n d So-
(pp. 1 0 7 - 1 2 4 ). New York: Wiley. c ia l P sy ch olog y , 3 9 , 3 9 0 - 4 0 5 .
Nystedt, L., & Smari, J. (1989). Assessment of the Scheier, M . F., Carver, C. S., & Gibbons, R. X . (1979).
Fenigstein, Scheier, and Buss Self-Consciousness Self-directed attention, awareness of bodily states,
Scale: A Swedish translation. J o u r n a l o f P erson ality and suggestibility. J o u r n a l o f P erson ality a n d S o c ia l
A ssessm en t, 5 3 , 3 4 2 - 3 5 2 . P sy ch olog y , 37, 1 5 7 6 - 1 5 8 8 .
Piliavin, J. A., & Charng, H. (1988). What is the Scheier, M. F., Fenigstein, A., & Buss, A. H. (1974).
factorial structure of the private and public self- Self-awareness and physical aggression. J o u r n a l o f
consciousness scales? P erson ality a n d S o c ia l P sy E x p e r im e n ta l S o c ia l P sy ch o lo g y , 10, 2 6 5 - 2 7 3 .
c h o lo g y B u lletin , 14, 5 8 7 - 5 9 5 . Schlenker, B. R. (1980). Im p res sio n m a n a g e m e n t: T h e
Pilkonis, P. A. (1977). Shyness, public and private, and s e lf-c o n c e p t, s o c ia l id en tity , a n d in te r p e r s o n a l r e la
its relationship to other measures of social behavior. tion s. Monterey, CA: Brooks/Cole.
J o u r n a l o f P erson ality , 4 5 , 5 8 5 - 5 9 5 . Schlenker, B. R., & Weigold, M. F. (1990). Self-
Porterfield, A. L., Mayer, F. S., Dougherty, K. G., consciousness and self-presentation: Being autono
Kredich, K. E., Kronberg, M. M ., Marsee, K. M ., et mous versus appearing autonomous. J o u r n a l o f P er
al. (1988). Private self-consciousness, canned laugh so n a lity a n d S o c ia l P sy ch o lo g y , 5 9, 8 2 0 - 8 2 8 .
ter, and responses to humorous stimuli. J o u r n a l o f Shapiro, D. (1965). N eu r o tic styles. New York: Basic
R e se a rc h in P erson ality , 2 2 , 4 0 9 - 4 2 3 . Books.
P y szcz y nski,T .,& Greenberg, J . (1987). Self-regulatory Shepherd, J . A., & Arkin, R. M . (1989). Determinants
perseveration and the depressive self-focusing style: of self-handicapping: The moderating role of public
A self-awareness theory of the development and self-consciousness and task importance. P erson ality
maintenance of depression. P s y c h o lo g ic a l B u lletin , a n d S o c ia l P sy c h o lo g y B u lletin , 15, 2 5 2 - 2 6 5 .
102, 122-138. Shrauger, J . S., & Osberg, T. M. (1981). The relative
Pyszczynski, T., Greenberg, J ., Hamilton, J . C., & Nix, accuracy of self-predictions and judgments of oth
J . (1991). On the relationship between self-focused ers in psychological assessment. P sy c h o lo g ic a l B u l
attention and psychological disorder: A critical re letin , 9 0 , 3 2 2 - 3 5 1 .
appraisal. P s y c h o lo g ic a l B u lletin , 110, 5 3 8 - 5 4 3 . Silvia, P. J ., & Gendolla, G. H. (2001). On introspec
Pyszczynski, T., Hamilton, J . C., Herring, F. H., & tion and self-perception: Does self-focused attention
Greenberg, J . (1989). Depression, self-focused at enable accurate self-knowledge? R e v ie w o f G e n e ra l
tention, and negative memory bias. J o u r n a l o f P er P sy ch olog y , 5 , 2 4 1 - 2 6 9 .
so n a lity a n d S o c ia l P sy ch olog y , 57, 3 5 1 - 3 5 7 . Singer, J. L. (1988). Sampling ongoing consciousness
Pyszczynski, T., Holt, K., & Greenberg, J . (1987). D e and emotional experience: Implications for health.
pression, self-focused attention, and expectancies In M. J. Horowitz (Ed.), P sy c h o d y n a m ics a n d c o g
for future positive and negative events for self and n ition (pp. 2 9 7 - 3 4 6 ) . Chicago: University of Chi
others. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o l cago Press.
og y , 5 2 , 9 9 4 - 1 0 0 1 . Smith, ). D., & Shaffer, D. R. (1986). Self-consciousness,
34. P riv a te and P u b lic S elf-C onsciousn ess 511
self-reported altruism, and helping behavior. S o c ia l Underwood, B., 8c Mo ore, B. S. (1981). Sources of
B e h a v io r a n d P erson ality , 14, 2 1 5 - 2 2 0 . behavioral consistency. J o u r n a l o f P erson ality a n d
Smith, T. W., Sc Greenberg, J. (1981). Depression and S o c ia l P sy ch olog y , 4 0 , 7 8 1 - 7 8 5 .
self-focused attention. M o tiv a tio n a n d E m o tio n , 5, Vleeming, R. G., 8c Engels, J. A. (1981). Assessment
323-331. of private and public self-consciousness: A Dutch
Solomon, M . R., & Schopler, J. (1982). Self- replication. J o u r n a l o f P erson ality A ssessm en t, 45,
consciousness and clothing. P erson ality a n d S o c ia l 385-389.
P sy c h o lo g y B u lletin , 8, 5 0 8 - 5 1 4 . Vorauer, J. D., 8c Ross, M . (1999). Self-awareness and
Strack, S., Blaney, P. H., Ganellen, R. J . , 8c Coyne, J. feeling transparent: Failing to suppress one’s self.
C. (1985). Pessimistic self-preoccupation, perfor J o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch olog y , .55,
mance deficits, and depression. J o u r n a l o f P ers o n a l 415-440.
ity a n d S o c ia l P sy ch olog y , 4 9, 1 0 7 6 - 1 0 8 5 . Wells, A., Si Matthews, G. (1994). A tten tio n a n d
Striegel-Moore, R. H., Silberstein, L. R., 8c Rodin, e m o tio n : A clin ica l p er sp e c tiv e . Hillsdale, NJ: E r
J. (1993). T he social self in bulimia nervosa: Pub lbaum.
lic self-conscious, social anxiety, and perceived Wheeler, S. C., Morrison, K. R., DeMarree, K. G., 8c
fraudulence. jo u r n a l o f A b n o r m a l P sy ch olog y , 1 0 2 , Petty, R. E. (2 008). Does private self-consciousness
297-303. increase or decrease priming effects?: It depends.
Swanson, D. W., Bohnert, P. J ., 8c Smith, J . (1970). T h e J o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch olog y , 4 4 ,
p a r a n o id . Boston: Little, Brown. 882-889.
Tobey, E. L., & Tunnell, G. (1981). Predicting our Wicklund, R. A., Sc Duval, S. (1971). Opinion change
impressions on others: Effects of public self- and performance facilitation as a result of objective
consciousness and acting, a self-monitoring sub self-awareness. J o u r n a l o f E x p e r im e n ta l S o c ia l P sy
scale. P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , c h o lo g y , 7, 3 1 9 - 3 4 2 .
7, 6 6 1 - 6 6 9 . Wicklund, R. A., 8c Gollwitzer, P. M . (1987). The
Tooby, J ., Sc Cosmides, L. (1990). On the universality fallacy of the private-public self-focus distinction.
of human nature and the uniqueness of the individ J o u r n a l o f P erson ality , 55, 4 9 2 - 5 2 3 .
ual: The role of genetics and adaptation. J o u r n a l o f Wilson, T. D. (1990). Self-persuasion via self-reflection.
P erson ality , 5 8 , 17-67. In J. Olson & M. P. Zanna (Eds.), S elf-in fe re n ce
Trapnell, P. D., 8c Campbell, J. D. (1999). Private self- p r o c e s s e s : T h e O n ta rio S y m p osiu m (Vol. 6, pp. 4 3 -
consciousness and the five-factor model of person 67). Hillsdale, NJ: Erlbaum.
ality: Distinguishing rumination from reflection. Wilson, T. D., Dunn, D. S., Kraft, D., 8c Lisle, D.
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 76, J. (1989). Introspection, attitude change, and
284-304. attitude-behavior consistency: The disruptive ef
Turner, R. G. (1978a). Effects of differentia] request pro fects of explaining why we feel the way we do. In L.
cedures and self-consciousness on trait attributions. Berkowitz (Hd.), A d v a n ces in e x p e r im e n ta l s o c ia l
J o u r n a l o f R esea rc h in P erson ality , 12, 4 3 1 - 4 3 8 . p s y c h o lo g y (Vol. 19, pp. 1 2 3 - 2 0 5 ) . Orlando, FL:
Turner, R. G. (1978b). Self-consciousness and speed Academic Press.
of processing self-relevant information. P erson ality Wine, J . D. (1971). Test anxiety and direction of atten
a n d S o c ia l P sy c h o lo g y B u lletin , 4, 4 5 6 - 4 6 0 . tion. P s y c h o lo g ic a l B u lletin , 76, 9 2 - 1 0 4 .
Turner, R. G. (1980). Self-consciousness and memory Wine, J. D. (1982). Evaluation anxiety: A cognitive-
of trait terms. P erso n a lity a n d S o c ia l P sy ch olog y attentional construct. In H. W. Krohne 8c L. C.
B u lletin , 6, 2 7 3 - 2 7 7 . I.aux (Eds.), A ch iev e m en t, stress, a n d a n x iety
Turner, R. G., Gilliland, L., 8C Klein, H. M. (1981). (pp. 2 0 7 - 2 2 2 ) . Washington, DC: Hemisphere.
Self-consciousness, evaluation of physical charac W ood, J . V., Saltzberg, J . A., 8c Goldsamt, L. A.
teristics, and physical attractiveness. J o u r n a l o f R e (1990). Does affect induce self-focused attention?
s e a r c h in P erson ality , 15, 1 8 2 - 1 9 0 . J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 5 8,
Turner, R. G., Scheier, M. F., Carver, C. S., 8c Ickes, 899-908.
W. (1978). Correlates of self-consciousness. Jo u r n a l Wood, J. V., Saltzberg, J. A., Neale, J . M., Stone, A. A.,
o f P erson ality A ssessm en t, 4 2 , 2 8 5 - 2 8 9 . Sc Rachmiel, T. B. (1990). Self-focused attention,
Tversky, A., Sc Kahneman, D. (1973). Availability: coping responses, and distressed mood in everyday
A heuristic for judging frequency and probability. life. J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y ,
C o g n itiv e P sy ch olog y , 5, 2 0 7 - 2 3 2 . 58, 1027-1036.
CHAPTER 35
Independent, Relational,
and Collective—Interdependent Self-Construals
Su sa n E. C r o ss
E r in E . H a r d in
B e r n a G e r c e k S w in g
512
35. S elf-C o n stru als 513
behavioral consistency) than by the broader More recently, Harb and Smith (2008)
dimensions. Thus there is a clear need for a created a Six-Fold Self-Construal Scale that
psychometrically sound measure of multidi integrates Brewer’s work on the personal, re
mensional self-construal. lational, and group selves (Brewer & Chen,
2007; Brewer & Gardner, 1996) with Sin-
gelis’s work on vertical and horizontal indi
Measuring R elational Self-C onstrual
vidualism and collectivism (IN D -C O L ; Sin-
Cross and her colleagues (Cross et al., 2 0 0 0 ) gelis, Triandis, Bhawuk, & Gelfand, 1995).
created the Relational-Interdependent Self- The scale assesses the horizontal and vertical
Construal Scale (RISC) to measure the rela collective self-construal, as well as the hori
tional form of interdependent self-construal. zontal and vertical relational self-construal,
Its explicit focus on the individual’s self the personal self-construal, and a humanity-
definition distinguishes it from other mea bou n d self-construal. In samples of college
sures of communion or expressivity. It has students from the United Kingdom, Leba
good internal reliability (> .85) and good non, Syria, and Jordan, reliabilities on all
stability over a 2-month period (test-retest six dimensions were adequate to good, with
reliability is .76; Cross et al., 20 0 0 ). Women most reliabilities in the m id-.80s. In light of
usually score higher than men (ds range Brewer and Chen’s (2007) call for research
from -.1 7 to -.5 7 ). Scores correlate moder ers to distinguish between the relational and
ately positively with other measures of relat collective (or group) interdependence, new
edness but do not correlate with measures measures such as this are likely to be of in
of independence. Examination of the incre creasing utility.
mental utility of the RISC scale showed that
it taps self-definition in a unique fashion that
Betw een- and W ithin-G roups Differences
is not tapped by other measures of related
in Self-C onstrual
ness, expressivity, or communalism (Cross
et al., 2000). These measures have facilitated the explo
sion in research on self-construal by al
lowing researchers explicitly to measure
Measuring Collective
self-construals and to test their relations to
or Group Self-C onstrual
a range of other cognitive, affective, and be
Whereas the relational-interdependent self havioral variables (see subsequent sections).
is defined in terms of significant dyadic re The development of these measures has also
lationships, the collective-interdependent allowed researchers to explore between- and
self is defined in terms of significant group within-groups differences in self-construal.
memberships. Brewer and Chen (2007) dem To the surprise of many, however, the results
onstrated that although most existing mea from such research tend to be inconsistent,
sures of collectivism and interdependence often showing that individuals from differ
assess both types of interdependence, rela- ent countries do not demonstrate the expect
tionally oriented items are more than twice ed differences in self-construal (see Oyser-
as common. Brewer and Chen argued that man, Coon, &C Kemmelmeier, 2002).
cross-cultural researchers need to clearly de These unexpected or absent between-
lineate between these two types of interde groups differences led Matsumoto (1999) to
pendence, as they have very different impli conclude that the theory of self-construal is
cations and predict different outcomes. For fundamentally flawed, whereas Levine and
example, whereas men tend to score lower colleagues (2003) concluded that measures
than women on measures of interdepen of self-construal are flawed. Other authors,
dence (see Cross & Madson, 1997), Gabriel however, have argued convincingly that
and Gardner (1999) demonstrated that dis contextual factors may explain the mixed
tinguishing between relational and collec results. Across several studies, Heine, Leh
tive interdependence yields more nuanced man, Peng, and Greenholtz (2002) obtained
findings: Men score lower than women in expected, unexpected, or absent between-
terms of relational interdependence but groups differences by manipulating the ref
higher than women in terms of collective in erence group that participants had in mind
terdependence. when responding to Singelis’s (1994) SCS.
35. S elf-C o n stru als 515
They also showed that these reference- (how it would increase the ruler’s status; the
group effects are attenuated or absent when IndSC prime) or on collective concerns (the
self-construal scale scores are compared general was a member of the ruler’s family;
for different ethnic groups within the same the InterSC prime; Trafimow, Triandis, &
country, allowing expected between-groups Goto, 1991). As expected, participants ex
differences to emerge in such samples (cf. posed to the IndSC primes described them
Levine et al., 2003). Thus reference-group selves on the T S T using more individual,
effects seem able to account for much of the personal terms than did those exposed to
mixed data on between- and within-groups the InterSC primes. Participants exposed to
differences in self-construal, demonstrat the InterSC primes reported more collective
ing that the apparent problems identified by and group-oriented responses than did those
Matsumoto have more to do with measure exposed to the IndSC primes.
ment issues than with theoretical flaws. Brewer and Gardner (1996) and Gardner,
Although demonstrating between- and Gabriel, and Lee (1999) introduced a second
within-groups differences in levels of self- manipulation of IndSC and InterSC. In this
construal may be important in supporting technique, participants read a story about
the theoretical link between culture and going on a trip, and they circled either sin
self-construal, the purpose of self-construal gular pronouns (I, me, m ine ; IndSC prime)
theory is not solely to explain cultural dif or plural pronouns [we, our, us; InterSC
ferences in cognition, emotion, motiva prime). Control conditions included a task
tion, and behavior. Even if cultures are in which third-person pronouns {they, them)
becoming more similar and do not reliably or impersonal pronouns (it) were circled.
differ in related constructs such as indi Brewer and Gardner hypothesized that when
vidualism and collectivism (cf. Matsumoto, the InterSC (we-us) is primed, others are in
1999), self-construal theory’s greatest con cluded in the self, resulting in an increased
tribution remains: the identification of inde perception of similarity to others (relative to
pendent, relational, and interdependent self the ffcey-primed control group). Using a re
systems as individual-differences variables action-time task, Brewer and Gardner found
that predict other psychological phenom that ti'e-primed participants made judg
ena in reliable and theoretically consistent ments of similarity more quickly than those
ways. in the ffcey-primed condition, but those in
the they- primed condition made judgments
of dissimilarity more quickly than those in
M anipulations o f S elf-C onstrual
the we-primed condition. Later, Gardner
The development of priming manipulations and colleagues found that InterSC-primed
allowed researchers to move from reliance participants endorsed collectivist values
on self-report or proxy measures of self- and obligations to help more than IndSC-
construal to experimental manipulations primed participants. These differences were
of these constructs. As a result, researchers mediated by differences in T ST responses:
can more confidently examine causal hy InterSC-primed participants tended to de
potheses and within-culture consequences scribe themselves in terms of their relation
of activation of the three components of ships and group memberships more than
self-construal. The premise in this work is did the IndSC-primed participants, and self
that all persons, no matter their cultural descriptions in turn predicted responses to
background, construct independent, rela the values and helping measures.
tional, and collective-interdependent self- Other priming tasks have also been used
construals. Cultural practices and affor- successfully in other studies. In their ear
dances, however, result in variability in the liest paper, Trafimow and his colleagues
elaboration and accessibility of these dimen (1991) asked participants to write about
sions. what made them similar to their friends and
Two approaches to manipulating self- family (priming the InterSC) or different
construal have dominated research. In from their friends and family (IndSC). Sta-
the first of these, participants read a story pel and Koomen (2001) primed IndSC and
about a ruler selecting a general to send to InterSC by having participants write a story
war based on either individualistic concerns about themselves, describing either “who I
516 V I. S E L F -R E L A T E D D IS P O S IT IO N S
am,” using the words 1, me, myself, mine relative to a no-prime condition than prim
in each sentence, or “who we are,” using ing the culturally dominant self-construal.
the words we, our, ourselves, ours.” Final Finally, an approach developed by Hong,
ly, Kiihnen and Hannover (2000) created a M orris, Chiu, and Benet-Martmez (2000)
scrambled-sentence task in which the four- exposes bicultural Asians to cultural icons
word sentences either focused on IndSC (“I from the East (e.g., the Great Wall of China)
like being unique”) or on InterSC (”I sup or from the West (e.g., Mickey Mouse) to
port my team”). Each scrambled sentence prime cultural knowledge systems. This ap
also included one word that was related to proach likely primes a variety of culture
type of prime— either a word reflecting in relevant thoughts, attitudes, beliefs, and
dependence (assertive) or a word reflecting goals— some of which will be related to self-
interdependence (help). The authors varied construals— but it does not specifically tar
the salience of the IndSC or InterSC prime get self-construal.
by either having participants write down the
unscrambled sentence (the “overt” prime) or
having the participants write down the un H ow D o Self-Construals
necessary word (the “subtle” prime). When Shape Behavior?
participants were primed with the overt
IndSC prime, they paradoxically viewed Self-C onstrual Influences Cognition
themselves as more similar to the target and Inform ation Processing
person than when primed with the Inter Thinking about the S elf
SC prime. But when exposed to the subtle
primes, the InterSC primes led to greater Markus and Kitayama (1991) argued that
perceived similarity than the IndSC prime. individuals with InterSC should be espe
cially likely to pay attention to others and
the social context of interaction, resulting in
Concerns about Priming Manipulations self-representations that include social con
Several questions remain about these prim texts and elaborate cognitive representations
ing manipulations. First, it is likely that the of others. Consistent with this hypothesis,
two most common InterSC manipulations— people from collectivist cultures (who are
the story of the ruler sending the general presumed to have high InterSC) report more
who is a member of his family into battle social, collective, or group-oriented respons
(Trafimow et al., 1991) and the story of a es on the T S T than do people from individu
trip to the city using plural pronouns (Brew alistic cultures (e.g., Kanagawa, Cross, 8c
er & Gardner, 1996)— make the relational M arkus, 2001). Few studies, however, have
self-construal accessible, rather than the col examined the association between measures
lective self. Brewer and Gardner (1996) at of self-construal and T S T responses, and
tempted to tease apart these two dimensions their results have not always shown the ex
of the InterSC, but they were only partly pected relations (see Bresnahan et al., 2 0 0 5 ;
successful. Grace & Cramer, 2003). This discrepancy
Second, researchers have seldom used may be due in part to lack of agreement on
priming manipulations with non-Western a standardized coding scheme for T ST re
populations. A few studies have included sponses.
either Asian Americans or Asian students,
and their findings are somewhat mixed.
Context-Sensitive S elf
Trafimow and colleagues (1991) included
a small group of Asian Americans, and the If persons with high InterSC are sensitive
pattern of individual and group cognitions to situational or relational context, then
were similar in the two priming conditions they should tend to describe themselves
to those of the European American partici differently in different situations. This hy
pants. Gardner and her colleagues (1999) pothesis has been supported in studies that
found that priming the culturally nondomi use culture as a proxy for self-construal
nant dimension of self-construal (e.g., Inter (e.g., Suh, 2 0 0 2 ), but the results of studies
SC for European Americans and IndSC for using measures or manipulations are mixed
East Asians) resulted in greater differences (e.g., Cross, Gore, & M orris, 2003). Kashi-
35. S elf-C o n stru als 517
ma and his colleagues (2004) found that They found more contextual interference ef
RelSC was associated with contextualized fects for participants high in InterSC than for
self-description for Japanese participants participants high in IndSC. In addition, in a
but not among members of Western soci paradigm that involved switching frequently
eties (Australia, the United Kingdom, and between two different cognitive tasks, par
Germany) and was negatively related to ticipants high in InterSC were less facile in
context-sensitive self-descriptions among their switches than were the participants
Korean participants. This inconsistency high in IndSC. This suggests that IndSC al
may be due in part to methods that make it lows the person to quickly and easily focus
more likely that Western participants will on a specific task and to inhibit attention to
feel a press to describe themselves consis previous or irrelevant tasks.
tently across situations (e.g., a one-time
questionnaire).
Relational Self-Construal
and Information Processing
Context-Dependent Cognition
Cross, M orris, and Gore (2002) have argued
Researchers have extended research on self- that the RelSC should influence informa
construal and context-dependent cognition tion processing without conscious control,
beyond self-description. In early work, peo and they examined the role of RelSC in a
ple with high InterSC tended to report great variety of implicit cognitive processes that
er attention to the context described in social centered on relationship-oriented material.
scenarios and were more likely to attribute They found that North American partici
outcomes in the scenarios to contextual ef pants with chronically high RelSC respond
fects than were others (Singelis, 1994). In an ed more positively to relationship-oriented
experimental study with German students, terms in an Implicit Association Test (IAT)
InterSC-primed participants were more sen task (Greenwald, McGhee, & Schwartz,
sitive to context effects on questions in a 1998) and had denser associative networks
questionnaire than were IndSC-primed par for relationship-oriented terms than did
ticipants (Haberstroh, Oyserman, Schwarz, those low in RelSC. Cross and her colleagues
Kiihnen, & Ji, 2002). also found that participants high in RelSC
Kiihnen, Hannover, and Schubert (2001) remembered more relationship-related infor
found that priming the IndSC (using mation about a target person and organized
sentence-completion and other tasks) result information about others in terms of their
ed in context-independent processing and relationships. In short, persons with chroni
that priming the InterSC resulted in context- cally high RelSC are “tuned” to pay atten
dependent processing. For example, when the tion to and to organize their worlds in terms
IndSC was primed, German and American of relationships.
participants were quicker to find geometric
figures embedded within more complex geo
Memory
metric designs. In contrast, InterSC-primed
participants performed better than IndSC- A focus on the IndSC or InterSC also influ
primed participants on a task that was espe ences what one remembers. In one study,
cially sensitive to context-dependent think both European American and Asian or Asian
ing. In another study, InterSC-primed North American participants were primed with
American participants were more sensitive IndSC or InterSC and asked to recall their
to contextual information presented in a earliest memories (Wang & Ross, 2005).
causal reasoning induction task than were The IndSC-primed participants tended to
IndSC-primed participants (Kim, Grimm, describe more individual-focused memories,
& M arkm an, 2007). whereas the InterSC-primed participants
Hannover, Pohlmann, Springer, and Roed- tended to describe more group-focused
er (2005) have examined additional cogni memories and memories that focused on
tive consequences of priming self-construals. social interaction. In a follow-up study
In one series of studies, they used a modified in which European American and Asian
Stroop task to examine the association be American participants read a child’s picture
tween self-construal and attentional focus. book about a bear going to market, InterSC-
518 V I. S E L F -R E L A T E D D IS P O S IT IO N S
primed participants again were more likely more likely than IndSC-primed participants
than IndSC-primed participants to remem to take a target person’s prior knowledge
ber details about social interactions in the into account.
story. In other research, InterSC-primed
participants had better memory for inciden
S elf-C onstrual Influences A ffect
tal contextual information when asked to
remember the location of items in an array Despite interest in cultural differences in
than did IndSC-primed participants (Kiih- emotion, little research has actually mea
nen & Oyserman, 2002). sured self-construals to investigate how they
relate to emotion. The few studies that have
tend to find that IndSC is associated with
Contrast versus Assimilation
lower levels of depression (e.g., Lam, 2 0 0 5 ;
Self-construal also influences information- Okazaki, 1997; Sato & M cCann, 1998), un
processing styles. For example, priming the happiness (Kim, Kasser, & Lee, 2003), gen
IndSC results in contrast or differentiation eral anxiety (e.g., ITardin, Varghese, Tran,
effects (distinguishing oneself from oth & Carlson, 2 0 0 6 ; Kim et al., 2003), and so
ers), and priming the InterSC results in as cial anxiety (Hardin et al., 2 0 0 6 ; Okazaki,
similation or integration effects (connect 1997), whereas InterSC is often associated
ing oneself to others; Stapel & Koomen, with higher levels of these negative affects
2001). Furthermore, chronic differences in (Hardin et al., 2 0 0 6 ; Okazaki, 1997; Sato &
the activation of these self-construals influ M cCann, 1998).
ence perceived similarity with others (Cross Such results raise the question of why
et al., 2002). North American participants these relations exist. The relation of self-
rated the descriptiveness of multiple traits, construal to social anxiety is intuitive; it
values, and abilities for themselves and a is not surprising that individuals high in
same-sex friend. Similarity scores were cal InterSC, for whom interpersonal relation
culated by computing intraclass correlations ships are central, would express more con
for each person’s pair of ratings in each do cern about appropriate behavior in social
main (traits, values, and abilities). In regres contexts. Interestingly, IndSC is often found
sion analyses that controlled for self-esteem, to be a better predictor of social anxiety
RelSC scores significantly predicted each than InterSC. “In other words, those who
type of similarity. were more concerned with asserting one’s
own judgment and emphasizing autonomy
from others were less likely to be socially
Perspective Taking
avoidant, distressed in social situations,
If others are connected to the self and and fearful of social evaluations” (Okazaki,
viewed as self-defining, then a person will 1997, p. 58). But what of the relation of self-
tend to take the other’s perspective in social construal to other types of negative affect,
interaction and decision making. Cross and such as unhappiness and depression? These
colleagues (2000) found that students with relations are likely artifactual, explained by
high RelSC were more likely to consider the the high correlations between social anxiety
needs and wishes of friends and family mem and other types of negative affect. Okazaki
bers when making decisions than were stu (1997) demonstrated that once social anxi
dents with low RelSC. Likewise, Gore and ety is controlled, neither IndSC nor InterSC
Cross (2006) found that North Americans is related to depression.
with high RelSC were more likely to include Such failure to control for social anxiety
other people in their rationale for pursu may account for other findings that, on the
ing important goals. People whose InterSC surface, appear to contradict predictions of
is chronically activated (or primed) tend to self-construal theory. For example, given
give more weight to others’ views or opin that InterSC is consistent with collectivist
ions about their goals and behaviors than values, we might expect interdependence
do people with high IndSC (e.g., Ybarra & to be associated with positive outcomes in
Trafimow, 1998). In priming studies, Hab- collectivist cultures. Some evidence does
erstroh and colleagues (2002) showed that support this hypothesis: InterSC was asso
InterSC-primed Western participants were ciated with greater life satisfaction in Hong
35. S elf-C o n stru als 519
Kong but was unrelated to life satisfaction Finally, Lam (2005) argued that family
in the United States (Kwan, Bond, & Singe- cohesion should be more important to the
lis, 1997). Surprisingly, then, greater relative self-esteem of interdependent adolescents,
InterSC predicted greater unhappiness and whereas peer support should be more impor
less happiness in South Koreans but was un tant to the self-esteem of independent ado
related to unhappiness in the United States lescents. Indeed, family cohesion fully medi
(Kim et al., 2003). If the measures of happi ated the relation between InterSC and global
ness and unhappiness used in this study are self-esteem, whereas peer support partially
correlated with social anxiety, the finding mediated the relation between IndSC and
that interdependence predicts unhappiness self-esteem.
in South Korea may be an artifact of inter
dependence predicting social anxiety. These
Relational Self-Construal and Affect
results, however, must be interpreted with
caution, as they are based on unidimension Although RelSC is not associated with gener
al self-construal scores. al psychological well-being among predomi
Other researchers have explored media nantly European American college students,
tors of the relation between self-construal it is associated with greater relational well
and affect. Given the importance of inter being (Cross et al., 2003). RelSC is related to
personal relationships for interdependent general well-being in other samples. Berkel
people and the importance of internal, pri and Constantine (2005) argued that a need
vate self-evaluations for independent people, for affiliation may be stronger and more ben
Kwan and colleagues (1997) argued— and eficial to women of color in predominantly
demonstrated— that relationship harmony white environments and hypothesized that
fully mediates the relation between InterSC RelSC would predict life satisfaction among
and life satisfaction, whereas global self these women of color. Indeed, in their sample
esteem fully mediates the relation between of African American and Asian American
IndSC and life satisfaction in samples from women recruited from a predominantly white
Hong Kong and the United States. university, greater RelSC predicted greater
Using the same measures of global self life satisfaction, even after controlling for re
esteem and self-construal in a sample of lationship harmony and family conflict.
Vietnamese American adolescents, Lam
(2005) also found that self-esteem fully me
Self-C onstrual Influences Motivation
diates the relation between IndSC and de
and Self-R egulation
pression. Contrary to the results of Kwan
and colleagues (1997), however, self-esteem Markus and Kitayama’s (1991) conceptu
also fully mediated the relation between In alization has several implications for the
terSC and depression. As in other samples, manifestation of motivations. First, for the
IndSC in Lam ’s sample was associated with InterSC, being a part of social groups and
greater self-esteem (Kwan et al., 1997) and maintaining harmonious relationships with
less depression (Okazaki, 1997; Sato & important others are of great importance, in
M cCann, 1998); contrary to past research, contrast to the autonomy-related needs of the
however, InterSC was also associated with IndSC. To date, however, only a few studies
greater self-esteem and less depression. This have used measures or manipulations of self-
is likely due to the more bicultural nature construal to examine motivational process
of Lam’s adolescent sample, which, in con es. One such study found that cultural ori
trast to other studies that have used col entation of individualism-collectivism and
lege students (Kwan et al., 1997; Okazaki, self-construal were both related to sensitiv
1997; Sato & M cC ann, 1998), consisted ity to a partner’s concerns for saving face in
of adolescents living in ethnically encultur- conflict situations (Oetzel &C Ting-Toomey,
ated families. Thus Lam argued that high 2003). Moreover, InterSC was positively re
er interdependence reflected an important lated to avoiding confrontation in a conflict
cultural consistency with the participants’ and engaging in integrative behavior. IndSC,
home environments, which in turn was as on the other hand, was related to dominat
sociated with greater self-esteem and less ing behavior in such a situation. These re
depression. sults imply that people with high InterSC
520 V I. S E L F -R E L A T E D D IS P O S IT IO N S
variables and life satisfaction in African American Grace, S. L., 8c Cramer, K. L. (2003). The elusive na
and Asian American college women. J o u r n a l o f ture of self-measurement: The self-construal scale
C o lle g e C o u n selin g , 8, 5 - 1 3 . versus the twenty statements test. J o u r n a l o f S o c ia l
Bresnahan, M . J ., Levine, T. R ., Shearman, S. M., P sy ch olog y , 143, 6 4 9 - 6 6 8 .
Lee, S. Y., Park, C., 8c Kiyomiya, T. (200 5 ). A Greenwald, A. G., McGhee, D. E., 8c Schwartz, J.
multimethod multitrait validity assessment of self- L. K. (1998). Measuring individual differences in
construal in Japan, Korea, and the United States. implicit cognition: The Implicit Association Test.
H u m an C o m m u n ic a tio n R e se a rc h , 3 1, 3 3 - 5 9 . J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 74,
Brewer, M . B., 8c Chen, Y. ( 2 0 07 ). Where (who) are 1464-1480.
collectives in collectivism?: Toward conceptual clar Gudykunst, W. B., Matsu moto, Y., Ting-Toomey, S.,
ification of individualism and collectivism. P s y c h o Nishida, T., Kim , K., 8c Heyman, H. (1996). The
lo g ic a l R e v iew , 114, 1 3 3 - 1 5 1 . influence of cultural individualism-collectivism,
Brewer, M . B., 8c Gardner, W. (1996). W h o is this self-construals, and individual values on communi
“we” ?: Levels of collective identity and self repre cation styles across cultures. H u m an C o m m u n ic a
sentations. J o u r n a l o f P erso n a lity a n d S o c ia l P sy tion R esea rch , 2 2 , 5 1 0 - 5 4 3 .
c h o lo g y , 71, 8 3 - 9 3 . Haberstroh, S., Oyserman, D., Schwarz, N., Kiihnen,
Brockner, J ., De Cremer, D., van den Bos, K., 8c Chen, U., 8c Ji, L. (200 2 ). Is the interdependent self more
Y. R. (2 005). The influence of interdependent self- sensitive to question context than the independent
construal on procedural fairness effects. O rg a n iz a self?: Self-construal and the observation of con
tio n a l B e h a v io r a n d H u m an D ec is io n P rocesses, versational norms. J o u r n a l o f E x p e r im e n ta l S o c ia l
9 6 , 1 5 5 -16 7. P sy ch olog y , 3 8 , 3 2 3 - 3 2 9 .
Cheng, R. W., 8c Lam, S. (20 07 ). Self-construal and Hannover, B., Pohlmann, C., Springer, A., 8c Roeder,
social comparison effects. B ritish J o u r n a l o f E d u U. (2005). Implications of independent versus inter
c a t io n a l P sy ch olog y , 77, 1 9 7 - 2 1 1 . dependent self-knowledge for motivated social cog
Cross, S. E., Bacon, P. L., 8c Morris, M . L. (2 000). nition: T he semantic procedural interface model of
The relational-interdependent self-construal and the self. S e lf a n d Id en tity , 4, 1 5 9 - 1 7 5 .
relationships. J o u r n a l o f P erso n a lity a n d S o c ia l Harb, C., 8c Smith, P. B. (2 008). Self-construals across
P sy ch olog y , 7 8 , 7 9 1 - 8 0 8 . cultures: Beyond independence-interdependence.
Cross, S. E., Gore, J. S., & Morris, M . L. (2003). J o u r n a l o f C ro ss-C u ltu ra l P sy ch o lo g y , 3 9 , 1 7 8 -
The relational-interdependent self-construal, self- 197.
concept consistency, and well-being. J o u r n a l o f P er Hardin, E. E. (20 0 6). Convergent evidence for the mul
s o n a lity a n d S o c ia l P sy ch o lo g y , 8 5, 9 3 3 - 9 4 4 . tidimensionality of self-construal. J o u r n a l o f C ross-
Cross, S. E., 8c Mad so n, L. (1997). Models of the self: C u ltu ral P sy ch o lo g y , 37, 5 1 6 - 5 2 1 .
Self-construals and gender. P sy c h o lo g ic a l B u lletin , Hardin, E. E., Leong, F. T. L., 8c Bhagwat, A. A.
122, 5-37. (20 0 4). Factor structure of the Self-Construal Scale
Cross, S. E., 8c Morris, M . L. (2003). Getting to know revisited: Implications for the multidimensionality
you: T he relational self-construal, relational cogni of self-construal. J o u r n a l o f C ro ss-C u ltu ra l P sy
tion, and well-being. P erson ality a n d S o c ia l P sy c h o lo g y , 3 5 , 3 2 7 - 3 4 5 .
c h o lo g y B u lletin , 2 9 , 5 1 2 - 5 2 3 . Hardin, E. E., Varghese, F. V., Tran, U. V., 8c Carlson,
Cross, S. E., Morris, M. L., 8c Gore, J. S. (2002). A. Z. ( 2 0 0 6). Anxiety and career exploration: G en
Thinking about oneself and others: T he relational- der differences in the role of self-construal. J o u r n a l
interdependent self-construal and social cognition. o f V o c a tio n a l B eh a v io r, 69, 3 4 6 - 3 5 8 .
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o lo g y , 8 2 , Heine, S. J., 8c Hamamura, T. (200 7 ). In search of East
399-418. Asian self-enhancement. P erso n a lity a n d S o c ia l
Gabriel, S., 8c Gardner, W. L. (1999). Are there “his” P sy ch o lo g y R e v iew , 11, 4 - 2 7 .
and “hers” types of interdependence?: The implica Heine, S. J., Kitayama, S., 8c Lehman, D. R. (2001).
tions of gender differences in collective versus re Cultural differences in self-evaluation: Japanese
lational interdependence for affect, behavior, and readily accept negative self-relevant information.
cognition. J o u r n a l o f P erson ality a n d S o c ia l P sy J o u r n a l o f C ro ss-C u ltu ra l P sy ch olog y , 3 2 , 4 3 4 -
c h o lo g y , 77, 6 4 2 - 6 5 5 . 443.
Gardner, W. L., Gabriel, S., 8c Hochschild, L. (2002). Heine, S. J. , Lehman, D. R ., Mar kus, H. R ., 8C Kitaya
When you and I are “we,” you are not threatening: ma, S. (1999). Is there a universal need for positive
The role of self-expansion in social comparison. self-regard? P s y c h o lo g ic a l R e v iew , 1 0 6 , 7 6 6 - 7 9 4 .
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 8 2 , Heine, S. J ., Lehman, D. R., Peng, K., 8c Greenholtz,
239-251. J . (2002). W h a t ’s wrong with cross-cultural co m
Gardner, W. L., Gabriel, S., 8c Lee, A. Y. (1999). “I ” parisons of subjective Likert scales?: The reference-
value freedom, but “we” value relationships: Self- group effect. J o u r n a l o f P erso n a lity a n d S o c ia l P sy
construal priming mirrors cultural differences in c h o lo g y , 8 2 , 9 0 3 - 9 1 8 .
judgment. P sy c h o lo g ic a l S cien ce, 10, 3 2 1 - 3 2 6 . Higgins, E. T. (1996). The “self-digest”: Self-knowledge
Gore, J. S., 8c Cross, S. E. (20 06 ). Pursuing goals for serving self-regulatory functions. J o u r n a l o f P er
us: Relationally autonomous reasons in long-term so n a lity a n d S o c ia l P sy ch o lo g y , 71, 1 0 6 2 - 1 0 8 3 .
goal pursuit. J o u r n a l o f P erson ality a n d S o c ia l P sy Higgins, E. T. (1997). Beyond pleasure and pain.
c h o lo g y , 9 0 , 8 4 8 - 8 6 1 . A m er ica n P sy ch o lo g ist, 5 2 , 1 2 8 0 - 1 3 0 0 .
Gore, J . S., Cross, S. E., 8c Morris, M . L. (200 6). L et’s Holland, R . W., Roeder, U., van Baaren, R . B., Brandt,
be friends: Relational self-construal and the devel A. C., 8c Hannover, B. (20 04 ). Do n’t stand so close
opment of intimacy. P ers o n a l R e la tio n s h ip s, 13, to me: T he effects of self-construal on interpersonal
83-102. closeness. P sy c h o lo g ic a l S cien ce, 15, 2 3 7 - 2 4 2 .
35. S elf-C o n stru als 525
Hong, Y., Morris, M . W., Chiu, C., & Benet-Martl'nez, Lam, B. T. (200 5 ). Self-construal and depression among
V. ( 2 0 00 ). Multicultural minds: A dynamic con Vietnamese American adolescents. I n te r n a tio n a l
structivist approach to culture and cognition. J o u r n a l o f In tercu ltu ral R e la tio n s , 2 9 , 2 3 9 - 2 5 0 .
A m er ic a n P sy c h o lo g ist, 5 5 , 7 0 9 - 7 2 0 . Lee, A. Y., Aaker, J. L., 8c Gardner, W. L. (2000).
Iyengar, S. S., 8c Lepper, M . R. (1999). Rethinking the The pleasures and pains of distinct self-construals:
value of choice: A cultural perspective on intrinsic The role of interdependence in regulatory focus.
motivation. J o u r n a l o f P erson ality a n d S o c ia l P sy J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 78,
c h o lo g y , 76, 3 4 9 - 3 6 6 . 1122-1134.
Johnson, R. E., Selenta, C., 8c Lord, R. G. (2006). Levine, T. R ., Bresnahan, M . J., Park, H. S., Lapinski,
When organizational justice and the self-concept M . K., Wittenbaum, G. M ., Shearman, S. M ., et al.
meet: Consequences for the organization and its (2003). Self-construal scales lack validity. H u m an
members. O rg a n iz a tio n a l B e h a v io r a n d H u m an C o m m u n ic a tio n R e se a rc h , 2 9, 2 1 0 - 2 5 2 .
D ec is io n P ro c esses, 99, 1 7 5 - 2 0 1 . Markus, H. R., 8c Kitayama, S. (1991). Culture and
Kanagawa, C., Cross, S. E., 8c Mar kus, H. R. (2001). the self: Implications for cognition, emotion, and
“W ho am I ? ”: The cultural psychology of the con motivation. P sy c h o lo g ic a l R e v iew , 9 8 , 2 2 4 - 2 5 3 .
ceptual self. P erso n a lity a n d S o c ia l P sy ch o lo g y B u l Markus, H. R., 8c Kitayama, S. (20 0 4). Models of
letin , 27, 9 0 - 1 0 3 . agency: Sociocultural diversity in the construction
Kashima, Y., Kashima, E., Farsides, T., Kim, U., o f action. In V. Murphy-Berman 8c J. J . Berman
Strack, F., Werth, L., et al. (20 04 ). Culture and (Eds.), N e b r a s k a S y m p osiu m on M o tiv a tio n : Vol.
context-sensitive self: The amount and meaning of 49. C ro ss-c u ltu ra l d iffe r e n c e s in p er sp e c tiv e s on
context sensitivity of phenomenal self differ across th e s e l f (pp. 1 - 5 7 ) . Lincoln: University of Nebraska
cultures. S e lf a n d Id en tity , 3 , 1 2 5 - 1 4 1 . Press.
Kim, K., Grimm, L. R ., & M a rk m a n , A. B. (2007). Matsumoto, D. (1999). Culture and self: An empirical
Self-construal and the processing of covariation in assessment of Markus and Kitayama’s theory of in
formation in causal reasoning. M e m o ry a n d C o g n i dependent and interdependent self-construal. A sian
tion , 3 5 , 1 3 3 7 - 1 3 4 3 . J o u r n a l o f S o c ia l P sy ch olog y , 2 , 2 8 9 - 3 1 0 .
Kim, M ., Aune, K. S., Hunter, J . E., Kim, H., 8c Kim, Oetzel, J . G. (1998). Explaining individual communi
J. (2001). The effect of culture and self-construals cation processes in homogeneous and heterogeneous
on predispositions toward verbal communication. groups through individualism-collectivism and
H u m an C o m m u n ic a tio n R e se a rc h , 27, 3 8 2 - 4 0 8 . self-construal. H u m an C o m m u n ic a tio n R esea rch ,
Kim, M ., Sharkey, W. F., & Singelis, T. M . (1994). The 25, 20 2-224.
relationships between individuals’ self-construals Oetzel, J. G., 8c Ting-Toomey, S. (2 003). Face concerns
and perceived importance of interactive constraints. in interpersonal conflict: A cross-cultural empirical
In te r n a tio n a l J o u r n a l o f In tercu ltu ral R e la tio n s , test of face negation theory. C o m m u n ic a tio n R e
18, 1 1 7 - 1 4 0 . sea rch , 3 0 , 5 9 9 - 6 2 4 .
Kim, Y., Kasser, T., 8c Lee, H. (2003). Self-concept, Okazaki, S. (1 997). Sources of ethnic differences be
aspirations, and well-being in South Korea and the tween Asian American and white American college
United States. J o u r n a l o f S o c ia l P sy ch olog y , 143, students on measures of depression and social a n x i
277-290. ety. J o u r n a l o f A b n o r m a l P sy c h o lo g y , 1 0 6 , 5 2 - 6 0 .
Kuhn, M . H., 8c McPartland, T. (1954). An empirical Oyserman, D., Coon , H. M ., 8c Kemmelmeier, M.
investigation of self-attitudes. A m er ica n S o c io lo g i (2002). Rethinking individualism and collectivism:
c a l R ev iew , 19, 5 8 - 7 6 . Evaluation of theoretical assumptions and meta
Kurman, J. (2001). Self-enhancement: Is it restricted analyses. P sy c h o lo g ic a l B u lletin , 1 2 8 , 3 - 7 2 .
to individualistic cultures? P erson ality a n d S o c ia l Sato, T., 8c M cC ann , D. (1998). Individual differences
P sy c h o lo g y B u lletin , 27, 1 7 0 5 - 1 7 1 6 . in relatedness and individuality: An exploration of
Kiihnen, U., & Hannover, B. ( 2 00 0). Assimilation and two constructs. P erson ality a n d In d iv id u a l D iffe r
contrast in social comparisons as a consequence of e n c e s, 2 4 , 8 4 7 - 8 5 9 .
self-construal activation. E u r o p e a n J o u r n a l o f S o Sedikides, C., &c Brewer, M . B. (Eds.). (2001). In d i
c ia l P sy ch olog y , 3 0 , 7 9 9 - 8 1 1 . v id u a l self, r e la tio n a l self, c o lle c tiv e self. New York:
Kiihnen, U., Hannover, B., 8c Schubert, B. (2001). The Psychology Press.
semantic-procedural interface model of the self: Sedikides, C., Gaertner, L., &c Toguchi, Y. (2003). Pan-
The role of self-knowledge for context-dependent cultural self-enhancement. J o u r n a l o f P erson ality
versus context-independent modes of thinking. a n d S o c ia l P sy ch olog y , 84(1), 6 0 - 7 9 .
J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o lo g y , 8 0, Seeley, E. A., 8c Gardner, W. L. (2003). The “self
397-409. less” and self-regulation: The role of chronic other-
Kiihnen, U., & Oyserman, D. (2002). Thinking about orientation in averting self-regulatory depletion.
the self influences thinking in general: Cognitive S e lf a n d Id en tity , 2 , 103-117.
consequences of salient self-concept. J o u r n a l o f E x Seeley Howard, E., Gardner, W. L., 8c Thompson, L.
p e r im e n t a l S o c ia l P sy c h o lo g y , 3 8 , 4 9 2 - 4 9 9 . (20 0 7). The role of the self-concept and the social
Kwan, V. S. Y., Bond, M . H., 8c Singelis, T. M. (1997). context in determining the behavior of power hold
Pancultural explanations for life satisfaction: Add ers: Self-construal in intergroup versus dyadic dis
ing relationship harmony to self-esteem. J o u r n a l o f pute resolution negotiations. J o u r n a l o f P erson ality
P erso n a lity a n d S o c ia l P sy ch olog y , 73, 1 0 3 8 - 1 0 5 1 . a n d S o c ia l P sy ch olog y , 9 3, 6 1 4 - 6 3 1 .
Lam, A. G., 8c Zane, N. W. S. (200 4). Ethnic differ Sinclair, L., 8c Fehr, B. (2005). Voice vs. loyalty: Self-
ences in coping with interpersonal stressors: A test construals and responses to dissatisfaction in ro
of self-construals as cultural mediators. J o u r n a l o f mantic relationships, j o u r n a l o f E x p e r im e n ta l S o
C ro ss-C u ltu ra l P sy c h o lo g y , 3 5 , 4 4 6 - 4 5 9 . c ia l P sy ch olog y , 41, 2 9 8 - 3 0 4 .
526 V I. S E L F -R E L A T E D D IS P O S IT IO N S
Singelis, T. M . (1994). The measurement of indepen Tesser, A. (1980). Self-esteem maintenance in family
dent and interdependent self-construals. P ers o n a l dynamics. J o u r n a l o f P erso n a lity a n d S o c ia l P sy
ity a n d S o c ia l P sy c h o lo g y B u lletin , 2 0 , 5 8 0 - 5 9 1 . c h o lo g y , 3 9 , 7 7 - 9 1 .
Singelis, T. M ., & Brown, W. J . (1995). Culture, self, Trafimow, D., Triandis, H. C., & Goto, S. G. (1991).
and collectivist communication: Linking culture to Some tests of the distinction between the private self
individual behavior. H u m an C o m m u n ic a tio n R e and the collective self. J o u r n a l o f P erso n a lity a n d
sea rc h , 2 1 , 3 5 4 - 3 8 9 . S o c ia l P sy ch olog y , 6 0 , 6 4 9 - 6 5 5 .
Singelis, T. M ., Triandis, H. C., Bhawuk, D., & Gel- van Baaren, R. B., Maddu x, W. W., Chartrand, T.
fand, M . J. (1995). Horizontal and vertical dimen L ., de Bouter, C., & van Knippenberg, A. (200 3).
sions of individualism and collectivism: A theoreti It takes two to mimic: Behavioral consequences of
cal and measurement refinement. C ro ss-C u ltu ra l self-construals. J o u r n a l o f P erson ality a n d S o c ia l
R e se a rc h : T h e j o u r n a l o f C o m p a ra tiv e S o c ia l S ci P sy ch olog y , 8 4 , 1 0 9 3 - 1 1 0 2 .
en c e , 2 9 , 2 4 0 - 2 7 5 . Wang, Q., & Ross, M . (2 005). W hat we remember and
Somech, A. (20 0 0). The independent and the interde what we tell: The effects of culture and self-priming
pendent selves: Different meanings in different cul on memory representations and narratives. M e m o
tures. I n te r n a tio n a l jo u r n a l o f In tercu ltu ral R e la ry, 13, 5 9 4 - 6 0 6 .
tio n s, 2 4 , 1 6 1 - 1 7 2 . Weisz, J . R ., Rothbaum, F. M ., & Blackburn, T. C.
Stapel, D. A., & Koomen, W. (2001). I, we, and the ef (1984). Standing out and standing in: The psychol
fects of others on me: How self-construal level mod ogy of control in America and Japan. A m er ica n P sy
erates social comparison effects .J o u r n a l o f P ers o n c h o lo g is t, 3 9, 9 5 5 -9 6 9 .
a lity a n d S o c ia l P sy ch olog y , 8 0 , 7 6 6 - 7 8 1 . Yb arra, O., & Trafimow, D. (1998). Ho w priming
Suh, E. M . (2 002). Culture, identity consistency, and the private self or collective self affects the relative
subjective well-being. J o u r n a l o f P erson ality a n d weights of attitudes and subjective norms. P e r s o n a l
S o c ia l P sy c h o lo g y , 8 3 , 1 3 7 8 - 1 3 9 1 . ity a n d S o c ia l P sy ch o lo g y B u lletin , 2 4 , 3 6 2 - 3 7 0 .
CHAPTER 36
Self-Esteem
J e n n if e r K . B o sso n
W il l ia m B . Sw a n n , J r
527
528 V I. S E L F -R E L A T E D D IS P O S IT IO N S
challenges to the self (e.g., American Psy problems but also that it could be accom
chiatric Association, 2 0 0 0 ; M orf & Rhode- plished by merely reciting a few affirmations
walt, 2 0 0 1 ; Paulhus, Robins, Trzesniewski, such as “I am lovable and capable.”
& Tracy, 2 0 0 4 ; Raskin, Novacek, &CHogan, Members of the academic community
1991; Westen, 1990). challenged the extravagant claims of the
By the late 1950s, behaviorism was begin self-esteem movement, noting that they
ning to lose its grip on psychology in Amer lacked a solid basis in reality (e.g., Dawes,
ica. As a result, more and more theorists 1994; Swann, 1996). Some authors recently
began focusing on issues related to the self, took the argument a step further, not only
although most avoided using the language echoing the criticisms of the self-esteem
of self-esteem. In his theory of social com movement but also questioning the viabil
parison, for example, Festinger (1954) pos ity of the self-esteem construct itself. M ost
ited that people learn about their abilities significantly, after reviewing a subset of the
and opinions by comparing themselves with self-esteem literature, Baumeister, Camp
others. Although Festinger did not state that bell, Krueger, and Vohs (2003) asserted that
social comparison could serve as a basis for measures of self-esteem fail to offer strong
self-esteem, such a conclusion is surely com predictions of socially important behaviors,
patible with his formulation. Similarly, al as promised by the California Task Force.
though Bern (1972) refrained from discuss Some have taken this gloomy assessment to
ing self-esteem in his self-perception theory, mean that self-esteem is not a viable con
his notion that people derive self-knowledge struct and that its effects should no longer
from observing their own behavior and the be studied (Scheff & Fearon, 2004).
conditions under which it occurs can be un Others, however, took issue with Baumeis
derstood as a means through which people ter and colleagues’ (2003) conclusions re
develop self-esteem. garding the viability of the self-esteem con
Not long after the introduction of Bern’s struct (e.g., Marsh & Craven, 2 0 0 6 ; Swann,
(1972) theory, there was an explosion of Chang-Schneider, & M cClarty, 2 0 0 7 ,2 0 0 8 ).
interest in the self within social psychol For example, Swann, Chang-Schneider, and
ogy. There were several reasons for this M cClarty (2007) countered Baumeister and
emerging interest, but one factor seems to colleagues’ claims by proposing that, in
have been the success of efforts to draw evaluating the capacity of a global construct
parallels between self-knowledge and other (self-esteem) to predict a host of specific
cognitive structures (e.g., Kuiper & Rog behaviors, Baumeister and colleagues had
ers, 1979; M arkus, 1977). By drawing on failed to heed a widely recognized doctrine
well-researched cognitive phenomena— such of psychometrics (see Ajzen & Fishbein,
as mental schemas, encoding, and priming 2005). We elaborate on this issue and related
effects— research on the self and self-esteem ones in the course of discussing the nature of
earned new credibility. self-esteem.
Independent of these developments in
academia, a self-esteem movement emerged
within the lay community in the late 1960s The N ature o f Self-Esteem
(Branden, 1994; see also Twenge & Camp
bell, 2001). The movement peaked in the A key aspect of understanding self-esteem is
1980s with the formation of the California recognizing its relationship to self-concepts
Task Force to Promote Self-Esteem and Per and other cognitive structures. Some au
sonal and Social Responsibility (1990). On thors contend that self-esteem is the “affec
the basis of no empirical evidence (in fact, tive” component of self-representation (i.e.,
evidence pointed to the contrary), the move what people feel about themselves) and that
ment characterized self-esteem as a panacea self-concepts are “cognitive” components of
that would cure a wide range of social ills, self-representation (i.e., what people believe
from teenage pregnancy and welfare depen about themselves). Although the affective-
dency to juvenile delinquency and low edu cognitive distinction is useful in some con
cational attainment. As a result, thousands texts, we do not believe that it is the most
of Americans came to believe not only that useful means of distinguishing self-esteem
raising self-esteem could cure all of society’s from self-concepts, as empirical support for
36. S elf-E steem 529
this distinction is lacking (M arsh, 1986; able (e.g., attitudes toward action films pre
Marsh & Hattie, 1996; Shavelson, Hubner, dict how many action films people watch in a
& Stanton, 1976). It is not difficult to see given year, but not the total number of mov
why. After all, many of the self-concepts ies they watch). Conversely, when a predic
that social-personality psychologists study tor variable is relatively general, the impact
are strongly affectively charged. People often of rival influences can be averaged out by
care a great deal, for example, about their combining numerous behaviors into the cri
beliefs that they are intelligent, athletic, or terion variable (e.g., attitudes toward movies
dominant. Likewise, social self-concepts in general predict how many movies of all
(self-concepts that align people to groups, types that people watch in a given year, but
such as Christian, American, or teacher) are not necessarily how many action films they
sometimes held so passionately that their watch). In short, specific predictors should
bearers make huge sacrifices for them, even be used to predict specific outcomes, and
to the point of giving up their lives. Not general predictors should be used to predict
only do self-concepts often have an affective general outcomes.
component, but self-esteem also has a be Applied to research on self-esteem, the
lief component; it is, after all, a belief about specificity-matching principle suggests that
one’s worth. Using the affective-cognitive researchers who use global self-esteem as a
distinction to distinguish self-esteem from predictor should focus on global outcome
self-concepts thus rests on shaky conceptual measures, such as several outcomes bun
as well as empirical grounds. dled together (see also Rosenberg, Schooler,
In this chapter we define self-esteem as a Schoenbach, & Rosenberg, 1995). From
global view of the self and self-concepts as the perspective of the specificity-matching
relatively specific views of the self along vari principle, then, Baumeister and colleagues’
ous dimensions (e.g., honest, clumsy, math (2003) review of the self-esteem literature
ematically inclined). Rather than making most likely underestimated the potential im
categorical distinctions between self-esteem portance of self-esteem because it focused
and self-concepts, we suggest that they rep on the capacity of global self-esteem to pre
resent different levels of specificity within dict specific outcomes (e.g., Does self-esteem
the superordinate category of self-views (see predict grades in a math class?).
also Swann, Chang-Schneider, & McClarty, The intricate interplay between self-
2007). concepts and self-esteem also figures im
This conceptualization of self-esteem has portantly in understanding the relationship
clear implications for how its consequences between constructs that have recently been
should be assessed. Specifically, if self-esteem integrated into the psychological literature.
and self-concepts are simply more or less Whereas the terms self-esteem and self-
specific members of the same overarching concept have traditionally been used to refer
category, it makes little sense to consider the to characteristics of single individuals, theo
predictive validity of one without simultane rists have recently popularized “groupier”
ously considering the predictive utility of the variations on these constructs, such as col
other. This point is related to a key insight lective self-esteem and group identity. As
from the past three decades of research on we discuss next, the key difference between
attitudes (e.g., Ajzen & Fishbein, 2005) and these distinct but related self-views lies in
traits (e.g., Epstein, 1979; Fleeson, 2004), how global (vs. specific) and group-like (vs.
dubbed the specificity-m atching principle. personal) their referents are.
To compensate for the fact that outcomes in
naturally occurring settings are often caused
T h e D im ensions o f Self-E steem
by multiple factors other than the predictor
and Self-C oncepts
variable of interest, the specificity-matching
principle holds that the specificity of predic We suggest that the referents of self-esteem
tors and criteria should be matched. When and self-concepts can be organized along
a predictor variable is relatively specific, the the two orthogonal dimensions of glohality
impact of rival influences on the predictor- and groupiness. As seen in Figure 36.1, self-
criterion relationship can be minimized by concepts or identities (we use these terms
selecting an equally specific criterion vari interchangeably) refer to personal qualities
530 V I. S E L F -R E L A T E D D IS P O S IT IO N S
that are relatively specific; hence they reside al feelings of worth that derive from one’s
in the lower left-hand quadrant of the figure. memberships in social groups. As such, it
Pelham and Swann’s (1989) Self-Attributes occupies the upper right-hand quadrant
Questionnaire, which asks respondents to of Figure 36.1. An example of a scale that
rank themselves relative to others along measures this type of self-view is Luhtanen
several dimensions (e.g., social skills, physi and Crocker’s (1992) Collective Self-Esteem
cal attractiveness, artistic ability), measures Scale, which asks respondents to indicate
this type of self-view. Similarly, Marsh and their agreement with statements such as “I
Shavelson’s (1985) Self-Description Ques am a worthy member of the social groups I
tionnaire assesses people’s self-concepts belong to” and “I’m glad to be a member of
along relatively specific dimensions, such as the social groups I belong to.”
academic, social, emotional, and physical.
Self-esteem also refers to a personal quality,
Certainty and Stability o f Self-E steem
but it is global in nature; hence it is located
and Self-C oncepts
in the upper left-hand quadrant of the fig
ure. To measure self-esteem, investigators In addition to varying along the dimensions
have people respond to statements such as of globality and groupiness, self-esteem and
“I feel that I am a person of worth, at least self-concepts differ in other meaningful
on an equal basis with others” (Rosenberg, ways. For instance, people differ in the ex
1965). tent to which their self-views are held with
To the right are self-views that are social certainty and are stable across time.
rather than personal in nature. Because group Generally speaking, the more converging
identity refers to relatively specific qualities evidence people have to support a given be
of groups (e.g., “Germans are industrious,” lief, the more certain of that belief they will
“Students care about their grades”), it ap be. Applying this principle to self-views, the
pears in the lower right-hand quadrant of more consistent evidence that people have
Figure 36.1. Although we are not aware of to support a particular view of themselves,
any scales designed explicitly to measure the more certain that self-view will be (e.g.,
group identity, measures of self-stereotyping Pelham, 1991). The certainty with which
(e.g., Biernat, Vescio, & Green, 1996) and people hold self-views, in turn, has impor
infrahumanization (e.g., Cortes, Demoulin, tant implications. For example, increases
Rodriguez, Rodriguez, & Leyens, 2 005) can in the certainty and confidence of people’s
be used to assess people’s beliefs about the self-knowledge predict increases in global
qualities that link them to their ingroups. self-esteem (Baumgardner, 1990; Campbell,
Finally, collective self-esteem refers to glob 1990).
36. S elf-E steem 531
1985). More recently, the quest for a mea odi, Swann, and their colleagues note that
sure of uncontaminated, “true” self-esteem although self-liking and self-competence are
led researchers to develop implicit tests of correlated, the correlation is moderate, and,
self-esteem (e.g., Greenwald, M cGhee, & more important, each component predicts
Schwartz, 1998). unique outcomes (e.g., Bosson & Swann,
Still, as noted, most self-esteem research 1999; Tafarodi & M ilne, 2002).
relies on respondents’ self-reports. This Another two-component approach dis
practice makes sense given that self-esteem tinguishes between trait self-esteem , which
is, by definition, the esteem that one has for refers to people’s baseline level of global
oneself. Asking people directly about their self-esteem that remains fairly stable across
feelings toward themselves is therefore a rea time, and state self-esteem , which fluctuates
sonable strategy for assessing such feelings. on a moment-to-moment basis in response
Among researchers who utilize self-report to self-relevant experiences. Heatherton
measures, however, there are widely diver and Polivy (1991) developed the State Self-
gent ideas about the number of distinct com Esteem Scale (SSES) to capture people’s im
ponents or aspects that presumably underlie mediate feelings about themselves within
global self-esteem. We group these diverging several domains (performance, social, and
perspectives here into single-component, appearance). However, this scale’s sub
two-component, and multiple-component stantial correlation with trait measures of
approaches. self-esteem (r ~ .75) raises questions about
whether it truly captures a distinct com
ponent of self-esteem. Others (e.g., Kernis,
The Single-Component Approach
2005) measure state self-esteem by adminis
Probably the most common approach to mea tering trait self-esteem scales multiple times
suring self-esteem is based on the assump throughout the day, with the instruction to
tion that it consists of a single, general di “respond according to how you feel about
mension that can be measured with a modest yourself right now.” Indeed, many standard
number of items (e.g., Coopersmith, 1967). trait self-esteem scales can be modified to
This assumption is evident in the most com assess state self-esteem by rewording their
monly used measure of self-esteem, Rosen instructions or by adding the phrase “at this
berg’s (1965) Self-Esteem Scale. Taking the m om ent. . . ” to individual items.
unifactorial assumption even further, some Another popular two-component ap
researchers recently developed a one-item proach is based on the distinction between
self-esteem scale that consists of the single implicit and explicit attitudes. Although dif
statement “I have high self-esteem” (Robins, ferent authors make different assumptions
Hendin, &c Trzesniewski, 2001). about the precise nature of explicit and im
plicit attitudes, one common view holds that
explicit self-esteem is controllable, deliber
The Two-Component Approach
ate, and easy to verbalize, whereas implicit
In recent years, it became increasingly popu self-esteem is uncontrollable, automatic,
lar to divide global self-esteem into two com and difficult to verbalize (Epstein & M or-
ponents. One approach— which harks back ling, 1995). Several unobtrusive methods are
to Osgood’s (1952) early work on the evalua used to capture implicit self-esteem, includ
tive and potency components of social judg ing measures of people’s preferences for their
ments, as well as Bakan’s (1966) distinction own initials relative to other letters (e.g., the
between communal and agentic aspects of Name Letter Task; Koole, Dijksterhuis, &
personality— distinguishes between people’s van Knippenberg, 2001) and reaction-time
assessments of their lovability (self-liking) tasks that assess the speed with which peo
and competence (self-com petence). Several ple associate positive versus negative stimuli
scales capture these components (e.g., Dig- with the self (e.g., the Implicit Association
gory, 1966; Franks & M arolla, 1976; Gecas, Test; Greenwald et al., 1998). The consis
1971), but the one that does so most explic tently low or nonexistent correlations be
itly is Tafarodi and Swann’s (2001) Self- tween explicitly and implicitly assessed self
Liking and Self-Competence Scale. Tafar esteem lend credence to the notion that they
36. S elf-E steem 533
are distinct while simultaneously raising and global self-esteem predicts aggregated
questions about whether implicit and explic outcomes better than specific self-concepts
it scales truly tap the same underlying con do (e.g., Trzesniewski et al., 2006). These
struct. Currently, theorists disagree on this patterns are consistent with the multiple-
point (see Buhrmester, Blanton, & Swann, component approach, which theorizes both
2 0 0 8 ; Olson, Fazio, & Hermann, 2 0 0 7 ) and global and specific dimensions of the self-
may continue to do so until further research concept.
sheds more light on this issue.
that focus on the sociocultural origins of that values the child’s inherent worth should
self-esteem. foster authentic feelings of true self-esteem
(Deci & Ryan, 1995).
Continuing through adolescence and into
N urture
adulthood, individuals continue to develop
If genes explain approximately 3 0 % of the specific self-concepts through comparisons
population variance in self-esteem, then this with others (Festinger, 1954), as well as
leaves roughly 70% of the variance to be ex observations of their own behavior (Bern,
plained by other factors, including environ 1972). The positivity versus negativity of
mental influences and gene x environment specific self-concepts, in turn, may influence
interactions. Much of the research on envi self-esteem via the importance that individu
ronmental influences on self-esteem explores als place on them. For example, individuals
how specific relationship partners— such as who place importance on success in a given
parents, siblings, peers, and teachers— as domain and who have positive self-concepts
well as the broader culture shape individu in this domain will enjoy higher global
als’ self-esteem. self-esteem than those who have negative
According to attachment theory (Bowlby, self-concepts in valued domains (Higgins,
1969), infants begin to formulate schemas 1987; Jam es, 1890/1950; Pelham, 1991).
(working models) about their worth, based Moreover, given the multidimensionality
on the treatment they receive from caregivers, of the self, successes in a given domain may
before they even have self-awareness. During predict increases in the positivity of specific
infancy and early childhood, working mod self-concepts w ithout also influencing glob
els reflect the consistency and responsiveness al feelings of self-esteem (Marsh & Craven,
of caregivers’ treatment. Specifically, consis 2006).
tent and responsive caregiving should instill On a broader level, self-concepts and self
in children the rudimentary foundations of esteem reflect the culture in which people
high self-esteem and favorable self-concepts are socialized. One consistent finding is
by teaching them that they are worthy of that, on average, people who are raised in
love and capable of efficacious action (Bowl individualistic cultures report substantially
by, 1973; Mikulincer, 1995; Verschueren, higher self-esteem and more favorable self-
M arcoen, & Schoefs, 1996). concepts than do people raised in collectiv-
During middle childhood (around the age istic cultures (Heine & Hamamura, 2007).
of 8), relatively sophisticated cognitive pro Indeed, in analyses that treat culture as the
cesses further refine children’s self-esteem unit of analysis, there is a strong positive cor
and self-concepts (Harter, 1990). For in relation between a culture’s individualism
stance, children at this age begin develop and the average self-esteem of its members
ing specific self-concepts by comparing their (Oyserman, Coon, &C Kemmelmeier, 2002).
traits and abilities with those of their peers Furthermore, as the length of exposure to an
(Festinger, 1954). They also begin looking individualistic culture increases, so does the
to others for feedback about the extent to self-esteem of visitors from a collectivistic
which they are valued (Cooley, 1902; Mead, culture (Heine & Lehman, 1997).
1934), and they internalize their perceptions These cross-cultural findings raise an inter
of others’ approval (or disapproval) as feel esting— and currently unresolved— question
ings of self-esteem. Thus, across childhood, about the “true” self-esteem of people from
high self-esteem is associated with positive individualistic versus collectivistic cultures.
self-concepts in valued domains and percep Some theorists suggest, for example, that the
tions of approval from significant relation tendency toward high self-esteem and posi
ship partners (Harter, 1999). Importantly, tive self-concepts is universal and that peo
the type of approval that children receive ple from collectivistic cultures merely appear
from others can influence their developing (relatively) low in self-esteem because of the
self-views. Whereas approval that is contin value they place on modest self-presentation
gent on the child accomplishing specific goals and “fitting in” rather than “standing out”
or meeting specific standards can foster self (Sedikides, Gaertner, & Toguchi, 2003). In
esteem that is unstable and fragile, approval support of this view, Sedikides and his col
36. S elf-E steem 535
leagues report that people from collectivistic (Kenrick, Groth, Trost, & Sadalla, 1993).
cultures display highly favorable views of When goals relevant to success in these
themselves on communal self-concepts that social domains are not met, negative self-
are valued within their culture, such as loy assessments and feelings of low self-esteem
alty (Sedikides, Gaertner, & Vevea, 2005). motivate the individual to either renew ef
Likewise, Tafarodi and Swann (1996) found forts toward goal achievement or redirect
that Chinese participants scored higher than energies elsewhere.
American participants on the self-liking di Another functional perspective suggests
mension of global self-esteem, whereas they that self-esteem feelings protect people from
scored lower than Americans on the self the existential anxiety that accompanies
competence dimension. In contrast, other awareness of their own mortality (Hart,
theorists call the tendency toward high self Shaver, & Goldenberg, 2 0 0 5 ; Pyszczynski,
esteem and positive self-concepts “strikingly Greenberg, Solomon, Arndt, & Schimel,
elusive” among people from collectivistic 20 0 4). According to this view, high self
East Asian cultures (Heine & Hamamura, esteem and positive self-concepts signal that
2007, p. 22) and argue that such individu one meets or exceeds the value standards as
als instead display a tendency toward self- sociated with one’s role(s) within a larger sys
criticism (Heine, Lehman, M arkus, & tem of meaning. Conversely, low self-esteem
Kitayama, 1999). Although this debate is and negative self-concepts signal a break
still going strong, one promising resolution down in the psychological “arm or” that
involves the development of a method for protects people from their deep-rooted fear
separating the self-presentational com po of death and its accompanying unknowns.
nent of self-esteem from “true” self-esteem. Thus drops in self-esteem and negative self-
For example, Kwan and Mandisodza (2007) assessments motivate behaviors geared to
identified three components of self-esteem: ward restoring one’s value in the eyes of oth
benevolence, merit, and bias. The bias ers and shoring up support for human-made
component is conceptually similar to self systems of meaning.
enhancement bias, whereas the benevolence Whereas the aforementioned perspectives
and merit components seem to reflect “true” suggest that self-esteem and self-concepts
self-esteem. This approach may provide a confer survival benefits, such arguments
starting point from which to pursue ques seem to come perilously close to mistaking
tions about the nature of self-esteem across an abstraction (i.e., self-esteem) for a thing
cultures. (i.e., a psychological entity that shapes rath
er than merely reflects reality). From this
vantage point, the survival benefits associat
Functional Perspectives
ed with self-esteem may merely reflect those
Rather than focusing on the origins of in qualities that give rise to self-esteem rather
dividuals’ self-esteem, several perspectives than self-esteem itself (cf. Baumeister et al.,
take a broader look by focusing on the 2003). Furthermore, excessive focus on self
origins of self-esteem itself. These perspec esteem may be problematic in and of itself.
tives ask: Why do humans have self-esteem Crocker and Park (2004), for example, sug
in the first place, and what function(s) are gest that preoccupation with one’s achieve
served by self-esteem? One such perspective ments in self-esteem-relevant domains can
proposes that self-esteem and self-concepts divert attention from other important needs,
reflect the operation of psychological mech such as the needs for relatedness, compe
anisms that evolved because they helped tence, autonomy, and self-regulation. Note,
humans negotiate the social world (Kirk however, that this view is not necessarily
patrick & Ellis, 2001). According to this incompatible with the functional views de
perspective, self-esteem and self-concepts scribed previously. Although self-esteem may
provide people with information about, for have evolved to serve the informational and/
example, their dominance status (Barkow, or protective functions noted earlier, valuing
1989), social inclusion versus exclusion self-esteem fo r its ow n sake may indeed yield
(Leary & Baumeister, 2 0 0 0 ), prestige (Hen- the maladaptive outcomes noted by Crocker
rich & Gil-W hite, 2001), and mate value and Park. Moreover, although self-esteem
536 V I. S E L F -R E L A T E D D IS P O S IT IO N S
is an abstraction, it can have motivational bust correlations with the valence of people’s
properties. For example, people who enjoy specific self-concepts, such that higher self
high self-esteem are likely to persist on tasks esteem is associated with more positive eval
in the wake of failure (McFarlin, Baumeister, uations of the self along specific dimensions
8c Blascovich, 1984), and those who suffer (Brown, Dutton, 8c Cook, 2 0 0 1 ; Pelham 8c
from negative self-views are prone to toler Swann, 1989), as well as smaller discrepan
ate various forms of poor treatment (Swann, cies between actual and ideal beliefs about
De La Ronde, 8c Flixon, 1994; Wiesenfeld, the self (Higgins, 1987). Higher self-esteem
Swann, Brockner, 8c Bartel, 2007). is also associated with smaller overall pro
portions of negative, relative to positive,
self-concepts (Hoyle, 2 0 0 6 ; Showers, 1992),
Correlates o f Self-Esteem and the negative self-concepts of people
and Self-Concepts with high self-esteem tend to be relatively
less complex and differentiated (Morgan
Research on the correlates of self-esteem 8c Janoff-Bulm an, 1994; W oolfolk, Nova-
and self-concepts is abundant. Given the lany, Gara, Allen, &C Polino, 1995). These
expansiveness of the literature, we can do structural features buffer people who have
little more than summarize broadly some of primarily favorable beliefs about themselves
the key findings. We organize these findings from the painful effects of negative self
temporally, beginning with the metacogni- relevant information (such as negative feed
tive features of self-knowledge and then pro back, memories of undesirable past behav
ceeding through goal setting, environment ior, etc.). Unfortunately, these same features
and partner selection, self-presentation, and do little to protect those who have many
cognitive and affective reactions and end negative self-views from painful reminders
ing with real-world outcomes. Note that in of their deficits (Showers, 1992).
keeping with our conviction that self-esteem
and self-concepts are members of the larger
D ecisions and Goals
self-view category, we include investigations
of both in our review. When it comes to decision making, research
Before proceeding, we must acknowledge paints a portrait of people with low self
the inherent difficulty of establishing cau esteem as being less decisive (Rosenberg 8c
sality when discussing individual-difference Owens, 2001) and more likely to procrasti
variables such as self-esteem and self- nate (Ferrari, 1994) than those high in self
concepts. Although it is certainly possible esteem. Persons lower in self-esteem are also
that self-esteem causes some of the variables more easily persuaded than those high in
with which it is associated, it is also pos self-esteem (Gibson, 1981), particularly in
sible that self-esteem is caused by some of response to forceful or heavy-handed com
these variables. Other possibilities include munications, which tend to yield reactance
the “third variable problem”— that is, the effects among those high in self-esteem
notion that a third, unmeasured variable (Brockner 8c Elkind, 1985). In a similar
causes changes in both self-esteem and the vein, people low as compared with high in
variables with which it correlates— and the self-esteem are also more risk averse when
prospect of dynamic interrelations in which making decisions, most likely because they
self-esteem causes some outcome, which have relatively low expectations of success
then influences self-esteem, and so on. Be (Wray 8c Stone, 2005) and are motivated to
cause of these difficulties, we avoid causal avoid feelings of regret should a risky deci
language when describing the correlates of sion yield negative consequences (Josephs,
self-esteem and self-concepts. Larrick, Steele, 8c N isbett, 1992).
In addition to making riskier decisions,
people high in self-esteem also tend to set
M etacognitive Features o f Self-K now ledge
higher goals for themselves and to persist
Metacognitive features include qualities such more doggedly through setbacks than those
as the content and structure of, and links low in self-esteem. Indeed, some research
among, different pieces of self-knowledge. suggests that persons high in self-esteem
For example, global self-esteem shows ro pursue goals with an eye to achieving excel
36. S elf-E steem 537
lence, whereas those low in self-esteem seek them more positive feedback (in the form of
merely to attain adequacy (Baumeister & financial compensation), whereas those low
Tice, 1985). Moreover, higher self-esteem is in self-esteem seek work environments that
associated with superior self-regulation dur offer fewer such financial rewards (Schroed-
ing goal pursuit. For example, people with er, Josephs, & Swann, 2006). Such tenden
high self-esteem persist more than those cies should ensure that people surround
with low self-esteem after a single failure, themselves with relationship partners, feed
but they persist less than persons low in back sources, and environments that bolster,
self-esteem after repeated failures (Di Paula rather than challenge, their self-esteem and
& Campbell, 2002). People with high self self-concepts. Moreover, to the extent that
esteem also persist more than those low in a given relationship or environment discon-
self-esteem if they believe that persistence firms people’s self-concepts or self-esteem,
is linked with success at a particular task, they are likely to leave in search of a better
but not if they believe that persistence is ir fitting niche (Schroeder et al., 2 0 0 6 ; Swann
relevant to success (McFarlin, 1985). These & Pelham, 2002).
findings suggest that persons high in self
esteem are particularly adept at modifying
Self-P resen ta tioti
their goal-pursuit strategies to reflect the
likelihood of goal attainment. Within their chosen relationships and en
vironments, people’s self-esteem and self-
concepts relate predictably to the manner in
Creating a N iche
which they present themselves. For example,
Once people make decisions and set goals, whereas persons high in self-esteem seek
they must select the environments and re to impress others— and thereby enhance
lationships within which to pursue those themselves— by presenting themselves in a
goals. According to self-verification theory, highly favorable manner, those low in self
the need for psychological coherence— or a esteem present themselves in a more modest,
sense that the world fits with past experienc self-protective fashion (Baumeister, Tice, &
es— is a primary motive behind the selection Hutton, 1989). Ironically, one of the ways in
of settings and interaction partners (Swann, which people high in self-esteem present a
Rentfrow, & Guinn, 2003). That is, people favorable image is by self-handicapping, or
actively seek and embed themselves within creating obstacles to their own success so as
social environments that sustain their stable to create plausible external attributions for
self-views. Evidence of this tendency appears poor performance (Jones & Berglas, 1978).
in people’s choices of relationship partners, To illustrate this phenomenon, Tice and
careers, home and work environments, Baumeister (1990) measured the amount
group memberships, and even home and of of time that people high versus low in self
fice decor (Gosling, Ko, M annarelli, & M or esteem spent practicing for an upcoming
ris, 2 0 0 2 ; Sadalla, Vershure, & Burroughs, test, under public versus private conditions.
1987). Only when they thought that others would
To illustrate, people low in self-esteem know how much time they practiced did
tend to withdraw and isolate themselves people high in self-esteem self-handicap by
from others, whereas those high in self engaging in less preparation than people low
esteem more readily seek others’ company in self-esteem. Thus the desire to present the
(Rosenberg & Owens, 2001). Once they self favorably may, at times, lead persons
enter social settings, people’s stable self with high self-esteem to behave in ways that
views predict their preferences for specific undermine their own performances.
interaction partners. Whereas people with
favorable self-concepts tend to seek out rela
Social Cognition
tionship partners who view them favorably,
those with negative self-concepts prefer the Social interactions provide the raw material
companionship of those who view them un for a host of social-cognitive processes that
favorably (Swann et al., 1994; Swann & Pel differ as a function of self-esteem and self-
ham, 2002). Similarly, people high in self concepts. In this section, we consider the
esteem seek work environments that offer links between self-views and social-cognitive
538 V I. S E L F -R E L A T E D D IS P O S IT IO N S
processes such as information seeking, at feedback and experiences that are congruent
tention, recall, interpretation, and mental relative to incongruent with the valence of
simulation. their self-esteem, and they recall incongru
ent feedback and experiences as more con
gruent than they really were (Christensen,
Information Seeking
Wood, & Barrett, 2 0 0 3 ; Story, 1998). Sim
Within their interactions, people tend to ilar effects have been found at the level of
seek self-relevant information that is consis self-concepts, with people displaying better
tent with their chronic, firmly held self-views memory for feedback that is congruent than
(e.g., Swann, 1983, 1990). Despite early incongruent with the positivity or negativity
findings suggesting that global self-esteem of their self-perceived likeability (Swann &
did not predict people’s reactions to posi Read, 1981). Interestingly, these congruency
tive or negative feedback (Swann, Pelham, effects in memory for self-relevant informa
& Krull, 1989), later studies revealed strong tion appear to be moderated by self-esteem
links between specific self-concepts and level, such that people high in self-esteem
information-seeking tendencies regarding exhibit a stronger congruency bias (i.e.,
those self-concepts. Thus, when researchers tendency to recall past behavior in a man
uphold the specificity-matching principle, ner congruent with self-concepts) than those
they find that people generally seek positive low in self-esteem (Campbell, 1990).
information about their favorable self-views Self-esteem differences in recall also
and negative information about their unfa emerge during threatening experiences. For
vorable self-views (e.g., Bosson & Swann, example, people high in self-esteem are more
1999). likely than those low in self-esteem to remem
ber other p e o p le’s negative behaviors follow
ing their own failure experiences (Crocker,
Attention
1993), and persons high in self-esteem spon
Just as people seek information that is con taneously recall more positive autobiograph
sistent with their self-views, they pay more ical memories than do persons low in self
attention to evaluatively consistent than in esteem when in an experimentally-induced
consistent information. In general, people negative mood (Setliff &c M armurek, 2002).
low as compared with high in self-esteem at Such recall biases presumably facilitate and
tend more to negative information and events hamper mood repair efforts among people
(Leitenberg, Yost, & Carroll-W ilson, 1986). with high and low self-esteem, respectively.
When it comes to self-relevant information,
people with negative self-concepts pay more
Interpretation
attention to unfavorable than favorable eval
uations of themselves, whereas the reverse is The manner in which people interpret their
true among those with positive self-concepts own and other people’s behaviors and out
(Swann & Read, 1981). In the wake of fail comes is linked predictably with their self
ure feedback, persons with low self-esteem esteem and self-concepts. For instance, peo
focus attention on their weaknesses, whereas ple interpret feedback that is congruent with
those high in self-esteem increase attention their self-concepts as accurate, whereas they
to their strengths (Dodgson & Wood, 1998). dismiss incongruent feedback as inaccurate
Finally, people high in self-esteem are more (Markus, 1977; Shrauger & Lund, 1975;
likely than those low in self-esteem to focus Swann, Griffin, Predmore, & Gaines, 1987).
on the ways in which their own outcomes Moreover, a large body of research on at
compare favorably to the outcomes obtained tribution processes shows that people high
by the friends, acquaintances, and strangers in self-esteem take credit for their successes
that they encounter in daily life (Wheeler & and blame their failures on external factors
Miyake, 1992). (for reviews, see Blaine & Crocker, 1993;
Campbell & Sedikides, 1999). In contrast,
people low in self-esteem are less inclined
Recall
to take credit for their successes and more
Perhaps reflecting these differences in at inclined to assume responsibility for their
tention, people display better memory for failures (e.g., Fitch, 1970).
36. S elf-E steem 539
Self-esteem also relates to the manner in wise, people higher in self-esteem tend to
which people interpret ambiguous social score higher in extraversion and positive af
stimuli. To illustrate, people who are high as fectivity (Watson et al., 2002), which reflect
compared with low in self-esteem are more chronic tendencies toward positive emotions
likely to interpret ambiguous phrases (“Is such as enthusiasm and joy. Not surprisingly,
this how you want it?”) as conveying posi research reveals strong and consistent posi
tive feelings toward them (Bosson, Swann, tive links between self-esteem and reports of
& Pennebaker, 2 0 0 0 ; Tafarodi, 1998). Fur subjective happiness (e.g., Diener & Diener,
thermore, people with low self-esteem may 1995), leading Baumeister and colleagues
not even interpret their own success experi (2003) to conclude-— in the midst of their
ences as successes unless a credible outsider otherwise disparaging review— that “high
tells them explicitly that they have done well self-esteem may pay off handsomely for the
(Josephs, Bosson, & Jacobs, 2003). individual in terms of subjective happiness”
(p. 26). Related to this self-esteem-happiness
link is a strong positive correlation between
Mental Simulation
self-esteem and optimism, or the tendency
Paralleling these self-esteem differences in to anticipate positive future outcomes for
interpretation are differences in people’s the self (Lyubomirsky, Tkach, & DiM atteo,
mental simulations, or thoughts regarding 2005).
alternative possible outcomes for themselves. Considerably less research explores the
Whereas people low in self-esteem tend to links between self-esteem and self-conscious
think more about how future outcomes emotions, but the existing work points to
“could be better,” those high in self-esteem strong negative correlations between self
think more about how future outcomes esteem and shame proneness (Leith &
“could be worse” (Sanna & Meier, 20 0 0 ). Baumeister, 1998), moderate negative cor
Similar self-esteem differences emerge when relations between self-esteem and hubris-
people generate alternative outcomes for past tic (all-encompassing) pride, and strong
events, with people low in self-esteem simu positive correlations between self-esteem
lating more “could have been better” sce and authentic (achievement-oriented) pride
narios and those high in self-esteem simulat (Tracy & Robins, 2007). Thus people high
ing more “could have been worse” scenarios in self-esteem neither react to their own fail
(Sanna, Turley-Ames, & Meier, 1999). ures and transgressions with painful feelings
of disgrace nor react to their successes with
overblown feelings of arrogance. Instead,
Affect
they appear to feel good or bad about their
Given the aforementioned differences in actions in a given context, rather than feeling
self-knowledge, choice of partners and envi good or bad about themselves as a whole.
ronments, and cognitive responses to their
worlds, it should come as no surprise that
L ife Outcomes
people’s self-esteem and self-concepts are
closely tied to their chronic and moment- In this section, we consider some of the ways
to-moment affective states. As noted, global in which the self-esteem and self-concept dif
self-esteem is strongly negatively corre ferences summarized here predict real-world
lated with neuroticism (Judge et al., 2002) outcomes in terms of people’s relationship
and negative affectivity (Suls, 2 0 0 6 ), both functioning, academic and athletic perfor
of which reflect people’s stable tendencies mances, criminal activity, health behaviors,
to experience unpleasant emotions. Thus and finances.
people who are higher in self-esteem tend As noted earlier, some theorists propose
to experience fewer negative emotions such that self-esteem evolved to alert people to
as depression, anxiety, and hostility. Indeed, survival-relevant fluctuations in their rela
the negative association between self-esteem tionship status (Leary & Baumeister, 2000).
and depression is so strong (r ~ .8 0 ; Watson According to this sociometer hypothesis,
et al., 2002) that some suggest conceptualiz painful drops in self-esteem inform people
ing self-esteem and depression as end points about possible threats to their social inclu
of a bipolar continuum (Suls, 2 0 0 6 ). Like sion (Leary, Tambor, Terdal, & Downs,
540 V I. S E L F -R E L A T E D D IS P O S IT IO N S
1995). Consistent with this idea, people low Lauver, Le, Langley, & Carlstrom, 2 0 0 4 ;
in self-esteem exhibit an attentional bias to Valentine, DuBois, & Cooper, 2004). In the
ward information that conveys interpersonal domain of athletics, physical self-concepts
rejection, whereas those high in self-esteem predict future exercise behavior, gymnas
pay particular attention to information that tic self-concepts predict future gymnastic
conveys acceptance (Dandeneau & Baldwin, performance, and swimming self-concepts
200 4 ). Unfortunately for those low in self predict performance during elite swim
esteem, their heightened sensitivity to rejec ming competitions (for a review, see Marsh
tion cues can have harmful implications for & Craven, 2006). It is worth noting that
their close relationships. To illustrate, the these effects emerge even when controlling
heightened rejection sensitivity of those low for past performance in the domain of in
in self-esteem undermines their confidence terest, indicating that self-concepts explain
in romantic partners’ love for them, which unique variance in people’s behavioral out
then leads them to withdraw psychologically comes. Conversely— and consistent with the
from partners (Murray, Holmes, M acD on specificity-matching principle— global self
ald, & Ellsworth, 1998). Moreover, people esteem predicts bundled outcomes, or sum
low in self-esteem may react to relationship mary indices that combine multiple behav
conflict in ways that anger and frustrate ioral observations. Some work, for example,
their partners, ultimately eliciting the very shows that people lower in self-esteem dur
rejection they fear most (Downey, Freitas, ing adolescence are more likely to develop
Michaelis, & Khouri, 1998). In contrast, the physical and mental health difficulties, to
expectations of acceptance of persons high in use tobacco, to commit crimes, to drop out
self-esteem allow them to use their romantic of school, and to suffer money and work
relationships as sources of self-affirmation problems in adulthood (Trzesniewski et al.,
in the face of failure, thus furthering their 2006). Thus many important life outcomes
confidence in their partners’ positive regard can be predicted by people’s specific self-
and increasing their commitment to those concepts and global self-esteem.
partners (Murray et al., 1998).
People’s specific self-concepts also figure
importantly in relationship functioning. Future D irections
Self-concepts predict the types of appraisals
that people seek and prefer from their part We began this chapter by acknowledging the
ners, as well as their feelings of commitment deep doubts that several influential critics
to and intimacy with partners who offer expressed recently regarding the self-esteem
them congruent appraisals. In relationships construct (e.g., Baumeister et al., 2 0 0 3 ;
ranging from college roommates to long Crocker & Park, 2004). Although we agree
term married partners, people with positive that simpleminded characterizations of self
self-concepts prefer partners who view them esteem as a panacea for all of society’s ills
favorably, whereas those with negative self- are wrongheaded, we believe that some crit
concepts prefer partners who view them neg ics have gone too far in arguing for the aban
atively (Swann & Pelham, 2 0 0 2 ; Swann et donment of the self-esteem construct. In
al., 1994). Indeed, people experience higher support of this viewpoint, we summarized a
levels of marital distress to the extent that vast literature that suggests that self-esteem
their spouses’ views of them disconfirm their and self-concepts are predictive of people’s
stable self-concepts (Schafer, W ickram a, & behaviors, thoughts, feelings, and life out
Keith, 1996). As such, securing relation comes. We now outline three suggestions for
ship partners who confirm their self-views improving the study of self-esteem.
may be important for people’s psychological First, when addressing matters related to
well-being (e.g., Swann et al., 2003). predictive validity, self-esteem should be re
In the academic domain, there are strong united with other members of the self-view
links between people’s relatively specific ac family. This will mean moving away from
ademic self-concepts and outcomes such as the knee-jerk use of Rosenberg’s (1965) glob
academic achievement, college grade point al self-esteem scale and toward assessing the
average, and persistence at academic pur key components of self-esteem (self-liking
suits (e.g., Marsh & Craven, 2 0 0 6 ; Robbins, vs. self-competence, implicit vs. explicit self
36. S elf-E steem 541
esteem), as well as the specific self-concepts translate into behaviors and outcomes out
that are most relevant to researchers’ out side of the laboratory.
come variables. In addition, researchers may Finally, in light of the debilitating predic
benefit from assessing the metacognitive fea tive outcomes of low self-esteem and nega
tures of self-esteem and self-concepts, such as tive self-concepts, it is critical to learn more
their certainty, importance, clarity, extrem about how they can be changed. We rec
ity, accessibility, organizational structure, ognize, of course, the irony of ending this
and temporal stability, to name a few. Such chapter with the question of how to change
shifts not only make sound conceptual sense, self-views, as changing self-views was the
but they are also consistent with the way original (and almost comically misguided)
that related psychological constructs, such goal of the much-maligned California task
as attitudes and traits, have been conceptu force. While acknowledging this irony, we
alized and studied. Furthermore, once other also defend our position by pointing out
members of the self-view family are thrown that self-esteem change, when based on
into the mix, specificity matching becomes empirically substantiated strategies, can
possible, and following this psychometric theoretically produce large improvements in
principle will lead to assessments that are people’s well-being and overall functioning.
simultaneously more meaningful and more In this regard, we are encouraged by recent
optimistic. N ote, however, that we do not evidence that self-esteem can be improved
recommend that researchers blur the distinc via elaborate programs (e.g., DuBois & Flay,
tion between global self-esteem and specific 2 0 0 4 ; Haney & Durlak, 1998). O f course,
self-concepts. To the contrary, we are simply self-esteem programs are not for everyone—
pointing out the importance of recognizing after all, most people in the general popula
that self-esteem and self-concepts are mem tion have high self-esteem and thus do not
bers of the same self-view category and that require self-esteem interventions. Further
following the specificity-matching principle more, successful self-esteem improvement
will undoubtedly improve researchers’ abil programs have all been multifaceted, and it
ity to predict the outcomes of self-esteem. is not clear which of their many components
Second, as in research on attitudes, theo are effective in generating change, or how
retical models of the factors that constrain they do so. Rather than boosting self-esteem
the links between self-views and behavior directly, it is possible that these programs
should be developed. Attitude researchers have their effects by increasing people’s so
have approached this challenge in two dis cial skills and interpersonal problem-solving
tinct ways. First, in their reasoned-action abilities, for example, which then leads to
model, Ajzen and Fishbein (2005) identified increases in social acceptance, which in turn
the many normative, contextual, and per improves self-esteem (e.g., Leary, 1999).
sonal variables that moderate the links be Rigorous empirical work is needed to pin
tween attitudes and behaviors, thus allow point the strategies that most effectively in
ing for heightened precision when predicting crease self-esteem, to uncover the underlying
behavioral outcomes from attitudes. Second, mechanism(s) that drive this change, and to
in his motivation and opportunity as deter explore whether increasing the positivity of
minants (MODE) model, Fazio (1990) of self-esteem and self-concepts can, in fact,
fered a process model of the chain of events engender some of the beneficial outcomes
that determines when attitudes will become that inspired the original efforts of the task
translated into behaviors. Fazio and his col force.
leagues (e.g., Olson et al., 2 0 0 7 ) have made
progress in applying the M ODE model to
the study of self-esteem, but additional R e feren ces
work is needed. For example, their initial
work focuses primarily on the conditions Ajzen, I., & Fishbein, M. (2005). T he influence of atti
under which people’s global self-esteem (an tudes on behavior. In D. Albarracin, B. T. Johnson,
& M. P. Zanna (Eds.), T h e h a n d b o o k o f a ttitu d es
attitude) translates into self-reports of self
(pp. 1 7 3 - 2 2 1 ). Mahwah, NJ: Flrlbaum.
esteem (a behavior). It is important to know American Psychiatric Association. (20 0 0). D ia g n o stic
as well the conditions under which both a n d sta tis tic a l m a n u a l o f m e n ta l d iso rd er s (4th ed.,
global self-esteem and specific self-concepts text rev.). Washington, DC : Author.
542 V I. S E L F -R E L A T E D D IS P O S IT IO N S
Ames, D. R., Rose, P., Sc Anderson, C. P. ( 2 0 06 ). The From the top down: Self-esteem and self-evaluation.
NPI-16 as a short measure of narcissism. J o u r n a l o f C o g n itio n a n d E m o tio n , 15, 6 1 5 - 6 3 1 .
R e se a rc h in P erson ality , 4 0 , 4 4 0 - 4 5 0 . Brown, R. P., Sc Zeigler-Hill, V. (20 04 ). Narcissism
Bakan, D. (1966). T h e d u a lity o f h u m a n e x is te n c e : and the non-equivalence of self-esteem measures: A
An essa y o n p s y c h o lo g y a n d relig ion . Oxford, UK: matter of dominance? J o u r n a l o f R e se a rc h in P er
Rand McNally. so n a lity , 3 8 , 5 8 5 - 5 9 2 .
Barkow, J. H. (1989). D a rw in , sex , a n d statu s: B io s o Buhrmester, M ., Blanton, H., Sc Swann, W. B., jr.
c ia l a p p r o a c h e s to m in d a n d cu ltu re. Toronto, O n (2 0 08 ). M easu rin g im p licit s e lf-e s te e m : N o t th ere
tario, Canada: University of Toronto Press. yet. Unpublished manuscript, University of Texas
Baumeister, R. F., Campbell, J. D., Krueger, J. I., Sc at Austin.
Vohs, K. D. (2 003). Does high self-esteem cause bet Bushman, B. J ., Sc Baumeister, R. F. (1998). T hreat
ter performance, interpersonal success, happiness, ened egotism, narcissism, self-esteem, and direct
or healthier lifestyles? P s y c h o lo g ic a l S cien c e in the and displaced aggression: Does self-love or self-hate
P u blic In terest, 4, 1 - 4 4 . lead to violence? J o u r n a l o f P erso n a lity a n d S o c ia l
Baumeister, R. F., & Tice, D. M . (1985). Self-esteem P sy ch olog y , 75, 2 1 9 - 2 2 9 .
and responses to success and failure: Subsequent California Task Force to Promote Self-Esteem and Per
performance and intrinsic motivation. J o u r n a l o f sonal and Social Responsibility. (1 990). T o w a rd a
P erson ality , S 3, 4 5 0 - 4 6 7 . sta te o f se lf-e s te e m . Sacramento: California State
Baumeister, R. F., Tice, D. M ., & Hutton, D. G. (1989). Department of Education.
Self-presentational motivations and personality dif Campbell, J . D. (1 990). Self-esteem and clarity of the
ferences in self-esteem. J o u r n a l o f P erson ality , 57, self-concept. J o u r n a l o f P erson ality a n d S o c ia l P sy
547-579. ch o lo g y , 59, 5 3 8 - 5 4 9 .
Baumgardner, A. H. (1990). To know oneself is to Campbell, W. K. (1999). N arcissism in ev e r y d a y life.
like oneself: Self-certainty and self-affect. J o u r n a l Unpublished manuscript, Case Western Reserve
o f P erson ality a n d S o c ia l P sy ch olog y , 5 8 , 1 0 6 2 - University.
1 072. Campbell, W. K., Bosson, J . K., Goheen, T. W., Lakey,
Bern, D. J . (1972). Self-perception theory. In L. Berkow C. E., & Kernis, M . H. (200 7 ). Do narcissists dis
itz (Ed.), A d v a n c e s in e x p e r im e n ta l s o c ia l p s y c h o l like themselves “deep down inside” ? P sy c h o lo g ic a l
o g y (Vol. 6, pp. 1 - 6 2 ). New York: Academic Press. S cien ce, 18 , 2 2 7 - 2 2 9 .
Biernat, M., Vescio, T. K., Sc Green, M. L. (1996). Se Campbell, W. K., Sc Sedikides, C. (1999). Self-threat
lective self-stereotyping. J o u r n a l o f P erso n a lity a n d magnifies the self-serving bias: A meta-analytic inte
S o c ia l P sy c h o lo g y , 71, 1 1 9 4 - 1 2 0 9 . gration. R e v ie w o f G e n e r a l P sy ch olog y , 3 , 2 3 - 4 3 .
Blaine, B., Sc Crocker, J . (1993). Self-esteem and self- Christensen, T. C., Wood, J. V., & Barrett, L. F.
serving biases in reactions to positive and negative (2003). Remembering everyday experience through
events. In R. F. Baumeister (Ed.), S e lf-es te em : T h e the prism of self-esteem. P erson ality a n d S o c ia l P sy
p u z z le o f lo w self-r eg a r d (pp. 5 5 - 8 5 ) . New York: c h o lo g y B u lletin , 2 9 , 5 1 - 6 2 .
Plenum Press. Cooley, C. H. (1902). H u m a n n atu re a n d th e s o c ia l
Blascovich, J ., Sc Tom aka, J. (1991). Measures of self ord er. New York: Scribner’s.
esteem. In J . Robinson, P. Shaver, Sc L. Wrightsman Coopersmith, S. (1967). T h e a n te c e d e n ts o f s e lf
(Eds.), M easu res o f p e r s o n a lity a n d s o c ia l p s y c h o es te e m . San Francisco: Freeman.
lo g ic a l a ttitu d es (pp. 1 1 5 -1 6 0 ). San Diego, CA: Cortes, B. P., Demoulin, S., Rodriguez, R. T., Rodri
Academic Press. guez, A. P., Sc Leyens, J.-P. (2005). Infrahumaniza
Bosson, J. K., Lakey, C. E., Campbell, W. K., Zeigler- tion or familiarity?: Attribution of uniquely human
Hill, V., Jordan, C. H., & Kernis, M. H. ( 2 0 0 8 , emotions to the self, the ingroup, and the outgroup.
April). Untangling the links between narcissism and P erso n a lity a n d S o c ia l P sy ch o lo g y B u lletin , 31,
self-esteem: A theoretical and empirical review. S o 243-253.
c ia l a n d P erson ality P sy ch o lo g y C o m p a s s , p. 2. Crocker, J. (1993). Memory for information about
Bosson, J . K., & Swann, W. B., Jr. (1999). Self-liking, others: Effects of self-esteem and performance
self-competence, and the quest for self-verification. feedback. Jo u r n a l o f R e se a rc h in P erson ality , 27,
P erso n a lity a n d S o c ia l P sy ch o lo g y B u lletin , 2 5 , 35-48.
1230-1241. Crocker, J ., Sc Park, L. E. (20 04 ). The costly pursuit
Bosson, J. K., Swann, W. B., Jr., Sc Pennebaker, J . W. of self-esteem. P sy c h o lo g ic a l B u lletin , 130, 3 9 2 -
( 2 0 00 ). Stalking the perfect measure of implicit self 414.
esteem: T he blind men and the elephant revisited? Dandeneau, S. D., Sc Baldwin, M. W. (20 04 ). The
J o u r n a l o f P ers o n a lity a n d S o c ia l P sy ch olog y , 79, inhibition of socially rejecting information among
631-643. people with high versus low self-esteem: T he role
Bowlby, J . (1969). A tta c h m e n t a n d lo s s: Vol. 1. A t of attentional bias and the effects of bias reduction
t a c h m en t. New York: Basic Books. training. J o u r n a l o f S o c ia l a n d C lin ica l P sy ch olog y ,
Bowlby, J . (1973). A tta c h m e n t a n d lo s s: Vol. 2 . S e p a r a 23, 5 84-602.
tion , a n x iety , a n d an ger. London: Penguin Books. Dawes, R. M. (1994). H o u s e o f ca rd s: P sy ch o lo g y
Branden, N. (1994). T h e six p illa rs o f se lf-e s te e m . New a n d p sy c h o th e r a p y b u ilt o n m yth. New York: Free
York: Bantam Books. Press.
Brockner, J . , Sc Elkind, M . (1985). Self-esteem and Deci, E. L., Sc Ryan, R. M. (1995). Human autonomy:
reactance: Further evidence of attitudinal and m o The basis for true self-esteem. In M . H. Kernis (Ed.),
tivational consequences. J o u r n a l o f E x p e r im e n ta l E ffic a c y , a g en cy , a n d s e lf-e s te e m (pp. 3 1 - 4 9 ) . New
S o c ia l P sy c h o lo g y , 2 1, 3 4 6 - 3 6 1 . York: Plenum Press.
Brown, J. D., Dutton, K. A., Sc Cook, K. E. (2001). Demo, D. H. (1985). The measurement of self-esteem:
36. S elf-E steem 543
Refining our methods. J o u r n a l o f P erso n a lity a n d Gosling, S. D., Ko, S. J ., Mannarelli, T., 8c Morris,
S o c ia l P sy c h o lo g y , 4 8 , 1 4 9 0 - 1 5 0 2 . M. E. (2 002). A room with a cue: Personality judg
Di Paula, A., 8c Campbell, J. D. (200 2). Self-esteem ments based on offices and bedrooms. J o u r n a l o f
and persistence in the face of failure. J o u r n a l o f P er P erson ality a n d S o c ia l P sy ch olog y , 8 2 , 3 7 9 - 3 9 8 .
so n a lity a n d S o c ia l P sy ch olog y , 8 3 , 7 1 1-72 4. Greenwald, A. G., McGhee, D. E., 8c Schwartz, J. L. K.
Diener, E., 8c Diener, M . (1995). Cross-cultural corre (1998). Measuring individual differences in implicit
lates of life satisfaction and self-esteem. J o u r n a l o f cognition: The Implicit Association Test .J o u r n a l o f
P erson ality a n d S o c ia l P sy ch olog y , 6 8 , 6 5 3 - 6 6 3 . P erson ality a n d S o c ia l P sy ch olog y , 74, 1 4 6 4 - 1 4 8 0.
Diggory, J . C. (1966). S e lf-ev a lu a tio n : C o n c e p ts a n d Haney, P., 8c Durlak, J. A. (1998). Changing self
stu d ies. New York: Wiley. esteem in children and adolescents: A meta-analytic
Dodgson, P. G., 8c Wood, J. V. (1998). Self-esteem and review. Jo u rn a l o f C lin ica l C h ild P sy ch olog y , 27,
the cognitive accessibility of strengths and weak 423-433.
nesses after failure. J o u r n a l o f P erson ality a n d S o Hart, J., Shaver, P. R., & Goldenberg, J . R. (2 005). At
c ia l P sy c h o lo g y , 75, 178-197. tachment, self-esteem, worldviews, and terror man
Donnellan, B., Trzesniewski, K., Robins, R., Moffitt, agement: Flvidence for a tripartite security system
T., 8t Caspi, A. (2005). Low self-esteem is related theory. J o u r n a l o f P erson ality a n d S o c ia l P sy ch ol
to aggression, antisocial behavior, and delinquency. ogy, 8 8 , 9 9 9 - 1 0 1 3 .
P sy c h o lo g ic a l S c ien c e, 16, 3 2 8 - 3 3 5 . Harter, S. (1990). Causes, correlates, and the function
Downey, G., Freitas, A. L., Michaelis, B., 8c Khouri, al role of global self-worth: A life-span perspective.
H. (1998). T he self-fulfilling prophecy in close rela In J. Kolligian & R. Sternberg (Eds.), P ercep tio n s
tionships: Rejection sensitivity and rejection by ro o f c o m p e te n c e a n d in c o m p e t e n c e a c r o s s the life
mantic partners. J o u r n a l o f P erso n a lity a n d S o c ia l sp a n (pp. 6 7 - 9 8 ) . New Haven, CT: Yale University
P sy ch olog y , 7 5 , 5 4 5 - 5 6 0 . Press.
DuBois, D. I.., & Flay, B. R. ( 2 00 4). The healthy pur Harter, S. (1999). T h e c o n str u c tio n o f th e se lf: A d e v e l
suit of self-esteem: Comment on and alternative to o p m e n t a l p er sp e c tiv e . New York: Guilford Press.
the Crocker and Park (20 04 ) formulation. P sy c h o Harter, S., 8c Pike, R. (1971). T he pictorial scale of
lo g ic a l B u lletin , 13 0 , 4 1 5 - 4 2 0 . Perceived Competence and Social Acceptance for
Epstein, S. (1979). Stability of behavior: On predicting Young Children. C h ild D e v e lo p m e n t, 5 5, 1 96 9—
most of the people much of the time. J o u r n a l o f P er 1982.
so n a lity a n d S o c ia l P sy ch olog y , 37, 1 0 9 7 - 1 1 2 6 . Heatherton, T. F., & Polivy, J. (1991). Development
Epstein, S., 8c Morling, B. (1995). Is the self motivated and validation of a scale for measuring state self
to do more than enhance and/or verify itself? In M. esteem. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o l
H. Kernis (Ed.), E ffic a c y , ag en cy , a n d s e lf-e stee m og y , 6 0, 8 9 5 - 9 1 0 .
(pp. 9 - 2 9 ) . New York: Plenum Press. Heine, S. J., & Hamamura, D. R. (200 7). In search of
Fazio, R. H. (19 90). Multiple processes by which at East Asian self-enhancement. P erson ality a n d S o
titudes guide behavior: The M O D E model as an c ia l P sy ch o lo g y R e v iew , 11, 1 -24.
integrative framework. A d v a n ce s in E x p e r im e n ta l Heine, S. J ., 8c Lehman, D. R. (1997). The cultural
S o c ia l P sy ch olog y , 2 3 , 7 5 - 1 0 9 . construction of self-enhancement: An examination
Ferrari, J. R. (1994). Dysfunctional procrastination of group-serving biases. J o u r n a l o f P erson ality a n d
and its relationship with self-esteem, interpersonal S o c ia l P sy ch o lo g y , 7 2, 1 2 6 8 - 1 2 8 3 .
dependency, and self-defeating behaviors. P ers o n a l Heine, S. J ., Lehman, D. R., Markus, H. R., 8c Kitaya
ity a n d In d iv id u a l D iffe r e n c e s , 17, 6 7 3 - 6 7 9 . ma, S. (1999). Is there a universal need for positive
Festinger, L. (1954). A theory of social comparison self-regard? P sy c h o lo g ic a l R ev iew , 1 0 6 , 7 6 6 - 7 9 4 .
processes. H u m an R e la tio n s , 7, 117-14 0. Henrich, J ., 8c Gil-White, F. J . (2001). The evolution of
Fitch, G. (1970). Effects of self-esteem, perceived prestige: Freely conferred deference as a mechanism
performance, and choice on causal attributions. for enhancing the benefits of cultural transmission.
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 16, E v o lu tio n a n d H u m an B eh a v io r, 2 2 , 1 6 5 - 1 9 6 .
311-315. Higgins, E. T. (1987). Self-discrepancy: A theory re
Fleeson, W. (2 004). Moving personality beyond the lating self and affect. P sy c h o lo g ic a l R e v iew , 94,
person-situation debate: T he challenge and oppor 319-340.
tunity of within-person variability. C u rren t D ire c Hoyle, R. H. (1991). Evaluating measurement models
tio n s in P sy c h o lo g ic a l S cien ce , 13, 8 3 - 8 7 . in clinical research: Covariance structure analysis
Franks, D. D., 8c Marolla, J . (1976). Efficacious ac of latent variable models of self-conception. J o u r n a l
tion and social approval as interacting dimensions o f C o n su ltin g a n d C lin ica l P sy ch olog y , 59, 6 7 - 7 6 .
of self-esteem: A tentative formulation through con Floyle, R. H. (20 0 6). Self-knowledge and self-esteem.
struct validation. S o c io m etr y , 39, 3 2 4 - 3 4 1 . In M. H. Kernis (Fid.), S elf-es te em issu es a n d a n
Freud, S. (1914/1957). On narcissism: An introduc sw ers: A s o u r c e b o o k o f cu rren t p ersp e ctiv e s
tion. In J. Strachey (Ed. 8c Trans.), T h e s ta n d a r d (pp. 2 0 8 - 2 1 5 ) . New York: Psychology Press.
ed itio n o f th e c o m p le t e p s y c h o lo g ic a l w o r k s o f S ig James, W. (1950). T h e p rin c ip le s o f p sy c h o lo g y . New
m u n d F reu d (Vol. 14, pp. 6 7 - 1 0 4 ) . London: Ho ga York: Dover. (Original work published 1890)
rth Press. (Original work published 1914) Jones, E. F.., 8c Berglas, S. (1978). Control of attri
Gecas, V. (1971). Parental behavior and dimensions of butions about the self through self-handicapping
adolescent self-evaluation. S o c io m etr y , 3 4 , 4 6 6 - strategies: The appeal of alcohol and the role of un
482. derachievement. P erso n a lity a n d S o c ia l P sy ch olog y
Gibson, J. L. (1981). Personality and elite political be B u lletin , 4, 2 0 0 - 2 0 6 .
havior: T he influence of self-esteem on judicial deci Josephs, R. A., Bosson, J. K., 8c Jacobs, C. G. (2003).
sion making. J o u r n a l o f P olitics, 4 3 , 1 0 4 - 1 2 5 . Self-esteem maintenance processes: Why low self
544 V I. S E L F -R E L A T E D D IS P O S IT IO N S
esteem may be resistant to change. P erson ality a n d (1986). Negative cognitive errors in children: Ques
S o c ia l P sy c h o lo g y B u lletin , 2 9 , 9 2 0 - 9 3 3 . tionnaire development, normative data, and co m
Josephs, R. A., Larrick, R. P., Steele, C. M ., 8c Nisbett, parisons between children with and without self-
R. E. (1992). Protecting the self from the negative reported symptoms of depression, low self-esteem,
consequences of risky decisions. J o u r n a l o f P ers o n and evaluation anxiety. J o u r n a l o f C o n su ltin g a n d
ality a n d S o c ia l P sy ch olog y , 6 2 , 2 6 - 3 7 . C lin ica l P sy ch olog y , 5 4 , 5 2 8 - 5 3 6 .
Judge, T. A., Erez, A., Bono, J. E., 8c Thoresen, C. j . Leith, K. P., & Baumeister, R. F. (1998). Empathy,
(2002). Are measures of self-esteem, neuroticism, shame, guilt, and narratives of interpersonal con
locus of control, and generalized self-efficacy indi flicts: Guilt-prone people are better at perspective
cators of a common core construct? J o u r n a l o f P er taking. J o u r n a l o f P erson ality , 6 6 , 1-37.
so n a lity a n d S o c ia l P sy ch olog y , 8 3 , 6 9 3 - 7 1 0 . Luhtanen, R., 8c Crocker, J. (1992). A collective self
Kendler, K. S., Gardner, C. O., Sc Prescott, C. A. esteem scale: Self-evaluation of one’s social identity.
(1998). A population-based twin study of self P erson ality a n d S o c ia l P sy c h o lo g y B u lletin , 18,
esteem and gender. P s y c h o lo g ic a l M ed icin e, 2 8 , 302-318.
1403-1409. Lyubomirsky, S., T k a ch, C., 8c DiM atteo , M . R.
Kenrick, D. T., Groth, G. E., Trost, M. R., 8c Sadalla, (2005). What are the differences between happiness
E. K. (1993). Integrating evolutionary and social ex and self-esteem? S o c ia l In d ic a to r s R e se a rch , 78,
change perspectives on relationships: Effects of gen 36 3 - 4 0 4 .
der, self-appraisal, and involvement level on mate Markus, H. (1977). Self-schemas and processing infor
selection criteria. J o u r n a l o f P erso n a lity a n d S o c ia l mation about the self. J o u r n a l o f P erson ality a n d
P sy ch olog y , 6 4 , 9 5 1 - 9 6 9 . S o c ia l P sy ch olog y , 35, 6 3 - 7 8 .
Kernis, M. H. (2001). Following the trail from narcis M ar sh , H. W. (1986). Global self-esteem: Its relation to
sism to fragile self-esteem. P sy c h o lo g ic a l In q u iry , specific facets of self-concept and their importance.
1 2 ,2 2 3 - 2 2 6 . J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 5 i ,
Kernis, M. H. (2 003). Toward a conceptualization 1224-1236.
o f optimal self-esteem. P sy c h o lo g ic a l In q u iry , 14, M ar sh , H. W., 8c Craven, R. G. (20 0 6). Reciprocal ef
1 -2 6. fects of self-concept and performance from a multi
Kernis, M . H. (200 5 ). Measuring self-esteem in con dimensional perspective: Beyond seductive pleasure
text: The importance of stability of self-esteem in and unidimensional perspectives. P ersp ectiv es on
psychological functioning. J o u r n a l o f P erson ality , P s y c h o lo g ic a l S cien ce, I, 1 3 3 - 1 6 3 .
73, 1-37. M ar sh , H. W., 8c Hattie, J. (1996). Theoretical per
Kirkpatrick, L. A., 8c Ellis, B. J. (2001). An spectives on the structure of self-concept. In B.
evolutionary-psychological perspective on self- es Bracken (Ed.), H a n d b o o k o f s e lf-c o n c e p t (pp. 3 8 -
teem: Multiple domains and multiple functions. In 90). New York: Wiley.
G. J. O. Fletcher 8c M. S. Clark (Eds.), B la c k w e ll Marsh, H. W., 8c Shavelson, R. (1985). Self-concept:
h a n d b o o k o f s o c ia l p s y c h o lo g y : Vol. 2. I n t e r p e r Its multifaceted, hierarchical structure. E d u c a tio n
s o n a l p r o c e s s e s (pp. 4 1 1 - 4 3 6 ) . Oxford, UK: Black- a l P sy ch o lo g ist, 2 0 , 1 07 -1 2 5 .
well. McFarlin, D. B. (1985). Persistence in the face of fail
Koole, S. L., Dijksterhuis, A., 8c van Knippenberg, A. ure: T he impact of self-esteem and contingency in
(2001). W h at’s in a name: Implicit self-esteem and formation. P erson ality a n d S o c ia l P sy c h o lo g y B u l
the automatic self. J o u r n a l o f P ers o n a lity a n d S o cia l letin , I I , 1 5 3 - 1 6 3 .
P sy ch olog y , 8 0 , 6 6 9 - 6 8 5 . McFarlin, D. B., Baumeister, R. F., 8c Blascovich, J.
Kuiper, N. A., 8c Rogers, T. B. (1979). Encoding of (1984). On knowing when to quit: Task failure,
personal information: Self-other differences. J o u r self-esteem, advice, and nonproductive persistence.
n a l o f P erson ality a n d S o c ia l P sy ch olog y , 37, 4 9 9 - J o u r n a l o f P erson ality , 5 2 , 1 3 8 - 1 5 5 .
514. McGuire, S., Manke, B., Saudino, K. J ., Reiss, D., Het-
Kwan, V. S. Y., 8c Mandisodza, A. N. (20 07 ). Self herington, E. M ., 8c Plomin, R. (1999). Perceived
esteem: On the relation between conceptualization competence and self-worth during adolescence: A
and measurement. In C. Sedikides 8c S. Spencer longitudinal behavioral genetic study. C h ild D ev e l
(Eds.), F ron tiers in s o c ia l p s y c h o lo g y : T h e s e lf o p m e n t, 70, 1 2 8 3 - 1 2 9 6 .
(pp. 2 5 9 - 2 8 2 ) . Philadelphia: Psychology Press. Mead, G. H. (1934). M in d, s e l f a n d s o c iety . Chicago:
Leary, M . R. (1999). The social and psychological im University of Chicago Press.
portance of self-esteem. In R. M. Kowalski 8c M. Mikulincer, M. (1995). Attachment style and the men
R. Leary (Eds.), T h e s o c ia l p s y c h o lo g y o f e m o tio n a l tal representation of the self. J o u r n a l o f P erson ality
a n d b e h a v io r a l p r o b le m s : In te r fa c e s o f s o c ia l a n d a n d S o c ia l P sy ch olog y , 69, 1 2 0 3 - 1 2 1 5 .
c lin ic a l p s y c h o lo g y (pp. 1 9 7 -2 21 ). Washington, M o rf, C. C., 8c Rhodewalt, F. (2001). Unraveling the
DC : American Psychological Association. paradoxes of narcissism: A dynamic self-regulatory
Leary, M. R., 8c Baumeister, R. F. ( 2 0 00 ). The nature processing model. P s y c h o lo g ic a l In q u iry , 1 2, 1 7 7 -
and function of self-esteem: Sociometer theory. In 196.
M. Zanna (Ed.), A d v a n ce s in e x p e r im e n ta l s o c ia l Mo rgan, H. J ., 8c Jano ff-Bulman, R. (1994). Positive
p s y c h o lo g y (Vol. 3 2 , pp. 1 -6 2 ). San Diego, CA: and negative self-complexity: Patterns of adjust
Academic Press. ment following traumatic versus non-traumatic life
Leary, M. R., Tambor, E. S., Terdal, S. K., 8c Downs, experiences. J o u r n a l o f S o c ia l a n d C lin ic a l P sy ch o l
D. L. (1995). Self-esteem as an interpersonal moni ogy, 13, 6 3 - 8 5 .
tor: The sociometer hypothesis. J o u r n a l o f P ers o n Murray, S. I.., Holmes, J. G., MacDon ald, G., 8c Ells
a lity a n d S o c ia l P sy ch olog y , 6 8 , 5 1 8 - 5 3 0 . w orth, P. (1998). Through the looking glass dark
Leitenberg, H., Yost, L. W., 8c Carroll-Wilson, M. ly?: When self-doubts turn into relationship insecu
3 6. S elf-E steem 545
rities. J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , Sadalla, F.. K., Vershure, B., & Burroughs, J . (1987).
75, 1 4 5 9 - 1 4 8 0 . Identity symbolism in housing. E n v ir o n m en t a n d
Neiss, M . B., Sedikides, C., & Stevenson, J . (2002). B eh av ior, 19, 5 6 9 - 5 8 7 .
Self-esteem: A behavioural genetic perspective. E u Sanna, 1.. J ., & Meier, S. ( 2 00 0). Looking for clouds in
r o p ea n J o u r n a l o f P erson ality , 16 , 3 5 1 - 3 6 7 . a silver lining: Self-esteem, mental simulations, and
Olson, M . A., Fazio, R. FI., & Hermann, A. D., Sr. temporal confidence changes. J o u r n a l o f R esea rch
(20 0 7). Reporting tendencies underlie discrepan in P erson ality , 3 4 , 2 3 6 - 2 5 1 .
cies between implicit and explicit measures of self Sanna, L. J . , Turlev-Ames, K. J., & Meier, S. (20 00 ).
esteem. P sy c h o lo g ic a l S cien ce, 18, 2 6 7 - 2 9 1 . M o od , self-esteem, and simulated alternatives:
Osgood, C. E. ( 1952). The nature and measurement of Thought-provoking affective influences on counter-
meaning. P sy c h o lo g ic a l B u lletin , 4 9, 19 7-2 3 7. factual direction. J o u r n a l o f P erson ality a n d S o cia l
Oyserman, D., Coon , H. M ., & Kemmelmeier, M. P sy ch olog y , 76, 5 4 3 - 5 5 8 .
(2002). Rethinking individualism and collectivism: Savin-Williams, R. C., &c Jaquish, G. A. (1981). The
Evaluation of theoretical assumptions and mera- assessment of adolescent self-esteem: A comparison
analvses. P s y c h o lo g ic a l B u lletin , 1 2 8 , 3 - 7 2 . of methods. J o u r n a l o f P erson ality , 4 9, 3 2 4 - 3 3 6 .
Paulhus, D. 1.., Robins, R. W., Trzesniewski, K. H., & Schafer, R. B., Wickrama, K. A. S., & Keith, P. M.
Tracy, J. L. (20 04 ). Two replicable suppressor situ (1996). Self-concept disconfirmation, psychological
ations in personality research. M u ltiv a riate B e h a v distress, and marital happiness. J o u r n a l o f M ar
io r a l R esea rc h , 3 9 , 3 0 3 - 3 2 9 . ria g e a n d th e F am ily, 5 8 , 167-177.
Pelham, B. W. (1991). On confidence and consequence: Scheff, T. J . , & Fearon, D. S. (200 4 ). Cognition and
The certainty and importance of self-knowledge. emotion?: The dead end in self-esteem research.
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 6 0, J o u r n a l f o r th e T h e o r y o f S o c ia l B eh a v io u r, 3 4 ,
518-530. 73-91.
Pelham, B. W., & Swann, W. B., Jr. (1989). From self- Schroeder, D. G., Josephs, R. A., &c Swann, W. B., Jr.
conceptions to self-worth: The sources and struc (2 00 6 ). F oreg oin g lu crativ e e m p lo y m e n t to p re se rv e
ture of self-esteem. J o u r n a l o f P erson ality a n d S o lo w se lf-e s te e m . Unpublished manuscript.
c ia l P sy ch olog y , 57, 6 7 2 - 6 8 0 . Sedikides, C., Gaertner, I.., & Toguchi, Y. (200 3). Pan-
Pyszczynski, T., Greenberg, J ., Solomon, S., Arndt, cultural self-enhancement. J o u r n a l o f P erson ality
J., & Schimel, J. (2004). Why do people need self a n d S o c ia l P sy ch olog y , 8 4 , 6 0 - 7 9 .
esteem?: A theoretical and empirical review. P sy Sedikides, C., Gaertner, I.., &c Vevea, J. (2005). Pancul-
c h o lo g ic a l B u lletin , 130, 4 3 5 - 4 6 8 . tural self-enhancement reloaded: A meta-analytic
Raimy, V. C. (1948). Self reference in counseling in reply to Heine (2 005). J o u r n a l o f P erson ality a n d
terviews. J o u r n a l o f C o n su ltin g P sy ch olog y , 12, S o c ia l P sy ch olog y , 8 9, 5 3 9 - 5 5 1 .
1 5 3 -1 63 . Setliff, A. E., & Marm urek, H. H. C. (2002). The
Raskin, R. N., & Hall, C. S. (1981). The Narcissistic mood regulatory function of autobiographical re
Personality Inventory: Alternative form reliability call is moderated by self-esteem. P erso n a lity a n d
and further evidence of construct validity. J o u r n a l In d iv id u a l D iffer en c es , 3 2 , 7 6 1 - 7 7 1 .
o f P erson ality a n d S o c ia l P sy ch olog y , 4 5 , 1 5 9 - 1 6 2 . Shavelson, R. J ., Hubner, J. J., & Stanton, G. C. (1976).
Raskin, R. N., Novacek, J ., & Hogan, R. (1991). N a r Validation of construct interpretations. R e v iew o f
cissism, self-esteem, and defensive self-enhancement. E d u c a tio n a l R esea rch , 4 6 , 4 0 7 - 4 4 1 .
J o u r n a l o f P erson ality , 59, 1 9 - 3 8 . Showers, C. J. (1992). Compartmentalization of posi
Rhodewalt, F., Madrian, J. C., & Cheney, S. (1998). tive and negative self-knowledge: Keeping bad ap
Narcissism, self-knowledge organization, and em o ples out of the bunch. J o u r n a l o f P erson ality a n d
tional reactivity: T he effect of daily experiences on S o c ia l P sy ch olog y , 6 2 , 1 0 3 6 - 1 0 4 9 .
self-esteem and affect. P erso n a lity a n d S o c ia l P sy Shrauger, J. S., & Lund, A. K. (1975). Self-evaluation
c h o lo g y B u lletin , 2 4 , 7 5 - 8 7 . and reactions to evaluations from others. J o u r n a l o f
Robbins, S. B., Lauver, K., I.e, H., I.angley, R., & Carl- P erson ality , 4 3 , 9 4 - 1 0 8 .
strom, A. (2004). Do psychosocial and study skill Story, A. L. (1998). Self-esteem and memory for favor
factors predict college outcomes?: A meta-analysis. able and unfavorable personality feedback. P ers o n
P sy c h o lo g ic a l B u lletin , 130, 2 6 1 - 2 8 8 . ality a n d S o c ia l P sy ch o lo g y B u lletin , 2 4 , 5 1 - 6 4 .
Robins, R. W., Hendin, H. M ., & Trzesniewski, K. H. Suls, J. ( 2 0 0 6). On the divergent and convergent valid
(20 01). Measuring global self-esteem: Construct ity of self-esteem. In M. H. Kernis (Fid.), S elf-esteem
validation of a single-item measure and the Rosen issu es a n d a n sw ers: <4 s o u r c e b o o k o f cu rren t p e r
berg Self-Esteem Scale. P erso n a lity a n d S o c ia l P sy sp e ctiv e s (pp. 3 6 - 4 3 ) . New York: Psychology Press.
c h o lo g y B u lletin , 27, 15 1-16 1. Swann, W. B., Jr. (1983). Self-verification: Bringing
Rosenberg, M. (1965). S o c iety a n d th e a d o le s c e n t self- social reality into harmony with the self. In J. Suls
im a g e. Princeton, NJ: Princeton University Press. & A. G. Greenwald (Fids.), P s y c h o lo g ic a l p e r s p e c
Rosenberg, M., & Owens, T. J. (2001). Low self tives o n th e s e l f (Vol. 2 , pp. 3 3 - 6 6 ) . Hillsdale, NJ:
esteem people: A collective portrait. In T. J. Owens, Erlbaum.
S. Stryker, & N. Goo dm an (Fids.), E x te n d in g s e lf Swann, W. B., Jr. (1990). To be adored or to be
es te e m th eo r y a n d re se a r c h : S o c io lo g ic a l a n d p s y known?: The interplay of self-enhancement and
c h o lo g ic a l cu rren ts (pp. 4 0 0 - 4 3 6 ) . New York: self-verification. In FI. T. Higgins & R. M. Sorren
Cambridge University Press. tino (Eds.), H a n d b o o k o f m o tiv a tio n a n d co g n itio n :
Rosenberg, M. R., Schooler, C., Schoenbach, C., & Vol. 2 . F o u n d a tio n s o f s o c ia l b e h a v io r (pp. 4 0 8 -
Rosenberg, F. (1995). Global self-esteem and specific 4 48 ). New York: Guilford Press.
self-esteem: Different concepts, different outcomes. Swann, W. B., Jr. (1996). S elf-trap s: T h e elu siv e q u es t
A m eric a n S o c io lo g ic a l R e v iew , 6 0 , 141 -1 56 . f o r h ig h er s e lf-e stee m . New York: Freeman.
546 V I. S E L F -R E L A T E D D IS P O S IT IO N S
Swann, W. B., Jr., Beavers, C., 8c McGeary, J. (2007). logical structure of pride: A tale of two facets.
G e n e tic in flu e n c es o n se lf-e s te e m . Unpublished J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 9 2 ,
manuscript, University of Texas. 506-525.
Swann, W. B., Jr., Chang-Schneider, C., & McClarty, Trzesniewski, K., Donnellan, B., Moffitt, T., Robins,
K. (20 07 ). Do our self-views matter?: Self-concept R., Poulton, R., & Caspi, A. (20 06 ). Low self
and self-esteem in everyday life. A m er ica n P sy ch o l esteem during adolescence predicts poor health,
og ist, 6 2 , 8 4 - 9 4 . criminal behavior, and limited economic prospects
Swann, W. B., Jr., Chang-Schneider, C., &C McClarty, during adulthood. D e v e lo p m e n ta l P sy ch olog y , 4 2 ,
K. (2008). Yes, cavalier attitudes can have perni 381-390.
cious consequences: A reply to Krueger, Vohs, &C Twenge, J . M ., &c Campbell, W. K. (2001). Age and
Baumeister. A m erica n P sy ch o lo g ist, 6 3 , 6 5 - 6 6 . birth cohort differences in self-esteem: A cross
Swann, W. B., Jr., De La Ronde, C., & Hixon, J. G. temporal meta-analysis. P erso n a lity a n d S o c ia l P sy
(1994). Authenticity and positivity strivings in m ar c h o lo g y R ev iew , 5, 3 2 1 - 3 4 4 .
riage and courtship. J o u r n a l o f P erson ality a n d S o Valentine, J . C., DuBois, D. L., & Cooper, H. (2 004).
c ia l P sy ch olog y , 6 6 , 8 5 7 - 8 6 9 . T he relation between self-beliefs and academic
Swann, W. B., Jr., Griffin, J . J ., Predmore, S. C., & achievement: A meta-analytic review. E d u c a tio n a l
Gaines, B. (1987). The cognitive-affective crossfire: P sy ch o lo g ist, 3 9, 1 11 -1 3 3 .
When self-consistency confronts self-enhancement. Verschueren, K., Marco en, A., & Schoefs, V. (1996).
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 5 2 , The internal working model of the self, attachment,
881-889. and competence in five-year-olds. C h ild D e v e lo p
Swann, W. B., Jr., & Pelham, B. W. (2002). W h o wants m en t, 6 7 , 2 4 9 3 - 2 5 1 1 .
out when the going gets good?: Psychological invest Waters, E'., Noyes, D. M ., Vaughn, B. E., & Ricks, M.
ment and preference for self-verifying college room (1985). Q-sor t definitions of social competence and
mates. S e l f a n d Id en tity , 1, 2 1 9 - 2 3 3 . self-esteem: Discriminant validity of related co n
Swann, W. B., Jr., Pelham, B. W., & Krull, D. S. (1989). structs in theory and data. D e v e lo p m e n ta l P sy ch o l
Agreeable fancy or disagreeable truth?: Reconciling og y , 2 1 , 5 0 8 - 5 2 2 .
self-enhancement and self-verification. J o u r n a l o f Watson, D., Suls, J., & Haig, J. (2002). Global self
P erson ality a n d S o c ia l P sy ch o lo g y , 57, 7 8 2 - 7 9 1 . esteem in relation to structural models of personal
Swann, W. B., Jr., & Read, S. J. (1981). Self-verification ity and affectivity. J o u r n a l o f P erson ality a n d S o c ia l
processes: How we sustain our self-conceptions. P sy ch o lo g y , 8 3 , 1 8 5 - 1 9 7 .
J o u r n a l o f E x p e r im e n ta l S o c ia l P sy ch olog y , 17, Webster, G. D. ( 2 0 0 7). Is the relationship between
351-372. self-esteem and physical aggression necessarily U-
Swann, W.B., Jr., Rentfrow, P. J ., & Guinn, J. (2 003). shaped? J o u r n a l o f R e se a rc h in P erson ality , 41,
Self-verification: The search for coherence. In M. R. 977-982.
Leary & J. P. Tangney (Eds.), H a n d b o o k o f s e l f a n d Westen, D. (1990). The relations among narcissism,
id en tity (pp. 3 6 7 - 3 8 3 ) . New York: Guilford Press. egocentrism, self-concept, and self-esteem: Experi
Tafarodi, R. W. (1998). Paradoxical self-esteem and mental, clinical, and theoretical considerations.
selectivity in the processing of social information. P sy ch o a n a ly sis a n d C o n te m p o r a r y T h o u g h t, 13,
J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 74, 183-239.
1181-1196. Wheeler, L., &c Miyake, K. (1992). Social comparison
Tafarodi, R. W., &c Milne, A. B. (200 2). Decompos in everyday life. J o u r n a l o f P erson ality a n d S o cia l
ing global self-esteem. J o u r n a l o f P erson ality , 70, P sy ch olog y , 6 2 , 7 6 0 - 7 7 3 .
443-483. Wiesenfeld, B. M ., Swann, W. B., Jr., Brockner, J.,
Tafarodi, R. W., & Swann, W. B., Jr. (1996). & Bartel, C. (20 0 7). Is more fairness always pre
Individualism-collectivism and global self-esteem: ferred?: Self-esteem moderates reactions to proce
Evidence for a cultural trade off. J o u r n a l o f C ross- dural justice. A c a d e m y o f M a n a g e m en t J o u r n a l,
C u ltu ral P sy ch olog y , 27, 6 5 1 - 6 7 2 . 50, 1235-1253.
Tafarodi, R. W., & Swann, W. B., Jr. (2001). Two- Woolfolk, R. L., Novalany, J ., Gara, M . A., Allen,
dimensional self-esteem: Theory and measure L. A., &c Polino, M . (1995). Self-complexity, self-
ment. P erso n a lity a n d In d iv id u a l D iffe r e n c e s , 31, evaluation, and depression: An examination of form
653-673. and content within the self-schema. J o u r n a l o f P er
Tice, D. M ., & Baumeister, R. F. (1990). Self-esteem, so n a lity a n d S o c ia l P sy ch olog y , 6 8 , 1 1 0 8 - 1 1 2 0 .
self-handicapping, and self-presentation: The strat Wray, I.. D., & Stone, E. R. (2005). The role of self
egy of inadequate practice. J o u r n a l o f P erson ality , esteem and anxiety in decision making for self ver
58, 4 4 3 -4 6 4 . sus others in relationships. J o u r n a l o f B eh a v io r a l
Tracy, J. L., & Robins, R. W. (20 07 ). The psycho D ecisio n M akin g , 18, 1 2 5 - 1 4 4 .
CHAPTER 37
Narcissism
F r e d e r ic k R h odew alt
B e n ja m in Pe t e r s o n
547
548 V I. S E L F -R E L A T E D D IS P O S IT IO N S
berg (1975), and Kohut (1971), however, ate an actual self that is close to these self
made the most substantial contributions in standards in order to experience high self-
terms of placing the construct of narcissism regard. Reich described behaviors such as
within the lexicon of psychiatry and clini compensatory self-inflation, aggression, and
cal psychology. Their positions are elaborate overdependence on social approval as part
and complex and differ from one another of the narcissist’s repertoire of self-esteem
in significant ways. A full description of the regulation behaviors. In brief, narcissists
Freud, Kernberg, and Kohut writings on embrace grandiose self-images to compen
narcissism is beyond the scope of this chap sate for deficits in their early interpersonal
ter, and the reader is referred to an edited experiences and spend their lives managing
volume by M orrison (1986), which contains the problem of grandiose but fragile self
many of the original sources and several ex esteem resulting from the pursuit of such
cellent overviews. self-ideals.
Here we extract what we view as their
essential ideas, those that have formed the
foundation for contemporary work on the Definition and O perationalization
narcissistic personality type. Collectively,
Freud’s, Kernberg’s, and Kohut’s central the Despite the richness of clinical theory and
oretical positions described the narcissist as description, they did not generate systematic
an individual who, as a result of a history of research on narcissism because there was
unsatisfactory social relationships, possesses little consensus among theorists on the de
a grandiose and defensive self-concept that fining features of narcissism. Consequently,
contains a conflicted psychological depen there was no consistently accepted manner of
dence on other people. Rhodewalt and M orf assessing the presence (or degree) of narcis
(2005) characterized psychoanalytic views sism that would allow the empirical study of
on narcissism as agreeing that adult nar the thoughts and behaviors of such individu
cissism results from a childhood history of als. Thus the field was left with a body of
problematic interpersonal relationships. As work that was highly theoretical and based
a consequence, narcissistic adults appear to mainly on clinical description.
possess a grandiose self-concept and invul This situation changed with the publica
nerability on the outside but emptiness and tion of the third edition of the Diagnostic
isolation on the inside. This combination and Statistical M anual o f M ental D isorders
leads to a highly conflicted overdependence (D SM -III, American Psychiatric Association,
on others to maintain self-esteem. 1980), which organized clinical description
Perhaps the theorist who best bridges clas into a working “definition” of narcissism
sic psychoanalytic and contemporary views with criteria that could be used for assess
on narcissism is Annie Reich (1960). Her ment and research. In this edition of the
discussion of narcissism as pathological self D SM -III, narcissism was included as an Axis
esteem regulation anticipates recent work II personality disorder. The specific criteria
on social-cognitive self-regulatory models that were ultimately adopted reflected the
of narcissism (M orf & Rhodewalt, 2 0 0 1 ; major influence of the theoretical models of
Rhodewalt & M orf, 2005). In accord with Kernberg (1975) and Kohut (1971).
the psychoanalytic views of the time, Reich As is the tradition for all DSM disorders,
described narcissism as an overinvestment an individual is considered to be a “narcis
of libido in the self at the expense of invest sist” (classified as having narcissistic person
ment in others (objects). Reich took a con ality disorder [NPD]) if he or she meets a
temporary view of self-esteem, stating that certain level of the criteria set forth. In other
it is the harmony or discrepancy between words, there is a certain cutoff point be
one’s self-representations and one’s wishful tween individuals who are narcissistic and
concept of the self— in other words, a dis others who fall just shy, with a qualitative
crepancy between actual and ideal selves (cf. difference assumed between the two groups
Higgins, 1987). Given that narcissists’ ideal (see Foster & Campbell, 2007, for further
self-images are grandiose and unrealistic, discussion of the taxonic vs. dimensional de
they must go to extraordinary lengths to cre scription of narcissism). The initial criteria
37. N arcissism 549
included in the D SM -III for narcissism were scales that were based on these criteria—
as follows: which, in turn, stimulated research on the
construct. Some scales were based on exist
(1) Grandiose sense o f self-importance and ing large-scale personality inventories such
uniqueness, e.g., exaggeration o f achieve as the Minnesota Multiphasic Personal
ments and talents, focus on the special nature ity Inventory (M M PI) (e.g., Ashby, Lee, &
of one’s problems. Duke, 1979; Morey, Waugh, & Blashfield,
(2) Preoccupation with fantasies o f unlimited su c
cess, power, brilliance, beauty, or ideal love.
1985; Wink & Gough, 1990), and other new
(3) Exhibitionism: the person requires constant scales were subsequently created (e.g., the
attention and adm iration. M argolis-Thom as Measure of Narcissism—
(4) Cool indifference or marked feelings of rage, Mullins & Kopelman, 1988; the M illon
inferiority, shame, humiliation, or emptiness Clinical M ultiaxial Inventory narcissism
in response to criticism, indifference o f oth subscale— Millon & Davis, 1997), but the
ers, or defeat. one that has received by far the most em
(5) At least two of the following characteristic of pirical attention is the Narcissistic Person
disturbances in interpersonal relationships: ality Inventory (NPI; Raskin & H all, 1979,
a. Entitlement: expectation o f special favors 1981). Using the DSM -III criteria directly
without assum in g reciprocal responsibili as a guide, Raskin and Hall (1979) origi
ties, e.g., surprise and anger that people
nally developed a 54-item scale that used a
will not do what is wanted;
forced-choice format, such that the respon
b. Interpersonal exploitativeness: taking a d
vantage of others to indulge own desires or dent had to choose the one of two statements
for self-aggrandizement; disregard for the that was most self-descriptive. Unlike some
personal integrity and rights of others; of the other scales that were developed, the
c. Relationships that characteristically alter stated purpose of the NPI was to assess nar
nate between the extremes o f overidealiza cissism in the subclinical general population
tion and devaluation; (as a personality “trait”). Thus narcissism
d. Lack of empathy: inability to recognize was seen as falling on a general continuum
how others feel, e.g., unable to appreciate of personality such that only extreme lev
the distress of someone who is seriously ill. els may be considered to characterize NPD
(American Psychiatric Association, 1980) in the clinical sense. Even though certain
elevated “subthreshold” narcissism scores
This definition clearly emphasizes the nar may not be considered as constituting NPD
cissist’s characteristic grandiosity in relation in a clinical sense, high levels of subclinical
to his or her self-image, and this emphasis narcissism should promote similar outcomes
has continued through subsequent revi and problems as does NPD.
sions to the criteria (D SM -III-R [American A great deal of research has been con
Psychiatric Association, 1987], DSM -IV ducted on the construct validity, reliability,
[American Psychiatric Association, 1994], internal consistency, and factor structure of
DSM -IV -TR [American Psychiatric Associa the NPI. From the original 54-item version,
tion, 2000]). It is this grandiosity aspect of two commonly used versions have been put
narcissism that most effectively distinguish forth, each paring off several of the original
es it from other “Cluster B ” personality dis items and proposing varying factor struc
orders (such as histrionic, antisocial, and tures (Emmons, 1987; Raskin & Terry,
borderline personalities) when DSM criteria 1988). Emmons’s (1984, 1987) 37-item
are evaluated for their ability to categorize version comprises four underlying factors:
disorders properly (Gunderson, Ronning- Leadership/Authority, Self-Absorption/Self-
stam, &c Smith, 1991). Such distinctions are Admiration, Superiority/Arrogance, and
important for psychiatric diagnosis, though Exploitativeness/Entitlement. Using this ver
this process may exclude important features sion, Emmons and others have argued that
of the construct that are potentially impor exploitativeness and entitlement may repre
tant to other researchers. sent the most maladaptive aspects of NPI-
The establishment of criteria for NPD was based narcissism. Raskin and Terry (1988)
important in many ways, most significantly also revised and improved on the original
in facilitating the development of self-report NPI with a 40-item version that could be
550 V I. S E L F -R E L A T E D D IS P O S IT IO N S
ment, employing interpersonal strategies de Sun, 1994), and self-handicapping (Rhode
signed to manipulate impressions that others walt et al., 2006).
hold of them, as well as the feedback they M orf and Rhodewalt (2001) also included
receive from others. They also are active the narcissist’s social relationships and con
with regard to the intrapersonal strategies texts as an element in the model. Narcissists
they employ, distorting and biasing their have an impact on their social worlds through
interpretations of outcomes and selectively their actions, interpretations, and choices of
recalling past events in self-enhancing ways. interaction partners, so their social environ
The dynamic aspect of the model is cap ments are objectively different from those of
tured in its recursive quality. Narcissistic less narcissistic individuals. Relationships
self-esteem regulation is shaped and molded and social contexts are attractive to nar
by ongoing and changing self-concerns and cissists to the extent that they provide self
social contexts. In turn, narcissistic individ enhancement opportunities, and it appears
uals behave in accord with their current con that they manage these with at least some
cerns about self-definition, and these behav short-term success. Paulhus (1998) reported
iors influence the social context. The social that narcissists are successful in garnering
context provides affordances for addressing admiration and positive regard early in their
issues involving self-concept and self-worth relationships but that, over time, these same
by highlighting, intensifying, or redirect interpersonal tactics result in rejection and
ing attention to specific aspects of the self. hostility. The reason is that interaction part
Rhodewalt and M orf (2005) argued that the ners come to see the narcissists as acting in
narcissistic self is context bound and that ways that are self-promoting and aggrandiz
transitions from one social context to an ing while also putting others down (Buss &
other lend to the fragility and vulnerability Chiodo, 1991).
of such individuals’ self-views. The final element in the model is the nar
The model has been helpful in organizing cissistic self-concept. Akhtar and Thomson
research findings and generating new hy (1982) summarized the clinical literature as
potheses. The research is too extensive to re indicating that narcissists’ public grandiosity
view here (see Rhodewalt & Sorrow, 2003), overlays an underlying fragility and feelings
so we highlight some key examples of the of worthlessness. A challenge that has clear
intra- and interpersonal self-regulation in implications for the assessment of narcissism
which narcissists engage. Perhaps the clear is to understand the coexistence of grandios
est example of narcissistic intrapersonal self ity and vulnerability within the narcissistic
regulation is their pronounced tendency to self-concept.
make self-aggrandizing attributions for pos Following the publication of the M o rf-
itive outcomes even when those outcomes Rhodewalt model, others have offered ex
are response noncontingent (Rhodewalt & tensions or alternatives to the dynamic
M orf, 1995, 1998). In line with Jones and self-regulatory processing framework.
Berglas’s (1978) reasoning about the genesis Baumeister and Vohs (2001) offered the
of self-handicapping, this self-aggrandizing analogy of addiction to describe narcissists’
attributional style in all likelihood contrib pursuit of self-esteem. They suggested that
utes to narcissists’ positive but fragile self narcissists crave feelings of superiority and
conceptions, as well as to their emotional approval, build up a tolerance to such feed
and interpersonal reactions to threats to back (thus requiring more and more), and
the self. Rhodewalt and colleagues (Rhode exhibit withdrawal (distress) when they fail
walt & Morf, 1998; Rhodewalt, Tragakis, to receive what they crave. This perspective
& Finnerty, 2 0 0 6 ) contend that narcissists is descriptive but not yet subjected to empiri
claim identities that, although highly posi cal test. In our view, it is a way of restating
tive, are uncertain and easily threatened. the self-regulation model and may not be
This unspoken uncertainty about the self lies substantially different in the final analysis.
behind the narcissist’s apparently insatiable More recently, Campbell and Foster
drive for admiration and regard from others. (2007) proposed the extended-agency model.
Interpersonal strategies include aggrandiz They noted that narcissists have positive self-
ing self-presentations (M orf, 1994), deroga images that are based on admiration and suc
tion of those who threaten them (Kernis & cess rather than on social acceptance and ap
552 V I. S E L F -R E L A T E D D IS P O S IT IO N S
proval (Raskin, Novacek, & Hogan, 1991). projected onto the circumplex, however,
Additionally, narcissists are not invested in Leadership/Authority and Self-Absorption/
warm communal relationships, and their Self-Admiration were associated with domi
self-regulatory strategies center on attempts nance but not low affiliation, Superiority/
to make themselves look powerful and com Arrogance was associated with both domi
petent. Campbell and Foster argued that, col nance and low affiliation, and Exploitative
lectively, these characteristics are accounted ness/Entitlement was associated with domi
for by the narcissist’s high agency and low nance, low affiliation, and high neuroticism.
communion. An interesting contribution is Ruiz and colleagues suggested caution in the
their notion that narcissism waxes and wanes interpretation of composite scores that may
as a function of whether self-esteem needs contain distinct personality characteristics.
are currently satisfied. That is, narcissism is That caution notwithstanding, in our view,
more state-like than trait-like. One corollary reducing narcissism to a combination of ex
of their model is that narcissism is not driven traversion (or dominance) and low agree
by the pursuit of an overarching goal such as ableness (or low affiliation) does not account
self-enhancement or self-esteem regulation. for the specific patterns of motivation and
Rather, narcissists’ goals are context depen self-regulatory behaviors indicated in the lit
dent. This is a provocative idea that contrasts erature. That is, narcissists may be high in
with the Rhodew alt-M orf assumption that extraversion and low in agreeableness, but
protecting, maintaining, and enhancing the this description does not account for their
self are chronically accessible goals that may being so sensitive to threats to the self and
be fulfilled polymorphously in any social relentless in seeking admiration.
context that affords the opportunity to ex In a meta-analysis of studies that included
perience positive self-esteem. Distinguishing measures of narcissism and measures of im
between these two models would provide an pulsivity, Vazire and Funder (2006) reported
important advance in the understanding of a mean effect size of r = .34 across 10 in
self-esteem regulation. This discussion may dependent samples and concluded that im
also shed light on gender differences in the pulsivity is a defining characteristic of nar
incidence of narcissism observed in clinical cissism. Impulsivity, they contend, accounts
populations. The higher incidence of NPD for much of narcissists’ self-enhancing as
among men may reflect compatibility be well as self-defeating behavior. Vazire and
tween narcissism and the agentic features of Funder suggested that impulsivity should
the male gender role. Narcissism in women be included in self-regulatory models of
may be masked by the fact that narcissism is narcissism, and we agree, with two impor
not compatible with the expression of com tant caveats. First, the correlations between
munion, the hallmark of the female gender NPI-defined narcissism and impulsivity, ex
role. traversion, and low agreeableness are sig
Several theorists have taken the approach nificant but modest, and many people who
that narcissism is best accounted for through possess these dispositions would not be con
its association with certain dispositions. sidered narcissists. Second, we suspect that
Paulhus (2001) proposed that the combina the linkages between narcissism and these
tion of extraversion and low agreeableness dispositions reflect their common associa
(high antagonism) is equivalent to narcis tions with temperament; all have significant
sism. Evidence suggests that narcissism oc heritability quotients. Thus a child who is
cupies the high extraversion/low agreeable temperamentally disposed toward extraver
ness location in “Big Five” space (Wiggins sion, impulsivity, and/or low agreeableness
& Pincus, 1989). Ruiz, Smith, and Rhode is more likely to become narcissistic given
walt (2001) projected measures of hostil a set of parent-child interactions than is an
ity and narcissism onto the interpersonal introverted, agreeable, well-controlled child
circumplex (IASR-B5; Trapnell & Wiggins, placed in the same socialization context.
1990) and found that both hostility and nar In sum, a number of contemporary models
cissism were associated with low affiliation of narcissism draw on either social-cognitive
(low agreeableness) but that only narcissism processing constructs (M orf & Rhodewalt,
was associated with high dominance (ex 2001) or dispositional models (Paulhus,
traversion). When the NPI subscales were 20 0 1 ; Vazire & Funder, 2 0 0 6 ) to account
37. N arcissism 553
for the psychological features that underlie tial feature. The NPI was developed using
the construct. This work largely relies on this definition, and, as noted, the NPI has
narcissism as assessed by the N PI, which, in been used in the vast majority of research
turn, is based on clinical definitions of nar on narcissism within personality and so
cissism (D SM -IV -TR; American Psychiatric cial psychology. Thus this research is based
Association, 200 0 ). The research is exten on an interpersonal type of narcissism that
sive and supports the view that narcissism Akhtar and Thomson (1982) labeled overt
is fundamentally a pattern of interpersonal narcissism. Overt narcissists openly seek
self-esteem regulation. However, pieces of admiration and use their interpersonal rela
the puzzle require further explanation and tionships for self-aggrandizement.
elaboration. One central question is wheth Although this conceptualization has pro
er these models characterize all narcissists duced a large and coherent research base,
or whether there are different forms of the it may tell only half the narcissism story,
characteristic. because there may be a second “type” of
narcissist, one that was also described by
Kernberg (1975) and Kohut (1971). This
Building on N PI D escription narcissist has been variously termed covert
and Self-R egulatory Models: (Akhtar & Thomson, 1982; Kernberg, 1975;
Som e Integrative Issues W ink, 1991), vulnerable (Dickinson & Pin
cus, 2 0 0 3 ; W ink, 1991), hypersensitive
Personality and social psychology research (Hendin & Cheek, 1997), and hypervigilant
ers have successfully taken the construct of (Gabbard, 1998), as the individual tends to
narcissism as described in the psychoanalytic be reserved and withdrawn interpersonally
literature and translated it into the narcissis while harboring some elements of charac
tic personality type, an individual difference teristic narcissistic expectations for self and
in dynamic, defensive self-regulation found others (e.g., entitlement). In other words, the
in nonclinical populations. This burgeon grandiose type of narcissist does not hide his
ing literature relies almost exclusively on the or her narcissistic expectations from others
NPI as its measure of narcissism. Although (he or she is “overt”), whereas the vulner
the overall characterization of narcissism as able type of narcissist may present in a dif
pathological self-esteem regulation is both ferent, more modest manner (he or she keeps
compelling and consistent with clinical re his or her grandiosity needs more “covert”).
search and theory, puzzling inconsistencies Because of this difference, some clinical
and paradoxes raise difficult questions and theorists and practitioners believe that the
calls for further refinement of narcissism narcissist captured by the DSM criteria and
theory and measurement. For example, re the NPI (and thus most social-personality
searchers are debating the true nature of research that uses the NPI) is incongruent
narcissism: Should it be considered a uni with what they see in therapy (see Cain, Pin
tary or multidimensional construct? Does cus, & Ansell, 2 0 0 8 , for a review).
it contain both adaptive and maladaptive There is support for the position that nar
aspects? How might potentially discrepant cissism occurs in several types. For example,
“inner” and “outer” self-beliefs be assessed Wink (1991) examined several M M PI nar
and incorporated into theory? W hat needs cissism scales, including some that focus
to be assessed to capture the possible multi on grandiose aspects and others that focus
faceted nature of narcissism? In this section, on more sensitive aspects. In a factor anal
we address the current state of some of these ysis of the six scales, Wink found that the
debates and provide some analysis and inte two sets loaded on separate factors, which
gration in the hope of organizing and stimu he called Grandiosity-Exhibitionism and
lating future research. Vulnerability-Sensitivity. Importantly, the
two factors had very distinct patterns of
correlates. Whereas the grandiose pattern
Is T h ere M ore T h a n O ne Type
resembled much of what has been discussed
o f Narcissism ?
up to this point (i.e., dominance, assertive
The DSM definition of narcissism (developed ness, aggressiveness), the vulnerable pattern
for NPD) focuses on grandiosity as its essen was one of “defensive, hypersensitive, anx
554 V I. S E L F -R E L A T E D D IS P O S IT IO N S
ious, and socially reticent individuals whose question. First, the NPI is based on the DSM
personal relations, however, were marked definition of narcissism, which recognizes
by self-indulgence, conceit, and arrogance, that pathological narcissism is a syndrome
and an insistence on having their own way” of characteristics. Would a person who has
(W ink, 1991, p. 596). More recent research a sense of leadership and authority in the ab
by Dickinson and Pincus (2003) found simi sence of entitlement, exploitativeness, supe
lar patterns using the NPI and its subscales. riority, arrogance, self-absorption, and self
At the clinical level, Fossati and colleagues admiration be categorized as a narcissist?
(2005) found evidence for two clusters of (We think not.) Second, in research using
symptoms reflecting overt and covert expres the NPI, the subscales are always correlated.
sions in a structural examination of DSM- Thus a person categorized as a narcissist is,
IV NPD criteria. Thus there appears to be in all likelihood, scoring somewhat highly
support for more than one type of narcis on all subscales. It is important to develop
sism and for the idea that the overwhelming assessment instruments that allow research
focus on grandiose aspects does not capture ers to distinguish between overt and covert
the full nature of the construct. narcissism, but we do not believe that de
The picture of the covert, vulnerable, and composing the NPI will prove to be the most
hypersensitive narcissist is one of a “seeth useful means of doing so.
ing” individual who expects a great deal
from him- or herself and the world but who
A re “ N o rm a l” Narcissists
has not experienced a great deal of confirma
M ore Psychologically H ealthy?
tion of those expectations. As a result, this
person becomes anxious and depressed and Research has generally found that narcis
withdraws from social interaction, out of sists report that they possess many positive
fear of further narcissistic injury in the form psychological resources, such as self-esteem,
of lack of admiration and rejection from happiness, optimism, and life satisfaction,
others. Even though this picture is in stark as well as a lack of negative psychological
contrast to that of its more overt and gran outcomes, such as depression, loneliness,
diose counterpart, the two types do share anxiety, and neuroticism. Although we do
some commonalities. M ost prominently, as not find these results surprising, the more
sessments of covert narcissism tend to cor interesting question may be, W hat do they
relate to a greater extent with overt (e.g., truly tell us about narcissism and its adap
NPI) elements involving entitlement and ex tive versus maladaptive features?
ploitativeness (Emmons, 1987). Dickinson Sedikides, Rudich, Gregg, Kumashiro,
and Pincus (2003), citing consistencies with and Rusbult (2004) suggested that such find
clinical theory (e.g., Akhtar & Thomson, ings tell us that self-esteem is the key to the
1982) and empirical research (e.g., Emmons, overall psychological “health” of narcissists.
1987), proposed that entitlement may be the Similarly, Rose and Campbell (2004) point
“core element” of narcissism that contrib ed to the appetitive goals that narcissists pur
utes to difficulties in regulating self-esteem. sue, combined with their subjective apprais
Thus overt/grandiose and covert/vulnerable als that they are successfully negotiating
narcissism may be different manifestations their interpersonal environments in pursuit
of this underlying core. Future research of such goals. We agree that narcissists dis
should look for ways of recognizing these play a positive bias in self-related accounts
two “types,” as well as attempting to inte but question whether this indicates positive
grate their core elements. mental health. First, the pattern of hyper
Researchers are often urged to report rela responsiveness to threats to the self belies a
tions between NPI subscales and dependent secure, confident, and mentally healthy in
variables, as well as the relations between dividual. Furthermore, narcissists’ self-views
the NPI total score and these outcomes. are quite positive, but evidence shows that
The logic is that such analyses will allow their self-evaluations are inflated compared
researchers to identify the “toxic” compo with objective reality (Gabriel, Critelli, &
nent of narcissism and perhaps reveal differ Ee, 1994; John & Robins, 1994; Rhode
ent forms of narcissism, but we are not cer walt & Eddings, 2 0 0 2 ; Rhodewalt &c Morf,
tain that this practice addresses the subtype 1998). Finally, many of the empirical find
37. N arcissism 555
ings relating narcissism to maladaptive out lation system, one that is sensitive to contex
comes (aggression, hostility, and defensive tual demands and threats.
self-aggrandizement) statistically control for So narcissists, especially those who may
self-esteem (e.g., Bushman & Baumeister, be characterized as overt and grandiose,
1998; M orf & Rhodewalt, 1993; Rhodewalt display high self-esteem and apparent psy
& M orf, 1998). The evidence thus suggests chological adjustment, but is this a tenuous
that although narcissists’ self-descriptions state they find themselves in? Even if such
appear to reflect a mentally healthy person, individuals are “better off” than those low
the coexistence of vulnerability calls into in narcissism on certain occasions, are they
question the proposition that “normal nar potentially setting themselves up for a hard
cissism” involves high self-esteem and men fall? Are the patterns of maladjustment that
tal health. Again, this debate suggests that we see in covert and vulnerable narcissists
more refined assessment methods are needed not far removed from the happier existence
to capture the juxtaposition of grandiosity of the overt and grandiose narcissist? The
and vulnerability that defines narcissism. grandiose and inflated self-views that pro
On the other side, whereas overt and mote perceptions of psychological adjust
grandiose narcissists tell us that they are ment may also leave the individual more
doing well, covert and vulnerable narcissists open to difficulties that threaten self-esteem
display poor psychological health, reporting and well-being. As noted previously, (overt)
low self-esteem and high rates of depres narcissists are self-aggrandizing in their at
sive and anxious symptomology, among tributions for positive outcomes, even out
other indicators (e.g., Dickinson & Pincus, comes that they in fact did not produce.
2 0 0 3 ; Rose, 2 0 0 2 ; Watson, Sawrie, Greene, Although this tendency to self-aggrandize
& Arredondo, 2 0 0 2 ; W ink, 1991). Why do appears to reflect positive mental health, it
overt narcissists seem to be so “healthy” backfires when the narcissist cannot sustain
psychologically whereas covert narcissists success (e.g., Rhodewalt & Morf, 1998).
appear considerably less so? These findings suggest that the propensity
In an attempt to reconcile this question, to describe oneself positively contributes to
Watson and his colleagues proposed a “con the emotional instability that is one of the
tinuum hypothesis” that places narcissism hallmarks of covert narcissism.
at varying levels of adjustment along a con
tinuum related to self-esteem (e.g., Watson,
A re A ll Narcissists D efending
Little, Sawrie, & Biderman, 1992; Watson
against Underlying Vulnerabilities?
et al., 2002). This continuum ranges from
purely maladaptive narcissism at the “un An additional question concerns whether
healthy” pole followed by an overlap be grandiosity and vulnerability are actual
tween maladaptive and adaptive narcissism ly found in the same individuals. In other
to more adaptive narcissism alone followed words, are overt and covert narcissists
by an overlap between more adaptive narcis qualitatively different from one another at
sism and healthy self-esteem and finally to the level of subjective experience and de
a fully healthy self-esteem at the “healthy” fensiveness? The work of Watson and his
pole (Watson et al., 2 0 0 2 , p. 86). Watson and colleagues (e.g., 2002) with the continuum
colleagues (2002) found that the maladap hypothesis is a step in the direction of inte
tive part of the continuum (greater covert grating the two “types” as opposite ends of
narcissism) is strongly related to depression, the same spectrum, but this idea needs to go
whereas healthy self-esteem is negatively re further to discuss the common basis of the
lated to depression (with the overt narcissists two. Whereas covert narcissists have vulner
on the more “adaptive” part in between). abilities that are displayed through a com
Furthermore, these researchers suggest that bination of a sense of entitlement and ex
one’s position along this continuum should plicitly stated feelings of worthlessness, the
not be considered stable; rather, situational vulnerabilities of overt narcissists include
changes and changes in life circumstances entitlement and hypersensitivity to threat.
can lead to movement up or down the con Still, these vulnerabilities may be another
tinuum of narcissism. This is consistent with common theme that binds the two forms to
the view of narcissism as a self-esteem regu gether, with overt and grandiose narcissists
556 V I. S E L F -R E L A T E D D IS P O S IT IO N S
simply more effective at defending against to how secure and confidently held these
these concerns. grandiose opinions are (e.g., Kernis, 2003).
The recent conceptualization and devel In a series of daily diary studies, Rhodewalt
opment of techniques for assessing implicit and his colleagues (Rhodewalt, M adrian, &
cognition and automatic evaluations (e.g., Cheney, 1998) had participants high and low
Greenwald & Banaji, 1995) have been ex in NPI provide daily descriptions of events
tended to the construct of self-esteem. Thus and state self-esteem across a number of
researchers can explore discrepancies be days. Narcissists not only displayed greater
tween explicit and implicit self-esteem as a day-to-day fluctuations in their self-esteem
way to track narcissistic vulnerability. Con than less narcissistic individuals, but their
sistent with this hypothesis, Jordan, Spen self-esteem was also more strongly related
cer, Zanna, Hoshino-Browne, and Correll to the quality of their social interactions
(2003) showed that the combination of high than it was for less narcissistic individuals.
explicit (self-reported) and low implicit self In particular, narcissists’ daily self-esteem
esteem significantly predicted higher scores was more highly correlated with the extent
on the NPI (see also Ziegler-Hill, 2006). to which the day’s social interactions were
Other research has shown that high explic perceived as positive or negative, the extent
it-low implicit individuals are extremely de to which the interactions made them feel like
fensive and self-serving, as well as reactive themselves, and, surprisingly, the extent to
to threats to self-esteem, similar to narcis which they felt accepted by the audience.
sists (e.g., Brown, Bosson, Ziegler-Hill, & Evidence of greater self-esteem instability
Swann, 2 0 0 3 ; Jordan et al., 2 0 0 3 ; Kernis among narcissists suggests that their grandi
et al., 2 0 0 5 ; McGregor & Marigold, 2003). osity masks feelings of vulnerability that are
But narcissism is not necessarily so easily ex brought to the surface in challenging social
plained by the combination of high explicit contexts.
and low implicit self-esteem. Other research shows that narcissistic
Campbell, Bosson, Goheen, Lakey, and self-esteem is highly contingent on the indi
Kernis (2007) pointed out that narcissism vidual’s ability to self-enhance through other
has not been found to be negatively associat people (M orf & Rhodewalt, 2001). Thus
ed with implicit self-esteem; rather, its level the positive self-views of narcissists in many
is simply not as high as would be expected ways depend on confirmation by others in
by narcissists’ explicit reports of self-esteem. their social environment. On top of this, nar
Also, this latter work demonstrates that the cissistic self-esteem is based to a high degree
implicit self-esteem of narcissists tends to be on competition (Crocker, Luhtanen, Cooper,
higher when participants are provided with & Bouvrette, 2003) such that outperforming
agentic as opposed to communal words to others and the ability to make favorable so
evaluate in the Implicit Association Test cial comparisons is essential to maintaining
(IAT). Thus research on the implicit self grandiose self-views. Such contingencies may
views of narcissists has provided inconsis contribute to the vulnerability of all narcis
tent support for the idea that narcissistic sists, as it would take an extremely high level
grandiosity is a defense against an under of self-regulatory resources and skill to con
lying vulnerability and sense of worthless tinually affect the social environment in de
ness, as indexed by implicit self-esteem. O f sired ways that maintain self-esteem.
course, the weak and inconsistent findings Such ideas about the vulnerability of the
may reflect a combining of covert and overt narcissistic self, even the grandiose type,
narcissism, and it may be that only covert find support in self-regulatory models such
narcissists display the explicit-im plicit self as that set forth by M orf and Rhodewalt
esteem discrepancy. (2001). It is also indicative of more general
Another way to assess the possible under models of the “fragile” self (e.g., Kernis,
lying vulnerability of narcissists is through 2 0 0 3 ; Rhodewalt & Peterson, 2008). The
an examination of their grandiose self-views model recently described by Rhodewalt
over time. In other words, the relative stabil and Peterson (2008) may help to reconcile
ity or instability of high self-esteem across the potential vulnerabilities of both covert
time, within different situations, and in re and overt narcissists, as well as discrepan
sponse to challenges and threats may point cies in the psychological adjustment of the
37. N arcissism 557
two types. In this model, self-esteem is seen though the initial version (56 items and eight
as both an outcome of and input to effec subscales) is burdensome for researchers and
tive self-regulation toward important goals. awaits further validation. At this point, the
Importantly, both the content and structure VNS appears to display a small positive cor
of self-beliefs can influence regulatory effec relation with the NPI (Peterson & Rhode
tiveness and thus have implications for both walt, 2007; Pimental et al., 20 0 6 ), combined
the level and stability of one’s self-esteem. A with a moderately negative correlation with
“fragile” self is one that is organized such self-esteem. Thus, although there is some
that it is difficult to sustain and effectively overlap, the NPI and VNS appear to capture
regulate toward important goals, leaving different points on a potential continuum
the individual vulnerable to threat, low and/ of self-esteem and regulatory effectiveness.
or unstable self-esteem, and, potentially, However, for the sake of integrated assess
depression. Applied to narcissism, the core ment of an overall narcissism construct,
features of the construct (e.g., entitlement research would benefit from an instrument
and beliefs in superiority, whether overtly that included overt, covert, and shared ele
expressed or not) will promote interpersonal ments in one scale that is less cumbersome
self-regulation toward the confirmation of to administer than both the NPI and VNS
such content. What may come to distinguish together. An integrated narcissism scale
the overt narcissist (relatively unstable high would allow researchers to test some of the
self-esteem) from the covert narcissist (rela potential relationships described here, such
tively unstable low self-esteem) is each one’s as whether more covert/vulnerable elements
history of self-regulatory effectiveness in the emerge over time as self-regulatory effective
pursuit of similar underlying goals. Whereas ness and self-esteem decrease and whether
overt/grandiose narcissists enjoy a relatively more overt/grandiose elements emerge with
high (though somewhat imperfect) level of more effective self-regulation and higher
success at confirming their grandiose self self-esteem.
beliefs, covert/vulnerable narcissists may Additionally, elements that have been
have encountered frustration in their at identified as potential “core” descriptors of
tempts to affect their environment in desired an integrated construct would be expected
ways, leaving them with low self-esteem and to remain more stable over time. In line with
hypersensitive to further failure in these pur thinking about core aspects of narcissism,
suits. In this account, self-esteem regulation Campbell and colleagues (2004) have cre
is the key to differences between overt and ated a short Psychological Entitlement Scale
covert narcissists. (PES), the validity of which was established
over nine studies. These researchers explic
itly defined entitlement as “a stable and per
H ow Should Narcissism B e Assessed
vasive sense that one deserves more and is
Most Effectively?
entitled to more than others” (p. 31). They
Given such questions in theory and research, found that high scorers on this 9-item scale
there are obvious limitations to the current not only reported that they deserved more
methods of assessing narcissism. The NPI, pay in a hypothetical employment setting
the most commonly used instrument in re but also displayed interpersonal behaviors
search, appears to assess a narcissism that is similar to those found in studies of narcis
mostly descriptive of the grandiose pole. Ad sists, including competition, selfishness in
ditionally, it has not been updated in more relationships, and aggression. From this and
than 20 years, and its reliability and internal other research (e.g., Dickinson & Pincus,
consistency have been questioned (e.g., del 20 0 3 ; Emmons, 1987), it appears that en
Rosario & W hite, 2 0 0 5 ; Kubarych, 2004). titlement would be an important part of any
Until recently, researchers were forced to fully integrated scale of narcissism.
look to the M M PI to find an assessment in Thus researchers now have several options
strument that captured the vulnerable pole of when choosing an assessment instrument for
narcissism (with the exception of entitlement narcissism and its associated features. The
and exploitativeness subscales in the NPI). choice at this point depends on the questions
A new Vulnerable Narcissism Scale (VNS; being asked, as there is not yet an integrated
Pimental et al., 2 0 0 6 ) was recently created, tool that captures all parts of the picture
558 V I. S E L F -R E L A T E D D IS P O S IT IO N S
being painted by theory and research. It American Psychiatric Association. (1994). D ia g n o stic
remains to be seen whether this is a viable a n d sta tistic a l m a n u a l o f m e n ta l d iso rd er s (4th ed.).
Washington, DC: Author.
endeavor or whether scales assessing only American Psychiatric Association. ( 2 0 00 ). D ia g n o stic
the core features of narcissism (i.e., entitle a n d s ta tistic a l m a n u a l o f m e n ta l d iso rd er s (4th ed.,
ment) are appropriate considering the wide text rev.). Washington, DC : Author.
spectrum of associated features that define Ames, D. R., Rose, P., & Anderson, C. P. (200 6). The
NP I-16 as a short measure of narcissism. J o u r n a l o f
narcissism, from the vulnerable to the gran R e se a rc h in P erson ality , 4 0 , 4 4 0 - 4 5 0 .
diose ends. Ang, R. P., 8c Yusof, N. (20 06 ). Development and ini
tial validation of the Narcissistic Personality Que s
tionnaire for Children: A preliminary investigation
Conclusion using school-based Asian samples. E d u c a tio n a l
P sy ch olog y , 2 6 , 1- 18.
Ashby, H. U., Lee, R. R., 8c Duke, E. H. (1979, Au
On its journey from the clinic to the labora gust). A n arcissistic p e r s o n a lity d is o r d e r M M P I
tory, the construct of narcissism has taken sc a le. Paper presented at the annual convention
an interesting and, we would argue, quite of the American Psychological Association, New
York.
fruitful path. Personality and social psy
Baumeister, R. F., 8c Vohs, K. D. (2001). Narcissism as
chologists have expanded the construct, an addiction to esteem. P s y c h o lo g ic a l In qu iry, 12,
originally a psychoanalytically defined per 206-210.
sonality disorder, to describe a subclini Brown, R. P., Bosson, J . K., Zeigler-Hill, V., 8c Swann,
W. B. (2003). Self-enhancement tendencies among
cal personality type. This effort has been people with high explicit self-esteem: The moderat
facilitated by consensus among clinicians ing role of implicit self-esteem. S e lf a n d Id en tity , 2,
concerning narcissism’s defining character 16 9 -18 7.
istics, as well as face-valid assessment de Bushman, B., 8c Baumeister, R. F. (1998). Threatened
vices based on this consensual definition. egotism, narcissism, self-esteem, and direct and dis
placed aggression: Does self-love or self-hate lead to
Contemporary researchers have proposed violence? J o u r n a l o f P erso n a lity a n d S o c ia l P sy c h o l
and validated process models of narcissism og y , 75, 2 1 9 - 2 2 9 .
that focus on interpersonal self-esteem regu Buss, D. M ., 8c Chiodo, L. M . (1991). Narcissistic
lation. In this regard, narcissism is clearly an acts in everyday life. J o u r n a l o f P erson ality , 19,
179-215.
interpersonally embedded individual differ
Cain, N. M ., Pincus, A. L., 8c Ansell, E. B. (2 008).
ence. As progress toward understanding the Narcissism at the crossroads: Phenotypic descrip
process of narcissistic self-esteem regulation tion of pathological narcissism across clinical the
advances, however, some of the controversies ory, social/personality psychology, and clinical di
that divided the psychoanalytic community agnosis. C lin ic a l P sy ch o lo g y R ev iew , 2 8 , 6 3 8 - 6 5 6 .
Calhoun, G. B., Glaser, B. A., Stefurak, T., 8c Brad
have reemerged. In particular, the various shaw, C. P. (20 00 ). Preliminary validation of the
psychoanalytic perspectives disagreed about Narcissistic Personality Inventory— Juvenile O f
how to best conceptualize the coexistence of fender. In te r n a tio n a l J o u r n a l o f O ffe n d e r T h era p y
narcissistic grandiosity and vulnerability. In a n d C o m p a ra tiv e C rim in o lo g y , 4 4 , 5 6 4 - 5 8 0 .
Campbell, W. K., Bonacci, A. M ., Shelton, J ., Exline,
this chapter we have discussed state-of-the-
J. J . , & Bushman, B. J. (20 0 4). Psychological entitle
art theory and research on the narcissistic ment: Interpersonal consequences and validation of
personality type with attention to measure a self-report measure. J o u r n a l o f P erson ality A s
ment issues and offered recommendations sessm en t, 8 3 , 2 9 - 4 5 .
for conceptualizing and assessing overt Campbell, W. K., Bosson, J . K., Goheen, T. W., Lakey,
C. E., 8c Kernis, M. H. (200 7). Do narcissists dis
and covert narcissism within contemporary
like themselves “deep down inside”? P s y c h o lo g ic a l
frameworks. S cien ce, 18, 2 2 7 - 2 2 9 .
Campbell, W. K., 8c Foster, J. D. (200 7 ). T he narcis
sistic self: Background, an extended-agency model,
R eferen ces and ongoing controversies. In C. Sedikides 8c S. J.
Spencer (Eds.), T h e s e l f (pp. 1 1 5 - 1 3 8 ). New York:
Akhtar, S., 8c Thomson , J . A. (1982). Overview: N a r Psychology Press.
cissistic personality disorder. A m erica n J o u r n a l o f Crocker, J ., Luhtanen, R. K., Cooper, M . L ., 8c Bou-
P sy ch iatry , 139, 1 2 - 2 0 . vrette, A. (2003). Contingencies of self-worth in col
American Psychiatric Association. (1980). D ia g n o stic lege students: Theory and measurement. J o u r n a l o f
a n d sta tis tic a l m a n u a l o f m e n ta l d is o r d e r s (3rd ed.). P erso n a lity a n d S o c ia l P sy ch olog y , 8 5, 8 9 4 - 9 0 8 .
Washington, DC . del Rosario, P. M ., 8c White, R. M. (20 05). The N a r
American Psychiatric Association. (1987). D ia g n o stic cissistic Personality Inventory: Test-retest stability
a n d sta tis tic a l m a n u a l o f m e n ta l d iso rd er s (3rd ed., and internal consistency. P erson ality a n d In d iv id u
revised). Washington, DC: Author. a l D iffe r e n c e s , 3 9, 1 0 7 5 -1 0 8 1 .
37. N arcissism 559
Dickinson, K. A., & Pincus, A. L. (2003). Interperson Kansi, J. (2003). The Narcissistic Personality Inven
al analysis of grandiose and vulnerable narcissism. tory: Applicability in a Swedish population sample.
J o u r n a l o f P erson ality D iso r d er s, 17, 18 8-2 0 7 . S ca n d in a v ia n J o u r n a l o f P sy ch olog y , 4 4 , 441 —
F.llis, H. (1878). Autoeroticism: A psychological study. 448.
A lien ist a n d N eu r o lo g ist, 19, 2 6 0 - 2 9 9 . Kernberg, O. F. (1975). B o rd e rlin e c o n d itio n s a n d
Emmons, R. A. (1984). Factor analysis and construct p a t h o lo g ic a l n arcissism . New York: Aronson.
validity of the Narcissistic Personality Inventory. Kernis, M . H. (20 03). Toward a conceptualization
J o u r n a l o f P erson ality a n d S o c ia l P sy ch o lo g y , 4 8 , of optimal self-esteem. P sy c h o lo g ic a l In qu iry, 14,
291-300. 1-26.
Emmons, R. A. (1987). Narcissism: Theory and mea Kernis, M . H., Abend, T. A., Gold man, B. M ., Shri-
surement. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ra, I., Paradise, A. N., & Hampton, C. (2005).
c h o lo g y , 5 2 , 11-17. Self-serving responses arising from discrepancies
Fossati, A., Beauchaine, T. P., Grazioli, F., Carretta, L., between explicit and implicit self-esteem. S e lf a n d
Cortinovis, F., & Maffei, C. (20 0 5). A latent struc Id en tity , 4, 3 1 1 - 3 3 0 .
ture analysis of D ia g n o stic a n d S ta tistica l M an u al Kernis, M . H., & Sun, C .- R. (1994). Narcissism and
o f M en ta l D iso r d er s, F o u rth E d itio n , narcissistic reactions to interpersonal feedback. J o u r n a l o f R e
personality disorder criteria. C o m p r e h e n s iv e P sy se a r c h in P erson ality , 2 8 , 4 - 1 3 .
ch iatry , 4 6 , 3 6 1 - 3 6 7 . Kohut, H. (1971). T h e an aly sis o f th e self. New York:
Foster, J. D., & Campbell, W. K. (200 7). Are there International Universities Press.
such things as “narcissists” in social psychology?: A Kubarych, T. S. (20 04 ). The Narcissistic Personality
taxometric analysis of the Narcissistic Personality Inventory: Factor structure in a nonclinical sample.
Inventorv. P erson ality a n d In d iv id u a l D iffer en c es , P erson ality a n d In d iv id u a l D iffe r e n c e s , 3 6 , 8 5 7 -
4 3 , 1321-1332. 87 2 .
Freud, S. (1953). On narcissism: An introduction. In J. Lasch, C. (1979). T h e cu ltu re o f n arcissism : A m e r i
Strachey (Ed. & Trans.), T h e s ta n d a r d e d itio n o f the can life in an a g e o f d im in ish in g e x p e c ta tio n s . New
c o m p le t e p s y c h o lo g ic a l w o r k s o f S igm u n d F reu d York: Norton.
(Vol. 14, pp. 6 9 - 1 0 2 ). London: Flogarth Press. McGregor, L, & Marigold, D. C. (2003). Defensive
(Original work published 1914) zeal and the uncertain self: W hat makes you so
Gabbard, G. O. (1998). Transference and counter sure? J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y ,
transference in the treatment of narcissistic patients. 8 5, 8 3 8 - 8 5 2 .
In E. Ronningstam (Ed.), D iso rd ers o f n arcissism : Millon, T., & Davis, R. D. (1997). The MC M I- I1 I:
D ia g n o stic , c lin ic a l, a n d em p ir ic a l im p lic a tio n s Present and future directions. J o u r n a l o f P ers o n a l
(pp. 1 2 5 -1 4 5 ). Washington, DC: American Psychi ity A ssessm en t, 6 8 , 6 9 - 8 5 .
atric Press. Morey, L. C., Waugh, M . H., & Blashfield, R. K.
Gabriel, M . T., Critelli, J. W., & Ee, J. S. (1994). Narcis (1985). M M P I scales for D SM -III personality dis
sistic illusions in self-evaluations of intelligence and orders: Their derivation and correlates. J o u r n a l o f
attractiveness. J o u r n a l o f P erson ality , 6 2 , 1 4 3 -15 5. P erson ality A ssessm en t, 5 2 , 6 1 0 - 6 2 5 .
Greenwald, A. G. (1980). The totalitarian ego: Fabri Morf, C. C. (1994). I n te r p e r s o n a l c o n s e q u e n c e s o f
cation and revision of personal history. A m erican n arcissists' c o n tin u a l e ffo r t s to m a in ta in a n d b o l
P sy ch olog ist, 3 5 , 6 0 3 - 6 1 8 . ste r s e lf-e stee m . Unpublished doctoral dissertation,
Greenwald, A. G., & Banaji, M . R. (1995). Implicit University of Utah.
social cognition: Attitudes, self-esteem, and stereo Morf, C. C., & Rhodewalt, F. (1993). Narcissism and
types. P sy c h o lo g ic a l R ev iew , 102, 4 - 2 7 . self-evaluation maintenance: Flxplorations in object
Gunderson, J . G., Ronningstam, E., & Smith, I.. E. relations. P erson ality a n d S o c ia l P sy ch o lo g y B u lle
(1991). Narcissistic personality disorder: A review tin, 19, 6 6 8 - 6 7 6 .
of data on D SM -III-R descriptions. J o u r n a l o f P er Morf, C. C., & Rhodewalt, F. (2001). Unraveling the
so n a lity D iso r d er s, 5, 167-177. paradoxes of narcissism: A dynamic self-regulatory
Hendin, H. M ., & Cheek, J . M. (1 997). Assessing hy processing model. P sy c h o lo g ic a l In q u iry , 12, 177—
persensitive narcissism: A reexamination of M u r 196.
ray’s narcism scale. J o u r n a l o f R e se a rc h in P ers o n Morrison, A. P. (1986) Introduction. In A. P. Morrison
ality , 3 1 , 5 8 8 - 5 9 9 . (Ed.), E s s e n tia l p a p e rs o n n arcissism (p. 11). New
Higgins, E. T. (1987). Self-discrepancy: A theory re York: New York University Press.
lating self and affect. P s y c h o lo g ic a l R ev iew , 8 0 , Mullins, L. S., & Kopehnan, R. E. (1988). Toward an
307-336. assessment of the construct validity of four mea
John, O. P., & Robins, R. (1994). Accuracy and bias sures of narcissism. J o u r n a l o f P erson ality A s se ss
in self-perception: Individual differences in self m en t, 5 2 , 6 1 0 - 6 2 5 .
enhancement and the role of narcissism. J o u r n a l o f Paulhus, D. I.. (1998). Interpersonal and intrapsychic
P erson ality a n d S o c ia l P sy ch o lo g y , 6 6 , 2 0 6 - 2 1 9 . adaptiveness of trait self-enhancement: A mixed
Jones, E. F.., & Berglas, S. (1978). Control of attri blessing? J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l
butions about the self through self-handicapping og y , 74, 1 1 9 7 - 1 2 0 8.
strategies: T he appeal of alcohol and the role of un Paulhus, D. L. (2001). Normal narcissism: Two mini
derachievement. P erson ality a n d S o c ia l P sy ch olog y malist views. P sy c h o lo g ic a l In qu iry, 12, 2 2 8 - 2 3 0 .
B u lletin , 4, 2 0 0 - 2 0 6 . Peterson, B., & Rhodewalt, F. (20 07 ). |The ostracized
Jordan, C. H., Spencer, S. J., Zanna, M. P., Hoshino- narcissist: Differential reactions to being ignored|.
Browne, E., & Correll, J . (2003). Secure and defen Unpublished raw data, University of Utah.
sive high self-esteem. J o u r n a l o f P erson ality a n d Pimentel, C. A., Ansell, E. B., Pincus, A. L., & Cain,
S o c ia l P sy c h o lo g y , 8 5 , 9 6 9 - 9 7 8 . N. M. (20 06 ). In itia l v a lid a tio n a n d d er iv a tio n o f
560 V I. S E L F -R E L A T E D D IS P O S IT IO N S
th e V u ln erab le N arcissism S ca le. Unpublished man Rhodewalt, F., & Sorrow, D. L. (2003). Interpersonal
uscript, Pennsylvania State University. self-regulation: Lessons from the study of narcis
Pulver, S. (1970). Narcissism: T he term and concept. sism. In M. R. Leary & J. P. Tangney (Eds.) H a n d
J o u r n a l o f th e A m erica n P sy ch o -A n a ly tic A s s o c ia b o o k o f s e l f a n d id en tity (pp. 5 1 9 - 5 3 5 ) . New York:
tion . 18, 3 1 9 - 3 4 1 . Guilford Press.
Raskin, R., & Hall, C. S. (1979). A narcissistic person Rhodewalt, F., Tragakis, M ., & Finnerty, J. (200 6).
ality inventory. P s y c h o lo g ic a l R e p o rts , 4 0 , 590. Narcissism and self-handicapping: Linking self-
Raskin, R., & Hall, C. S. (1981). The Narcissistic Per aggrandizement to behavior. J o u r n a l o f R e se a rc h in
sonality Inventory: Alternate form reliability and P erson ality , 4 0 , 5 7 3 - 5 9 7 .
further evidence of construct validity. J o u r n a l o f Rose, P. (2002). The happy and unhappy faces of nar
P erso n a lity A ssessm en t, 4 5, 1 5 9 - 1 6 2 . cissism. P erson ality a n d In d iv id u a l D iffe r e n c e s , 3 3 ,
Raskin, R., Novacek, J., & Hogan, R. (1991). Narcis 379-391.
sism, self-esteem, and defensive self-enhancement. Rose, P., & Campbell, W. K. (200 4 ). Greatness feels
J o u r n a l o f P erson ality , 59, 2 0 - 3 8 . good: A telic model of narcissism and subjective
Raskin, R., & Terry, H. (1988). A principal- well-being. In S. P. Shohov (Ed.), A d v a n ce s in p s y
components analysis of the Narcissistic Personal c h o lo g y research (Vol. 31, pp. 3 - 2 7 ) . Hauppauge,
ity Inventory and further evidence for its construct NY: Nova Science.
validity. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l Ruiz, J . M ., Smith, T. W., & Rhodewalt, F. (2001).
og y , 5 4 , 8 9 0 - 9 0 2 . Distinguishing narcissism and hostility: Similari
Reich, A. (1960). Pathologic forms of self-esteem ties and differences in interpersonal circumplex and
regulation. P sy c h o a n a ly tic S tu dy o f th e C h ild , 18, five-factor correlates. J o u r n a l o f P erso n a lity A s se ss
218-238. m en t, 76, 5 3 7 - 5 5 5 .
Rhodewalt, F. (2001). The social mind of the narcis Sedikides, C., Rudich, E. A., Gregg, A. P., Kumashi-
sist: Cognitive and motivational aspects of interper ro, M ., &c Rusbult, C. (2004). Are normal narcis
sonal self-construction. In J. P. Forgas, K. Williams, sists psychologically healthy?: Self-esteem matters.
& L. Wheeler (Eds.), T h e s o c ia l m in d : C o g n itiv e J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y , 87,
a n d m o tiv a tio n a l a s p e c t s o f in te r p e r s o n a l b e h a v io r 400-416.
(pp. 1 7 7 -1 98 ). New York: Cambridge University Trapnell, P. D., & Wiggins, J . S. (1990). Extension of
Press. the interpersonal adjective scales to include the Big
Rhodewalt, F., & Eddings, S. (2002). Narcissus re Five dimensions of personality. J o u r n a l o f P e r s o n a l
flects: Memory distortion in response to ego relevant ity a n d S o c ia l P sy ch olog y , 5 9, 7 8 1 - 7 9 0 .
feedback in high and low narcissistic men. J o u r n a l Vazire, S., & Funder, D. ( 2 00 6). Impulsivity and the
o f R e se a rc h in P erson ality , 3 6 , 9 7 - 1 1 6 . self-defeating behavior of narcissists. P erson ality
Rhodewalt, F., Madrian, J. C., & Cheney, S. (1998). a n d S o c ia l P sy ch o lo g y R e v iew , 10, 1 5 4 - 1 6 5 .
Narcissism, self-knowledge organization, and em o Watson, P. J ., Little, T., Sawrie, S. M., & Biderman,
tional reactivity: T he effect of daily experiences on M. D. (1992). Measures of the narcissistic person
self-esteem and affect. P erso n a lity a n d S o c ia l P sy ality: Complexity of relationships with self-esteem
c h o lo g y B u lletin , 2 4 , 7 5 - 8 7 . and empathy. J o u r n a l o f P erso n a lity D iso r d er s, 6,
Rhodewalt, F., & Morf , C. C. (1995). Self and inter 434-449.
personal correlates of the Narcissistic Personality Watson, P. )., Sawrie, S. M ., Greene, R. L., & Arredon
Inventory: A review and new findings. J o u r n a l o f do, R. (2 002). Narcissism and depression: M M P I -2
R e se a rc h in P erson ality , 2 9 , 1 - 2 3 . evidence for the continuum hypothesis in clinical
Rhodewalt, F., & Mo rf, C. C. (1998). On self- samples. J o u r n a l o f P erson ality A ssessm en t, 79,
aggrandizement and anger: A temporal analysis of 85-109.
narcissism and affective reactions to success and Wiggins, J. S., & Pincus, A. L. (1989). Conceptions of
failure. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o l personality disorders and dimensions of personal
og y , 74, 6 7 2 - 6 8 5 . ity. J o u r n a l o f C lin ica l a n d C o n su ltin g P sy ch olog y ,
Rhodewalt, F., & Morf, C. C. (2005). Reflections in 1, 3 0 5 - 3 1 6 .
troubled waters: Narcissism and the vicissitudes of W ink, P. (1991). Two faces of narcissism. J o u r n a l o f
an interpersonally contextualized self. In A. Tesser, P erson ality a n d S o c ia l P sy c h o lo g y , 61, 5 9 0 - 5 9 7 .
J . V. Wood, & D. A. Staper (Eds.), O n b u ild in g , d e W ink, P., & Gough, H. G. (1990). New narcissism
fen d in g , a n d reg u latin g th e s e l f (pp. 1 2 7 -1 5 2 ). New scales for the California Psychological Inventory
York: Psychology Press. and M M P I .J o u r n a l o f P erson ality A ssessm en t, 5 4,
Rhodewalt, F., & Peterson, B. (2 008). T he self and 446-462.
social behavior: T he fragile self and interpersonal Ziegler-Hill, V. ( 2 0 06 ). Discrepancies between implicit
self-regulation. In F. Rhodewalt (Ed.), P erson ality and explicit self-esteem: Implications for narcissism
a n d s o c ia l b e h a v io r (pp. 4 9 - 7 8 ) . New York: Taylor and self-esteem instability. Jo u r n a l o f P erson ality ,
& Francis. 74, 1 1 9 -1 43 .
CHAPTER 38
Self-Compassion
K r is t in N eff
561
562 V I. S E L F -R E L A T E D D IS P O S IT IO N S
flawed condition to the shared human condi ogy and especially attachment theory. Gil
tion, so that one’s own characteristics and bert argues that self-compassion taps into
experiences are considered from a broad, an evolved mammalian physiological system
inclusive perspective. In the same way, life guiding attachment and caregiving behavior.
difficulties and struggles are framed in light When accessed via external signals (other
of the shared human experience, so that one people’s behavior) or internal signals (self
feels connected to others when experiencing directed thoughts and emotions) of kindness
pain. Often, however, people feel isolated and caring, individuals experience feelings
and cut off from others when considering of connectedness and soothing. In contrast,
their personal flaws, as if the failing were an self-criticism taps into the threat-focused
aberration not shared by the rest of human physiological systems of social ranking,
kind. Similarly, people often fall into the which involve aggressive dominance and
trap of believing that they are the only ones fearful submission (Gilbert, 1989, 2005).
struggling when they experience difficult life From this perspective, self-compassion in
circumstances, and they feel a sense of isola volves an interdependent set of motives and
tion and separation from other people who competencies that relate to prototypic car
are presumably leading “normal,” happy ing: concern for individuals’ well-being,
lives. sensitivity to individuals’ distress and needs,
Mindfulness, the third component of sympathy, distress tolerance, empathy, and
self-compassion, involves being aware of nonjudgment. These are called the compas
present-moment experience in a clear and sion circle and are directed toward others or
balanced manner so that one neither ignores toward oneself.
nor ruminates on disliked aspects of oneself
or one’s life (Brown & Ryan, 2003). First, it
is necessary to recognize that one is suffer R esearch on Self-Com passion
ing in order to be able to extend compassion
toward the self. While it might seem that Much of the research on self-compassion has
personal suffering is blindingly obvious, been conducted using the Self-Compassion
people do not always pause to acknowledge Scale (SCS; Neff, 2003a), though research
their own pain when they are busy judging ers are also starting to use mood inductions
themselves or coping with life’s challenges. or therapeutic interventions as a means of
Mindfulness involves a sort of stepping out examining the impact of self-compassion on
of oneself, taking a metaperspective on one’s functioning (e.g., Gilbert & Proctor, 2 0 0 6 ;
own experience so that it can be considered Leary, Tate, Adams, Allen, & Hancock,
with greater objectivity and perspective. 2007). The SCS is a 26-item self-report
Thus mindfulness enables a way of relating scale that is composed of six subscales: self
to oneself that involves one aspect of the self kindness, self-judgment, common humanity,
giving compassion to another aspect of the perceived isolation, mindfulness, and overi
self. Mindfulness also prevents being swept dentification. The subscales are highly inter
up in and carried away by the story line of correlated, however, and confirmatory factor
one’s own pain, a process that N eff (2003b) analyses have indicated that these intercorre
has termed overidentification. When caught lations can be explained by a single overarch
up in this manner, one tends to ruminate and ing factor termed “Self-Compassion” (Neff,
obsessively fixate on negative self-relevant 2003a). M ost research to date has focused
thoughts and emotions, so that the mental on overall self-compassion scores rather
space needed to be self-compassionate is un than examining the various subcomponents
available. of self-compassion separately. The SCS evi
dences strong internal reliability (consistent
ly above .90), as well as test-retest reliability
O ther Conceptualizations
(.93 over a 3-week interval) (Neff, 2003a).
o f Self-C om passion
Convergent validity for the scale is strong,
It should be noted that other ways of defin with self-reported SCS scores substantially
ing self-compassion exist in the literature. overlapping with observer reports (either by
Gilbert (1989, 2005) views self-compassion romantic partners or therapists) (Neff, 2 0 0 6 ;
through the lens of evolutionary psychol Neff, Kirkpatrick, & Rude, 2007). The scale
38. S elf-C o m p assion 563
also shows discriminant validity: Practicing as we, and making more social references to
Buddhists report higher SCS scores than friends, family, and other humans.
non-Buddhists (Neff, 2003a), for instance. Similarly, Leary and colleagues (2007) in
vestigated the way that self-compassionate
people deal with negative life events using
Self-C om passion and
experience-sampling techniques, asking
Psychological Resilience
participants to report about problems they
One of the most robust and consistent find experienced over a 20-day period. Individu
ings in the research literature is that greater als with higher levels of self-compassion
self-compassion (as reported on the SCS) is had more perspective on their problems and
linked to lower anxiety and depression, with were less likely to feel isolated by them. For
zero-order correlations typically falling in example, they were more likely to feel that
the range o f - .5 0 to - .6 0 for depression and their struggles were not any worse than
- .6 0 to - .7 0 for anxiety (Neff, 2 0 0 3 a ; Neff, what lots of other people go through and
Hseih, & Dejitthirat, 2 0 0 5 ; Neff, Kirkpat were less likely to think that their lives were
rick, & Rude, 20 0 7 ; Neff, Pisitsungkagarn, more “screwed up” than those of others.
& Hseih, 2008). O f course, a key feature of They also experienced less anxiety and self-
self-compassion is that individuals do not consciousness when thinking about their
harshly judge and criticize themselves when problems.
they notice something about themselves they The emotional resilience provided by self
do not like, and self-criticism is known to be compassion is further evidenced by findings
an important predictor of anxiety and depres that self-compassionate individuals tend to
sion (Blatt, 1995). However, self-compassion engage in less rumination and thought sup
is still a robust negative predictor of anxiety pression than those scoring low on the trait
and depression even after controlling for (Neff, 2 003a; Neff, Kirkpatrick, & Rude,
self-criticism (Neff, 2003a), suggesting that 2007). Self-compassion entails taking a bal
self-compassion provides unique buffering anced approach to one’s emotional experi
effects. Similarly, self-compassion is a nega ence— so that one neither runs away from
tive predictor of anxiety even when control or away with one’s feelings. Thus it appears
ling for negative affect (Neff, Kirkpatrick, that self-compassionate individuals can face
& Rude, 20 0 7 ). Thus self-compassion is up to personal weaknesses and life challeng
not merely a matter of looking on the bright es with fewer emotional overreactions. Simi
side of experiences and avoiding negative larly, self-compassion is related to emotional
feelings. Self-compassionate individuals rec intelligence. Individuals with higher levels
ognize when they are suffering, but when of self-compassion report greater emotional
doing so they provide themselves feelings of coping skills, clarity of feelings, and ability
warmth, kindness, and interconnectedness to repair negative emotional states (Neff,
with the rest of humanity. As Gilbert and 2003a).
Irons (2005) suggested, self-compassion may
help activate the self-soothing system (relat
Self-C om passion and
ed physiologically to the parental caregiving
Psychological Strengths
system), and therefore help reduce feelings of
fear and isolation. In addition to providing protection against
In support of this proposition, Neff, Kirk negative mental states, self-compassion also
patrick, and Rude (2007) conducted a study appears to bolster positive emotional mind
involving a mock interview task in which par sets. For instance, self-compassion has been
ticipants were asked to write their answers linked to greater feelings of social connected
to a difficult interview question: “Please de ness and life satisfaction, important elements
scribe your greatest weakness.” Individuals of a meaningful life (Neff, 2 003a; Neff, Kirk
with higher levels of self-compassion experi patrick, & Rude, 2007; N eff et al., 2008). It
enced less anxiety after the task. They also has also been associated with feelings of au
tended to use less isolating language when tonomy, competence, and relatedness (Neff,
writing about their weaknesses, using fewer 2003a), indicating that self-compassion
first-person singular pronouns such as /, helps meet basic psychological needs that
using more first-person plural pronouns such Deci and Ryan (1995) identified as funda
564 V I. S E L F -R E L A T E D D IS P O S IT IO N S
comfortable with their bodies and had less ling, and verbally or physically aggressive.
anxiety regarding social evaluations of their Self-compassion was also associated with
physiques. more relationship satisfaction (as reported
Self-compassion may help people learn by participants and their partners) and
to deal with the intense pressures to be thin greater attachment security. Because self-
and attractive in Western society, while still compassionate people give themselves car
promoting healthy eating patterns. One ing, understanding, and support, they appear
study investigated whether inducing a state to have more emotional resources available
of self-compassion attenuates certain disor to give to their romantic partners. Also, the
dered eating behaviors. Highly restrictive ability to admit mistakes without ego de
eaters (i.e., dieters) often display a paradoxi fensiveness means that self-compassionate
cal tendency— if they break their diets and people may have less need to project their
eat high-calorie foods, they tend to eat even faults onto partners via angry accusations
more afterward (a process known as the (Feldman & Ciowen, 1998).
disinhibition effect). Heatherton and Polivy An interesting question concerns whether
(1990) proposed that this pattern of over self-compassionate people are more com
eating is an attempt to reduce the negative passionate toward others in general. On the
affect associated with the lapse of a desired one hand, cultivating an openhearted stance
goal. Adams and Leary (2007) asked college toward oneself that recognizes human inter
women to eat an unhealthy food (a dough connectedness should theoretically facilitate
nut) as part of an experiment and either in being kind, forgiving, and empathetic toward
duced them to think self-compassionately others. On the other hand, given that people
about eating the donut or else gave them no who lack self-compassion say they are much
intervention. Participants were later given kinder to others than they are to themselves
the opportunity to eat as much candy as they (Neff, 2003a), it may be that the tendency to
wanted while unobserved. Results showed be kind and giving toward others is relative
that the self-compassion induction reduced ly independent from the tendency to be com
negative affect and attenuated the amount of passionate toward oneself. Although there is
candy eaten after the doughnut among high very little research on this topic, preliminary
ly restrictive eaters (who displayed similar findings suggest that the link between self
patterns to nondieters). In contrast, highly compassion and other-focused concern is
restrictive eaters in the control condition ate mixed. N eff (2008) found that individuals
more candy afterward. Again, having com with higher levels of self-compassion report
passion for mistakes and failures allows such ed a significantly greater tendency to forgive
lapses to be taken less personally (in other others than those low in self-compassion.
words, they do not define the self as “bad” or They also reported being more likely to take
“unworthy”). Thus self-compassionate indi others’ perspectives and to feel less person
viduals are able to emotionally recover from al distress when considering other people’s
transgressions more quickly and to continue misfortunes. However, self-compassion had
to work toward their goals of growth and only a very weak association with the ten
change. dency to experience compassionate love to
ward others (Sprecher & Fehr, 2005). Also,
self-compassion was not significantly linked
Self-C om passion
to empathy for others or to altruism. These
and Interpersonal Functioning
results suggest that there may be some as
Just as there is a fair amount of evidence that pects of other-focused concern that are facil
self-compassion psychologically benefits itated by self-compassion— such as the abil
oneself, there are also some indications that ity to detach oneself from one’s own point
self-compassion also benefits others within of view and take another’s perspective or to
interpersonal relationships. In a study of recognize that all humans make mistakes
heterosexual couples (Neff, 2 0 0 6 ), self- and are worthy of forgiveness. However,
compassionate individuals were described self-compassion does not appear to predict
by their partners as being more emotionally general emotional responsiveness toward
connected, accepting, and autonomy sup others in terms of kindness, compassionate
porting while being less detached, control love, or empathy. Clearly, more research will
566 V I. S E L F -R E L A T E D D IS P O S IT IO N S
be needed to examine these issues, and it a real-life setting provides strong support for
will be important to examine the impact of the link between self-compassion and psy
self-compassion on behavior in experimen chological well-being.
tal settings rather than relying solely on self-
reports.
Self-Com passion versus Self-Esteem
Self-Cotnpassion in Therapeutic Settings
Interest in self-compassion has been spurred
An exciting area of research concerns the ap by the observation that self-compassion is
plication of self-compassion in clinical con associated with many of the benefits of high
texts. Neff, Kirkpatrick, and Rude (2007) self-esteem, while having fewer of the down
conducted a study that tracked changes in sides associated with self-esteem pursuit. For
self-compassion experienced by therapy instance, Leary and colleagues (2007) exam
clients over a 1-month interval. Therapists ined self-compassionate individuals’ reac
used a Gestalt two-chair technique designed tions to a mildly awkward and embarrassing
to help clients lessen self-criticism and have task— being videotaped while looking into
greater compassion for themselves (Green a camera and making up a children’s story
berg, 1983; Safran, 1998). Results indicated that began “Once upon a time there was a
that increased self-compassion levels over little bear.” Self-compassionate people rated
the month-long period (which were assessed their tapes more favorably and felt better
under the guise of an unrelated study) were while watching their tapes than those who
linked to fewer experiences of self-criticism, were low in self-compassion, indicating that,
depression, rumination, thought suppres like self-esteem, self-compassion is a source
sion, and anxiety. of positive emotions toward the self. How
Gilbert and Procter (2006) have de ever, self-compassionate individuals also
veloped a group-based therapy interven rated how they appeared on the tape (e.g.,
tion called “compassionate mind train awkward, competent, attractive, nervous)
ing” (CM T). The model is designed to help in a similar way to objective observers. This
people develop skills of self-compassion, suggests that self-compassionate individuals
especially when their more habitual form did not display the type of self-enhancement
of self-relating involves self-attacking. In a bias often associated with high self-esteem
pilot study of C M T among patients in a day (Robins & Beer, 2001).
treatment program for people suffering from Research indicates that people some
intense shame and self-criticism, individuals times engage in dysfunctional behaviors in
were led through weekly 2-hour C M T ses order to maintain a sense of high self-worth
sions for 12 weeks. Participants were in (for reviews, see Blaine & Crocker, 1993;
structed about the qualities involved in self Crocker & Park, 2004). People who are
compassion (e.g., developing empathy for highly invested in having high self-esteem
one’s own distress), explored their fears of often display narcissistic tendencies (M orf
being too self-compassionate (e.g., “it makes & Rhodewalt, 2001), a maladaptive pattern
me feel vulnerable”), and were helped to un that causes interpersonal problems (Camp
derstand their own self-critical tendencies bell & Baumeister, 2001). Those wanting to
without judgment. Participants were also in maintain high self-esteem sometimes trivial
vited to create an ideal image of caring and ize personal failings or blame them on ex
compassion, a figure embodying qualities of ternal causes, hindering their ability to grow
wisdom, strength, warmth, and nonjudg- and change (Sedikides, 1993). Other ways
mental acceptance. The training resulted in to protect high self-esteem involve becom
significant pre-post changes in depression, ing angry toward those who threaten one’s
self-attacking, feelings of inferiority, sub ego (Baumeister, Smart, & Boden, 1996), or
missive behavior, and shame. Moreover, al engaging in downward social comparisons
most all of the participants felt ready to be (Fein & Spencer, 1997). The motivation to
discharged from their hospital day program protect feelings of self-worth can also lead to
at the end of the study. This demonstration a type of closed-mindedness known as “need
of the healing qualities of self-compassion in for cognitive closure” (Jost, Glaser, Krug-
38. S elf-C o m p assion 567
lanski, & Sulloway, 2 0 0 3 ; Taris, 2 0 0 0 ), in ety (Neff, 2003a). The two constructs dif
which alternative viewpoints are not toler fer in important ways, moreover. Whereas
ated. high self-esteem depends on successful per
Because global self-esteem rests in part on formances and positive self-evaluations,
evaluations of self-worth in various life do self-compassion is relevant precisely when
mains, high self-esteem is often contingent self-esteem tends to falter— when one fails
on particular outcomes (Crocker, Luhtanen, or feels inadequate. Thus self-compassion
Cooper, & Bouvrette, 2003). This means provides a way of dealing with negative life
that self-esteem can fluctuate according to experiences that self-esteem cannot pro
particular circumstances. Even though trait vide. In the Neff, Kirkpatrick, and Rude
levels of global self-esteem tend to remain (2007) mock interview study asking people
relatively constant over time, state feelings to describe their greatest weaknesses, for
of self-worth may be highly unstable and instance, self-compassion provided a sig
change quite frequently (Kernis, Paradise, nificant buffer against anxiety, whereas trait
W hitaker, W heatman, & Goldman, 2000). self-esteem did not.
In fact, individuals with high levels of com Leary and colleagues (2007) found that
petence may be most vulnerable to experi when considering hypothetical scenarios
encing drops in state self-esteem (Crocker & involving failure or embarrassment (e.g.,
Park, 20 0 4 ), because they have more oppor being responsible for losing an athletic com
tunities to fall short of their personal stan petition for one’s team), participants with
dards (e.g., the A student who receives a B+ greater self-compassion reported less nega
on an exam). As the Hollywood saying goes, tive affect (e.g., sadness or humiliation) and
you’re only as good as your latest success (at more emotional equanimity (e.g., remaining
least when viewing the world through the calm and unflustered). In contrast, global
lens of self-esteem). levels of trait self-esteem predicted no vari
Research indicates that self-compassion ance in emotional reactions after controlling
is moderately associated with trait levels of for self-compassion levels. In another study,
global self-esteem, with correlations around participants were asked to give a brief intro
.5 5 -.6 0 using the Rosenberg (1965) measure duction of themselves on video (describing
(Leary et al., 200 7 ; Neff, 2 0 0 3 a ; N eff et al., interests, future plans, etc.) and were then
2008). This is unsurprising given that both given positive or negative feedback about
constructs represent a positive emotional the introduction that was ostensibly made
stance toward the self. Similarly, self-esteem by an observer. Participants’ reactions to
and self-compassion are both associated the feedback were then assessed, including
with emotional well-being, such as lower their attributions for the observer’s feed
levels of anxiety and depression, as well as back. Individuals with low self-compassion
higher levels of happiness, optimism, and gave defensive attributions. They were more
life satisfaction. Unlike self-esteem, how likely to attribute the observer’s feedback to
ever, the healthy states of mind associated their own personalities when the feedback
with self-compassion do not stem from posi was positive rather than negative. Individu
tive evaluations of the self, meeting set stan als high in self-compassion, however, were
dards, or favorable comparisons with others. equally likely to attribute the feedback to
Rather, they stem from recognizing the need their personalities regardless of whether the
to be kind to oneself in instances of suffer feedback was positive or negative. An op
ing and framing one’s experience in light of posite pattern was found for self-esteem.
the shared human experience— fragile and Individuals with low self-esteem were
imperfect as it is. Thus self-compassion ap equally likely to attribute the feedback to
pears to provide emotional resilience over their personalities when feedback was posi
and above that attributable to self-esteem. tive or negative, but participants with high
For example, when controlling for self self-esteem were more likely to attribute the
esteem, self-compassion is still a robust feedback to their own personalities when the
predictor of happiness, optimism, and posi feedback was positive rather than negative.
tive affect (Neff &C Vonk, 2009), and it also This suggests that self-compassion enables
negatively predicts depression and anxi people to admit and accept negative as well
568 V I. S E L F -R E L A T E D D IS P O S IT IO N S
sociated with feelings of secure attachment, linked to significantly lower levels of neu
safeness, and the oxytocin-opiate system). roticism. This is perhaps unsurprising, given
In contrast, self-esteem is thought to be an that the feelings of self-judgment, isolation,
evaluation of superiority-inferiority that and ruminative emotional processing inher
helps to establish social-rank stability and ent in the lack of self-compassion are similar
is related to alerting, energizing impulses to those described by the neuroticism con
and dopamine activation (Gilbert & Irons, struct. Self-compassion was also positively
2005). Self-compassion, therefore, enhances associated with agreeableness, extraversion,
feelings of interconnectedness, whereas self and conscientiousness (correlations ranged
esteem positions the self in competition with from .3 2 -.4 2 ), but no association was found
others and amplifies feelings of distinctive with openness to experience. The socially
ness and separation. oriented nature of people high in agreeable
Self-compassion may be a more useful ness and extraversion may help them to be
way to conceptualize a healthy way of relat kind to themselves and to take a broader
ing to oneself than the more ubiquitous con human perspective on their negative expe
struct of self-esteem, as it provides a more riences. Similarly, being conscientious may
stable foundation of positive self-regard that help individuals to pay greater attention to
is less ego reactive and contingent on exter their own needs and to respond to difficult
nal sources. In fact, self-compassion may situations in a responsible and nonreac
be a key source of the “optimal” or “true” tive manner. Importantly, however, self
self-esteem extolled by some theorists. Deci compassion still predicted positive psycho
and Ryan (1995) have proposed that some logical functioning after controlling for the
people possess “true self-esteem,” a self- “Big Five,” suggesting that self-compassion
determined and autonomous way of evaluat is not reducible to these personality traits.
ing oneself that is not dependent on particu It should also be remembered that the direc
lar outcomes or social approval. Similarly, tionality of the link between these personal
Kernis (2003) has proposed the concept of ity traits and self-compassion probably goes
“optimal self-esteem,” which is founded on both ways, and it is possible that developing
stable and noncontingent self-evaluations. greater self-compassion leads to a healthier
Self-compassion provides greater self-worth personality (e.g., lessened neuroticism).
stability and noncontingency than trait self Early family experiences are also likely to
esteem because its source is internal rather play a key role in the development of self
than external and because it avoids process compassion or lack thereof. Gilbert (2005)
es of self-judgment and evaluation altogeth suggested that self-compassion stems largely
er. For this reason, self-compassion does not from the attachment system, so that individ
require feeling “above average” or superior uals who are raised in safe, secure environ
to others, and it provides emotional stability ments and who experience supportive and
when facing up to personal inadequacies. validating relationships with caregivers are
more able to relate to themselves in a caring
and compassionate manner. In contrast, in
The Origins o f Self-Com passion dividuals who are raised in insecure, stress
ful, or threatening environments and who
Although there is evidence that self experience constant criticism and aggression
compassion is associated with psychological from caregivers tend to be self-critical rather
well-being and that individuals can be taught than self-compassionate (Gilbert & Proctor,
to be more self-compassionate, less is known 2006). This occurs because individuals with
about why people have greater or lesser lev insecure attachment relationships have an
els of self-compassion in the first place. Some insufficiently developed self-soothing system
variation may be due to broader personality and few internalized models of compassion
traits. In an examination of self-compassion to draw on. Also, children may develop de
and major personality traits as measured by fense mechanisms of self-attacking because
the N EO -FFI (Neff, Rude, & Kirkpatrick, it is too risky to blame powerful others for
20 0 7 ), it was found that self-compassion their punitive or neglectful behavior. Recent
had the strongest association with neuroti data collected with a sample of adolescents
cism (r = -.6 5 ), with greater self-compassion and young adults gives some tentative sup
570 V I. S E L F -R E L A T E D D IS P O S IT IO N S
port to these propositions. Neff and M cGe- ferences did not explain group differences
hee (2008) found that maternal criticism and in self-compassion. Self-construal theory
stressful family relationships were negatively has been used to argue that Asians are more
related to self-compassion among youths. self-critical than Westerners (Heine, Leh
Those who felt accepted and validated by man, M arkus, & Kitayama, 1999), which
their parents, on the other hand, reported implies that they should also be less self-
having greater self-compassion. Secure at compassionate. Because people with inter
tachment was positively associated with dependent self-construals are more invested
self-compassion, whereas preoccupied or in conforming the self’s behavior to the re
fearful attachment was negatively linked to quirements of social relationships, the theory
self-compassion, supporting the notion that goes, they tend to constantly criticize them
attachment schemas play a role in the abil selves in order to keep themselves in line.
ity to be self-compassionate. Another way This did not appear to hold true for Thais,
that familial environments might affect the however, who had almost identical levels of
development of self-compassion is the extent interdependent self-construal as did Taiwan
to which parents model self-compassionate ese yet were much more self-compassionate.
versus self-critical reactions to their own fail Moreover, the link between self-construal
ures or life difficulties, although this propo and self-compassion itself varied across cul
sition has yet to be empirically examined. tures. Self-compassion was associated with
interdependent self-construal in Thailand
but with independent self-construal in Tai
C ross-C ultural Variations wan and the United States. This suggests
in Self-Com passion that the meanings of independence and inter
dependence may vary across cultures. Inter
There has been a small amount of research dependence involves being deeply embedded
exploring whether self-compassion differs in a particular social system. If that system
across cultures. N eff and colleagues (2008) promotes the value of self-compassion, as
examined self-compassion, independent and it does in Thailand, then being more inter
interdependent self-construal, and psycho dependent within that system may promote
logical well-being in Thailand, Taiwan, and self-compassion and decrease self-judgment.
the United States. Mean self-compassion If the culture does not actively promote self
levels were highest in Thailand and lowest compassion, however, which appears to be
in Taiwan, with the United States falling in the case in Taiwan and the United States,
between (all cultures differed significantly being independent of the prevailing cul
from one another, although within-culture tural ethos may facilitate the type of self-
variations in self-compassion were as great understanding and self-care required to be
as between-culture variations). These cross- compassionate toward oneself. In all three
cultural differences may be explained by the cultures, however, greater self-compassion
fact that Thais are strongly influenced by significantly predicted less depression and
Buddhism and that the value of compassion greater life satisfaction, suggesting that there
is emphasized in parenting practices and may be universal benefits to self-compassion
everyday interactions in Thailand. In con despite cultural differences in its prevalence.
trast, the Taiwanese are more influenced by
Confucianism, and shame and self-criticism
is more strongly emphasized as a means Sum m ary and Conclusions
of parental and social control in Taiwan.
Americans may have reported middling lev Self-compassion is a relatively new construct
els of self-compassion because U.S. culture in personality and social psychology, but the
displays more mixed messages with regard data gathered so far suggest that the ability
to self-compassion (e.g., a strong emphasis to be self-compassionate is linked to greater
on positive self-affect but also an isolating, emotional resilience and psychological well
competitive ethos). being. Self-compassionate people are less de
Interestingly, cross-cultural differences pressed and anxious, have better emotional
remained even when controlling for self- coping skills, are less afraid of failure, are
construal, suggesting that self-construal dif more intrinsically motivated to learn and
38. S elf-C o m p assion 571
grow, are happier, are more curious and is aimed at the management of chronic pain
wise, and feel more connected to others. Im and the treatment of stress disorders (Gross
portantly, these mental health benefits are man, Niemann, Schmidt, & Walach, 2004).
not obtained through a process of judging Although M BSR training primarily focuses
or evaluating the self— by stuffing oneself on teaching mindfulness skills, it also teach
into a box labeled “good” versus “bad.” es meditation practices aimed at developing
This type of self-evaluation often requires compassion for self and others and has been
comparing oneself to others, with cognitive shown to increase self-compassion among
distortion being used to overevaluate one’s participants (Shapiro, Astin, Bishop, &c
own competencies and underevaluate those Cordova, 20 0 5 ; Shapiro, Brown, & Biegel,
of others (Taylor & Brown, 1988). The need 20 07). Future research should be aimed at
to feel special and above average can lead determining whether self-compassion can
to increased feelings of isolation and sepa be successfully taught to children and ado
ration from fellow humans and is counter lescents in the schools, as it might provide
productive to feelings of interconnectedness. youths with greater emotional resilience
With self-compassion, however, the bound when facing the problems and difficulties of
aries between self and other are softened. living a human life.
All human beings are worthy of compassion,
oneself included. Thus self-compassion is a
useful alternative to self-esteem when con R e fe re n ce s
ceptualizing healthy forms of self-relating.
It provides similar mental health benefits to Adams, C. E ., & Leary, M . R. (2 0 0 7 ). Promoting
self-com passionate attitudes toward eating among
high self-esteem without being linked to pat
restrictive and guilty eaters, jo u r n a l o f S o c ia l a n d
terns of narcissism, social comparison, ego C lin ica l P sy ch olog y , 2 6 , 1 1 2 0 -1 1 4 4 .
defensiveness, or self-worth contingency and Baer, R. A. (2003). M indfulness training as a clinical
instability that have been associated with intervention: A conceptual and empirical review.
self-esteem pursuit (Crocker & Park, 2004). C lin ical P sy ch olog y : S cien ce a n d P ractice, 10,
1 2 5 -1 4 3 .
Even if the quest for high self-esteem were
Baum eister, R. F., Cam pbell, J . D., Krueger, J. L, &
not so potentially problematic, it has proved Vohs, K. D. (2 0 0 3 ). Does high self-esteem cause bet
difficult to raise people’s levels of self-esteem ter perform ance, interpersonal success, happiness,
in any case (Baumeister, Campbell, Krueger, or healthier lifestyles? P s y c h o lo g ic a l S c ien c e in the
& Vohs, 2003). The reason is partly that P u b lic In terest, 4, 1 -4 4 .
Baum eister, R. F., Sm art, L ., & Boden, J . M . (1996).
people with low self-esteem identify with
Relation of threatened egotism to violence and ag
their perceived lack of competence and often gression: The dark side of high self-esteem. P sy c h o
prefer to verify and maintain their identi lo g ic a l R ev iew , 103, 5 - 3 3 .
ties rather than engaging in the positive Blaine, B ., & C rocker, J. (1993). Self-esteem and self-
self-illusions that are common among those serving biases in reactions to positive and negative
events: An integrative review. In R. E. Baumeister
high in self-esteem (Swann, 1996). It may be (Ed.), S e lf-es te em : T h e p u z z le o f lo w se lf-reg a rd
more possible to raise people’s levels of self (pp. 5 5 - 8 5 ). H illsdale, N J: Flrlbaum.
compassion, given that it requires them to B latt, S. J . (1995). Representational structures in psy
merely acknowledge and accept their human chopathology. In D. Cicchetti & S. Toth (Eds.),
limitations with kindness, rather than chang R o c h e s t e r sy m p o siu m on d e v e lo p m e n t a l p s y c h o
p a t h o lo g y : E m o tio n , c o g n itio n , a n d re p re se n ta tio n
ing their self-evaluations from negative to (Vol. 6 , pp. 1 -3 4 ). Rochester, NY: University of
positive. Research demonstrates that self Rochester Press.
compassion can be enhanced in the short and Brach, T. (2 0 0 3 ). R a d ic a l a c c e p ta n c e : E m b r a cin g y o u r
long term (Gilbert & Proctor, 2 0 0 6 ; Leary life w ith th e h e a r t o f a B u d d h a . New York: Bantam
et al., 200 7 ; Neff, Kirkpatrick, & Rude, Books.
Brow n, K. W ., & Ryan, R. M . (2 0 0 3 ). T he benefits of
20 0 7 ), suggesting that programs designed to being present: M indfulness and its role in psycho
increase self-compassion have some chance logical well-being, j o u r n a l o f P erson ality a n d S o
of success. In fact, there is increasing interest c ia l P sy ch olog y , 8 4 , 8 2 2 - 8 4 8 .
in mindfulness-based interventions for their Cam pbell, W. K ., & Baum eister, R. F. (2001). Is loving
the self necessary for loving another?: An exam i
ability to reduce stress and improve men
nation of identity and intimacy. In M . Clark 8c G.
tal health (Baer, 2003). Mindfulness-based Fletcher ( Eds.), B la c k w e ll h a n d b o o k o f s o c ia l p sy
stress reduction (M BSR; Kabat-Zinn, 1982) c h o lo g y : Vol. 2. In te r p e r s o n a l p r o c e s s e s (pp. 4 3 7 -
is the best known of these interventions and 456). London: Blackwell.
572 V I. S E L F -R E L A T E D D IS P O S IT IO N S
Self-Monitoring
Pa u l T . F u g l e s t a d
M a r k Sn y d e r
574
39. S elf-M o n ito rin g 575
the researchers could ascertain whether par Nelson (1981) has noted two distinct pat
ticular phenomena (e.g., attitude-behavior terns in language acquisition by children ages
consistency) were associated with particu 2 -3 . Some children, labeled referential, first
lar dimensions (e.g., Other-Directedness) or learn to use language as a referential system
whether diverse phenomena would cluster for the communication of happenings in the
together and load most highly on the gen world. Other children, labeled expressive,
eral self-monitoring axis. O f the nine groups learn early on that linguistic expressions are
of criteria, seven clustered around the self dependent on social context and that lan
monitoring axis. The axis defined by the av guage can be used to obtain attention from
erage of these seven categories was displaced others. It is possible that these early differ
only 3° from the self-monitoring axis. Fur ences in language acquisition patterns reflect
thermore, 78% of the criterion variables cor the same underlying trait that manifests it
related most highly with the self-monitoring self in adult self-monitoring orientations.
axis, 13% with the extraversion axis, and Graziano, Leone, Musser, and Laut-
9% with the other-directedness axis. Clear enschlager (1987) developed the Junior
ly, the results suggest that self-monitoring Self-Monitoring Scale to differentiate self
represents a unified construct related to a monitoring propensities in children (see also
wide range of self-presentational behaviors. Eder, 1987, for a self-monitoring scale to be
used with children as young as age 3). Like
the original Self-Monitoring Scale, items
The Origins o f Self-M onitoring were designed to tap five related domains
of self-presentation and expressive behav
Flow do differences in self-monitoring ior. In one study they found that high self
emerge? Are high and low self-monitors monitoring children were more likely to ex
born or made? Is self-monitoring a prod amine social comparison information than
uct of culture? Such questions address self were low self-monitoring children while
monitoring’s origins. completing an opinion survey and anticipat
ing a discussion with other students.
Musser and Browne (1991) examined the
Self-M onitoring in Childhood
stability of self-monitoring scores for first-,
Gangestad and Snyder (1985) suggest that third-, and fifth-graders over 15 months, as
self-monitoring differences in adulthood well as relations of self-monitoring to other
may partially emerge because of differences personality variables, competence, and peer
in childhood communication patterns and/or acceptance. They found that self-monitoring
receiving differential treatment from adults scores were stable over time for each age
in childhood. They propose that initially group. Similar to the findings for adults,
small differences between individuals be self-monitoring was modestly related to ex
come amplified over time, producing larger traversion, but not to other aspects of per
differences (i.e., divergent causality; Meehl, sonality. For boys, especially older boys, self
1978). These differences have a substantial monitoring was positively related to number
genetic component (Dworkin, 1977; Gang of friends, peer popularity, and self-esteem;
estad, 1984), and the development of early self-monitoring was not reliably related to
self-monitoring orientations may be sensitive these measures for girls. The positive asso
to some specific environmental parameter, ciation of self-monitoring with popularity
such as attention received from caretakers. and self-esteem for boys may reflect the gen
It could be that children who receive little at der appropriateness of high self-monitoring
tention from their caretakers develop a strat behavior. Boys tend to associate in large
egy designed to gain the attention and regard groups, but girls tend to have more intimate,
of others— a high self-monitoring strategy. two-person friendships over longer periods
Predispositions may influence the effects of of time (Rubin, 1980). If children’s self
environmental factors on self-monitoring. monitoring friendship patterns are similar
For example, for some individuals, a lack of to those of adults in that high self-monitors
parental attention may need to be somewhat tend to engage in particular activities with
extreme to influence the adoption of a high particular people and low self-monitors tend
self-monitoring strategy. to engage in most activities with the same
39. S elf-M on itorin g 577
well-liked people (Snyder, Gangestad, & et al., 1989; Gudykunst, Yang, 8c Nishida,
Simpson, 1983), then high self-monitoring 1987). Although it appears that strategic
would be more gender appropriate for boys self-presentation as captured by the Self-
than for girls. This could be one possible ex Monitoring Scale is somewhat more preva
planation for the consistent trend for men lent in individualistic cultures, future studies
to score slightly higher than women on the are needed to examine developmental and
Self-Monitoring Scale (Day, Schleicher, Un- behavioral differences in self-monitoring
ckless, & Hiller, 2 0 0 2 ; Snyder, 1987). across cultures.
can more effectively assume the mannerisms ior but also on the motivational context of
of a typical extravert and introvert. the interaction (certain participants were
High self-monitors, relative to low self instructed to either gain affection or win
monitors, project a general self-presentation respect; Jones 8c Baumeister, 1976). On the
that is more friendly and outgoing and less other hand, low self-monitors generally ac
worried and anxious (Lippa, 1976). Onto cepted people’s behavior at face value, irre
this general background, high self-monitors spective of the motivational context.
project specific images that are tailored to
situational features and, therefore, display
Self- Conceptions
more behavioral variability across situations
than do low self-monitors (Lippa 8c Don Self-monitoring orientations relate to basic
aldson, 1990; Snyder, 1979). In studies of understandings of what constitutes a “self.”
social interaction, high self-monitors, more High self-monitors regard themselves as
so than low self-monitors, choose clearly de adaptive individuals who pragmatically tai
fined situations, use scripts for typical situ lor their expressive behavior to features of
ations, formulate effective plans of action interpersonal contexts. They tend to attri
before social interaction, use other people’s bute their own behavior to situational influ
behavior as guides to their own behavior, ences and to define their identities in terms of
and actively facilitate smooth and pleasing situational features (Sampson, 1978; Snyder,
conversations (e.g., Douglas, 1983; Ickes 1976). On the other hand, low self-monitors
8c Barnes, 1977; Jordan 8c Roloff, 1997; think of themselves as principled individu
Snyder & Gangestad, 1982). In an exam i als who value consistency between who they
nation of impression-management tactics, are and what they do. They tend to attribute
Bolino and Turnley (2003) found that high their own behavior to dispositional influ
self-monitors were likely to be classified as ences and to define their identities in terms
positive impression managers (i.e., using in of enduring dispositions (Sampson, 1978;
gratiation, self-promotion, and exemplifica Snyder, 1976).
tion, but not supplication and intimidation), When deciding how to express and pres
whereas low self-monitors were likely to be ent oneself in a social situation, the proto
classified as aggressive (i.e., using all tactics typical high self-monitor is thought to ask,
to some extent) or passive impression man “Who does this situation want me to be and
agers (eschewing self-presentational tactics). how can I be that person?” In this way, the
behavior of high self-monitors is shaped by
knowledge of the prototypical person called
Social-Cognitive Processes
for by a given situation. For example, when
High and low self-monitors also diverge in faced with a party, a high self-monitor may
social-cognitive and attributional processes. conjure up the image of an “ideal extravert”
For example, in a study in which participants and engage in expressive behavior that is
were allowed to observe a potential dating outgoing, witty, and friendly. In contrast,
partner, high self-monitors were more likely when deciding how to express and present
than low self-monitors to remember infor oneself in a social situation, the prototypi
mation about their partners and to infer their cal low self-monitor is thought to ask, “Who
partners’ dispositions; that is, they were able am I and how can I be me in this situation?”
to construct clear images of their partners In this way, the behavior of a prototypical
(Berscheid, Graziano, M onson, 8c Dermer, low self-monitor is shaped by knowledge of
1976). Furthermore, high self-monitors pay his or her self-conceptions and typical ways
particular attention to social comparison in of behaving in a given situation. In support
formation, are more responsive to expecta of these conceptions of self-monitoring pro
tions, and are more likely to look to others cesses, Snyder and Cantor (1980) found that
for guidance in unfamiliar social situations low self-monitors have richer and better
(e.g., Harris 8c Rosenthal, 1986; Rarick, articulated self-images in a wide variety of
Soldow, 8c Geizer, 1976; Snyder, 1974). In social situations, whereas high self-monitors
a study of evaluations of people in dyadic have richer and better articulated images
interactions, high self-monitors based judg of prototypic persons in the same situa
ments not only on a person’s actual behav tions. For both high and low self-monitors,
39. S elf-M o n ito rin g 579
faction (e.g., Norris & Zweigenhaft, 1999; 1989), low self-monitors should be more
Snyder &c Simpson, 1984), in sexual behav likely to develop trust in their romantic rela
ior in short-term relationships (e.g., Snyder, tionships, because they prefer partners with
Simpson, & Gangestad, 1986), in relation traits that foster these attributes (e.g., kind
ship longevity (e.g., Leone & Hall, 2 0 0 3 ; ness, honesty, loyalty; Jones, 1993). High
Snyder & Simpson, 1984), and in reactions self-monitors, on the other hand, desire at
to relationship dissolution (e.g., Snyder & tributes such as social status and physical
Simpson, 1984), among other relationship attractiveness to a greater extent (Jones,
processes. 1993; Snyder et al., 1985). Indeed, Norris
In a study by Snyder and colleagues (1985), and Zweigenhaft (1999) found that low-self-
men examined the profiles (which contained monitoring couples had greater trust than
information about personality and attitudes, did high-self-monitoring couples and were
as well as photos) of potential dating part more likely to see their relationships devel
ners. High self-monitors spent more time oping into marriage.
examining the photos, whereas low self In marriage, there is no assortative mat
monitors spent more time examining the ing with respect to self-monitoring (Leone
personal information. In a study that forced & Hawkins, 2 0 0 6 ; Snyder, 1987). How
a choice between a possible date with a good ever, similar to dating relationships, self
personality and an unattractive appearance monitoring orientations influence marital
and a date with a bad personality and an at satisfaction and functioning (Leone & Hall,
tractive appearance, high self-monitors chose 2 0 0 3 ; Leone & Hawkins, 2006). Low self
the attractive date and low self-monitors monitors, relative to high self-monitors, re
chose the date with the good personality port greater consensus in matters such as
(Snyder & Simpson, 1984). In an exam ina finances and religion, greater partner en
tion of dating motivations, low self-monitors gagement in positive relationship activities
reported greater desire for similar values, such as calm problem solving, greater dis
honesty, loyalty, and kindness in a potential plays of affection to their partners, greater
dating partner, whereas high self-monitors investment of resources, more commitment,
expressed greater desire for attractiveness, more psychological intimacy, and greater
sex appeal, and social status (Jones, 1993). marital satisfaction. Moreover, currently
The patterns of low and high self-monitors married high self-monitors are more likely
in romantic relationships reflect two distinct than currently married low self-monitors to
orientations toward close relationships— have been divorced.
the restricted sociosexual orientation of low
self-monitors and the unrestricted sociosex
C onsum er Behavior
ual orientation of high self-monitors (Snyder
et al., 1986). In romantic relationships, low Self-monitoring processes have also been
self-monitors strive to cultivate intimate, linked to consumer decision making and
long-term relationships with partners cho behavior (e.g., DeBono, 2 0 0 6 ; DeBono &
sen on the basis of dispositional compatibil Snyder, 1989). Advertisements for consumer
ity. In contrast, high self-monitors are more products tend to emphasize one of two broad
willing to “play the field” of potential dat themes— images associated with the product
ing partners. High self-monitors view love or utilitarian features of the product itself
as a social game and believe there may be (Fox, 1984). Because high self-monitors are
more than one person whom they can love concerned with image projection and the so
(Neto, 1993; Snyder, 1987). In contrast, low cial appropriateness of their behavior, they
self-monitors are more likely to view love as should be responsive to advertisements that
a psychologically close and emotionally in emphasize desired images. Conversely, be
tense endeavor or as a quest to find a similar cause low self-monitors value consistency
other who will be a compatible life partner; between believing and doing, they should
they are also more likely to believe that there be responsive to advertisements that em
is only one person who is ideally suited to phasize a product’s actual qualities. In stud
them (Neto, 1993; Snyder, 1987). ies of self-monitoring and advertising, high
Because trust is often a function of pre self-monitors generally find image-based ad
dictability and faith (Holmes & Rempel, vertisements more appealing, whereas low
582 V I. S E L F -R E L A T E D D IS P O S IT IO N S
self-monitors generally respond more favor egy), whereas low self-monitors based their
ably to quality-based appeals (e.g., Snyder judgments solely on the actual taste of the
& DeBono, 1985). However, as discussed cheese (a performance- based strategy).
by DeBono (2006), the results of these and Similarly, high self-monitors are influenced
subsequent studies paint a more complex by packaging attractiveness, whereas low
picture. For example, Snyder and DeBono self-monitors are not (DeBono, Leavitt, &
(1985) found that the image-versus-quality Backus, 2 0 0 3 ; DeBono &C Snyder, 1989).
results were stronger for some products (e.g., Aaker (1999) has demonstrated that low
Canadian Club whiskey) than others (e.g., self-monitors are more likely than high self
Irish M ocha M int coffee). Other research monitors to prefer products with brand
has found little or no support for the asso images (e.g., exciting, rugged) that match
ciation between self-monitoring and adver their own self-images. In contrast, high self
tising appeals (e.g., Bearden, Shuptrine, & monitors are more likely to prefer products
Teel, 1989). with brand images that match hypotheti
In attempting to account for these dis cal social situations (e.g., preferring rugged
crepant findings, DeBono (2006) suggests brands for an informal barbecue after river
that information processing plays a role in rafting). It appears that low self-monitors
the persuasiveness of advertising appeals. prefer products that consistently reflect who
The elaboration likelihood model of persua they are, whereas high self-monitors prefer
sion (Petty &C Cacioppo, 1986) posits that to match specific products to specific situa
the persuasiveness of a given message will tions. Using a theatrical metaphor, high self
depend on the strength of the arguments monitors prefer to have the right props for
as well as the ability and motivation of the the role.
recipient to process the message. One fac Other research suggests that high self
tor that can influence motivation to process monitors, relative to low self-monitors,
is whether or not the content of a message prefer products that match their own self-
matches the functional basis of individuals’ images when the products are used in pub
attitudes (Lavine & Snyder, 2 0 0 0 ; Petty & lic (e.g., Chevy Camaros, Reebok shoes) as
Wegener, 1998). When there is a match, in opposed to private settings (e.g., R ead er’s
dividuals are more motivated to process mes Digest, Budweiser beer) (Graeff, 1996). The
sages, and argument quality will play a more discrepancy between this finding and stud
crucial role in persuasion than when there is ies by Aaker (1999) may be explained by the
not a match. In support of these ideas, Petty fact that products such as the car one drives
and Wegener (1998) found that when partic are relatively invariant to the situation, and
ipants were exposed to ads that functionally high self-monitors should prefer such items
matched (i.e., high self-monitors exposed to the extent that they convey desired im
to image-based ads; low self-monitors ex ages. On the other hand, for items that can
posed to quality-based ads), the strength of be consumed or used in a wide variety of
the arguments played a crucial role in deter contexts (e.g., beverages), high self-monitors
mining the effectiveness of the ads. Specifi may base usage decisions on relevant situ
cally, high self-monitors reacted favorably to ational cues.
image-based ads with strong arguments and
unfavorably to image-based ads with weak
E m ploym ent Decisions
arguments, and low self-monitors reacted
and O rganizational Behavior
favorably to quality-based ads with strong
arguments and unfavorably to quality-based Self-monitoring figures prominently in
ads with weak arguments. work-related behaviors and decisions. Just
Beyond responses to advertising, self as high self-monitors focus on the appear
monitoring is also related to product evalu ances of prospective romantic partners, they
ation strategies and the influence of brand also use appearance in making employment
and self-concept matching on product judg decisions (e.g., Jawahar & M attsson, 2005;
ments. For example, DeBono and Rubin Snyder, Berscheid, & Matwychuk, 1988).
(1995) found that high self-monitors based In one study, participants made hiring deci
their quality judgments of cheese solely on sions for either a sales clerk or camp coun
the country of origin (an image-based strat selor (Snyder et al., 1988). The information
39. S elf-M o n ito rin g 583
Research by Flynn and Ames (2006) has cal issues in organizations, provided they
shown that high-self-monitoring women have the right principles and values. From
outperform low-self-monitoring women in the perspective of the individual, high self
settings that are stereotypically masculine. monitors may be in a better position than
Because many performance contexts in or low self-monitors to obtain information and
ganizations are gender stereotyped in that opportunities within organizations because
men are expected to outperform women they tend to occupy central positions in
(Eagly & Karau, 2002), they hypothesized their social networks at work (Mehra et al.,
that high-self-monitoring women should 2001). Having access to information could
be better able to counteract the negative contribute to “making sense” of the work
expectations of others by cultivating social place, which could lead to “getting along”
images that convey competence. High self and “getting ahead.”
monitoring should be less advantageous
for men in these situations because gender-
typed behaviors are the norm and because The Evolution o f Self-M onitoring
others may be inclined to perceive them in T h eory and R esearch
expectation-congruent ways. Additionally,
to be viewed as competent, women must act Originally, self-monitoring theorizing pro
in agentic ways (e.g., being assertive, control posed that high self-monitors have a concern
ling, confident), but this behavior violates for the social appropriateness of their expres
gender norms and may result in backlash sive behavior and that low self-monitors do
from others (Eagly & Karau, 2002). Because not have such a concern for social appropri
women are faced with self-presentational di ateness but are instead driven from internal
lemmas in the workplace, the ability to man states and dispositions. Although concern
age conflicting expectations is vital for suc for social appropriateness has been at the
cessful performance. In Flynn and Ames’s heart of self-monitoring theorizing, self
first study, small groups of M BA students monitoring should not be confused with need
completed a semester-long project. At the for social approval (Crowne &c Marlowe,
end of the semester, high self-monitoring 1964), as evidenced by small correlations
women were seen as more influential and between the Self-Monitoring Scale and the
more valuable to the group than low self M arlow e-Crow ne Social Desirability Scale
monitoring women; self-monitoring did not (in the - .2 range) and differential prediction
predict group evaluations for men. M ore of criterion behaviors such as emotional ex
over, in a mixed-sex negotiation task, high pression and attention to social comparison
self-monitoring women outperformed low information (e.g., Snyder, 1974). Hence the
self-monitoring women. As a whole, these concern for social appropriateness of the
results suggest that the self-presentational high self-monitor is distinct from defensive
flexibility and skills of the high self-monitor posturing to avoid disapproval. Early self
are quite beneficial for women working in monitoring theorizing did not precisely and
masculine-typed settings. explicitly address the motivations behind
Day and Schleicher (2006) also speculate self-monitoring propensities. However, as
on Hogan’s (1991) third broad motivational theory and research on the self-monitoring
aspect of “making sense.” The relatively construct have progressed over the past
more variable behavior of high self-monitors 30 years, so, too, have conceptions of self
could be seen as unpredictability by others. monitoring processes and ideas regarding
As such, a high self-monitor who is in a top- the motivational underpinnings of both low-
level position could be seen as an inconsis and high-self-monitoring orientations.
tent leader, which could undermine work In considering the evolution of self
ers’ trust and commitment. Additionally, if monitoring, we first present Gangestad
high self-monitors are ethically pragmatic, and Snyder’s (2000) reappraisal of the self
they may have difficulty in consistently deal monitoring construct. We then discuss sev
ing with ambiguous ethical dilemmas. Low eral lines of empirical work that directly re
self-monitors, drawing on principles and late to Gangestad and Snyder’s analysis of
values, may more effectively deal with ethi self-monitoring with regard to status-related
39. S elf-M o n ito rin g 585
concerns and motivations. Last, we consider As discussed, seven of nine criterion cat
a recent theoretical view emphasizing the egories were related to self-monitoring and
motivational role of affect in self-monitoring can help to clarify the construct of self
processes. monitoring. O f those, four categories were
most strongly related to self-monitoring:
behavioral variability, sensitivity to expecta
Self-M onitoring R eappraisal
tions and other cues, interpersonal orienta
Beyond addressing the construct validity tions, and responses to physical appearance.
of self-monitoring, Gangestad and Snyder’s Together, these criteria suggest that the
(2000) quantitative review further refined specific impression management engaged in
the construct by showing what phenomena by high self-monitors involves actively con
are unrelated and what phenomena are most structing public appearances that lead to
closely related to self-monitoring. In regard favorable outcomes and social cachet (e.g.,
to specifying what self-monitoring is not, knowing what image to project based on
two areas of inquiry were not related to self situational cues, choosing friends based on
monitoring. First, self-peer ratings of traits playing a role). Additionally, other catego
did not vary by self-monitoring; that is, ries of phenomena related to self-monitoring
agreement between self-ratings of one’s traits include expressive control and nonverbal de
and peer ratings of traits were similar for low coding skills, both of which are skills useful
and high self-monitors. This suggests that to image cultivation, and attitude-behavior
self-monitoring cannot resolve the person- consistency, which involves the correspon
situation debate as it specifically pertains to dence between public images and private be
trait consistency. However, as evidenced by liefs. As a whole, these criteria relate to image
the empirical review and prior research, self management and cultivation, concerns that
monitoring does relate to attitude-behavior have been central to self-monitoring theory.
consistency and behavioral variability. The Gangestad and Snyder (2000) further
idea that high self-monitors would show suggested that high self-monitors engage
consistency in traits is not antithetical to in assim ilative (vs. accommodative; Ickes,
self-monitoring theory. As argued by Snyder Reidhead, & Patterson, 1986) and acquisi
(1979) and supported by research (Lippa, tive (vs. self-protective; Wolfe, Lennox, &
1976, 1978; Snyder & Monson, 1975), high Cutler, 1986) forms of self-presentation to
self-monitors show cross-situational con enhance status (i.e., one’s social and occupa
sistency in background self-presentation, tional standing) and effectively operate with
appearing friendly and nonanxious, and in perceived hierarchical social structures
cross-situational variability in foreground (i.e., perceptions of relative influence and/or
self-presentation based on the specifics of a power in relationships or groups). Research
given interaction or situation. on the interpersonal orientations of high and
Another set of criterion behaviors that are low self-monitors supports this view— high
unrelated to self-monitoring are impression- self-monitors, relative to low self-monitors,
management behaviors involving close atten generally have less committed and stable so
tion and responsivity to others. People who cial relationships, choose friends based on
engage in this type of impression manage activity expertise, and desire high-status ro
ment are socially anxious, concerned about mantic partners (e.g., Jones, 1993; Snyder et
negative social evaluation, and seek to ap al., 1983, 1986; Snyder & Simpson, 1984).
pease others. It appears as though vigilantly High self-monitors may be more invested
attending to others and being highly respon in negotiating status within unequal-status
sive to others’ behavior are tendencies as social structures, whereas low self-monitors
sociated with a defensive self-presentational may be more invested in establishing equal-
style, which is not related to self-monitoring. status bonds in which they are free to be
Those with defensive self-presentational themselves and in which trust can develop.
styles may adapt to others, but, unlike a high Furthermore, low self-monitors, instead of
self-monitor, they engage in restrained and being unconcerned with public images, may
appeasing social behavior that may be rela actively try to cultivate images of genuine
tively consistent across situations. ness and sincerity.
586 V I. S E L F -R E L A T E D D IS P O S IT IO N S
E m pirical Investigations o f Status Concerns shaking, the other person will tend to engage
in these same behaviors. Importantly, prim
Perceptions o f Status and Status Cultivation
ing an affiliation goal (either consciously or
in Exchange Relationships
nonconsciously) leads peoples to engage in
In a direct examination of self-monitoring more mimicry (Lakin & Chartrand, 2003).
and status concerns, Flynn and colleagues Because high self-monitors tend to be more
(2006) found that high self-monitors report responsive to situational cues and desire to
ed greater desire than did low self-monitors make favorable impressions and cultivate
for social status and were more accurate in status, they should be more likely to engage
perceiving status in dyadic exchange relation in nonconscious mimicry when situations
ships. In addition, high self-monitors were call for affiliation.
perceived by former coworkers to have had Cheng and Chartrand (2003) examined
higher social status and to have displayed mimicry in dyadic interactions in which
greater generosity than low self-monitors. participants’ interaction partners were os
A mediational analysis suggested that high tensibly either high school students, fellow
self-monitors were able to achieve status, at psychology students, or graduate students.
least in part, by helping others in workplace Because a psychology student is someone
networks. In a study of exchange relations with whom participants could have future
of M BA students, each student indicated to contact, affiliation should be relevant. In
whom he or she would go for help or advice this study, high self-monitors engaged in
and who would be most likely to come to more mimicry than low self-monitors, but
him or her for help and advice. Additionally, only when they thought they were interact
students provided ratings of who would go ing with a psychology student. However,
to whom for help among a subset of their high self-monitors did not pay particular at
peers. High self-monitors were particularly tention to the psychology student, were not
accurate in their perceptions of exchange consciously aware of foot shaking, and did
relationships (i.e., they knew who would go not explicitly report a greater desire to af
to whom for help), less likely seek out help, filiate. A second study examined mimicry in
and more likely to be sought out for help. the context of leader-worker interactions.
Additionally, high self-monitors were likely High self-monitors engaged in more mim
to occupy positions of higher status (i.e., icry when interacting with a more powerful
being sought for help from someone, but not versus less powerful other; low self-monitors
seeking help from that person). By more ac did not vary by condition. This finding pro
curately perceiving exchange relationships vides further evidence for the contention that
and developing a positive reputation and so high self-monitors are attuned to and moti
cial standing, high self-monitors may be in vated by social status (Gangestad & Snyder,
a better position than low self-monitors to 2000). By engaging in conscious processes
obtain resources and influence others in the such as help giving and nonconscious pro
workplace. cesses such as behavioral mimicry, high self
monitors are better able to facilitate smooth
social interactions and cultivate desired im
Status and Nonconscious Mimicry
ages.
Although the interaction behavior of high
self-monitors is often quite deliberate and
Attractiveness and Self-Evaluation
effortful (see Ickes, Holloway, Stinson, &
Hoodenpyle, 2 0 0 6 , for a discussion), other Physical appearance is an important and
research suggests that high self-monitors salient feature of high self-monitors’ so
also engage in nonconscious behaviors cial worlds (e.g., Snyder et al., 1985, 1988).
that facilitate social interactions and self- Low self-monitors, however, seem relatively
presentational goals (Cheng & Chartrand, unconcerned with physical appearance in
2003). People tend to nonconsciously engage the ways they view and structure their so
in behavioral mimicry (Chartrand & Bargh, cial worlds. Accordingly, why are high self
1999). For example, if one person in a dyadic monitors so attentive to and influenced by
interaction engages in face touching or foot the physical appearance of others? Fuglestad
39. S elf-M on itorin g 587
and Snyder (2005) suggest that physical (e.g., Gray-Little & Burks, 1983), an attenu
appearance is an important instrument in ation of this association was expected for
the facilitation of social status and should high self-monitors. In line with these hy
be an important aspect of self to high self potheses, low self-monitors perceived greater
monitors but of lesser importance to low power equality than did high self-monitors;
self-monitors. they also reported greater relationship qual
A meta-analysis by Langlois and col ity in symmetrical relationships and less re
leagues (2000) demonstrated that physically lationship quality in asymmetrical relation
attractive people are treated more favorably ships. For high self-monitors, the negative
(have less negative and more positive inter effects of power asymmetries were attenu
actions) and achieve more occupational and ated. Thus power and influence asymmetries
interpersonal rewards (e.g., greater income, may be more of a happenstance for high self
more popularity) than unattractive people. monitors or even an integral feature of their
Given that physical attractiveness has these relationships. And high self-monitors’ toler
implications for how one is treated by oth ance for asymmetries could lead to long-term
ers, it follows that attractiveness would af relationships less satisfying than those of low
fect how one feels about oneself. However, self-monitors. Low and high self-monitors in
overall correlations between attractiveness shorter relationships did not differ in their
and self-evaluations are small (Langlois et perceptions of asymmetries in the relation
al., 2000 ). Because high self-monitors place ship. However, in longer relationships, high
greater importance on physical appear self-monitors perceived more asymmetries
ance, the relation between attractiveness (results consistent with the finding that high
and self-evaluations should be stronger for self-monitors are less satisfied in their mar
them than for low self-monitors. Indeed, riages than are low self-monitors; Leone &
self-monitoring does moderate the relation H all, 2003).
between attractiveness and self-evaluations
(Fuglestad & Snyder, 2005). Attractiveness
T h e Role o f A ffect
and self-evaluations are positively related
in Self-M onitoring Processes
for high self-monitors but unrelated for low
self-monitors. Thus attractiveness appears In a review of self-monitoring processes in
to be a more important and salient aspect dyadic interactions, Ickes and colleagues
of self for high self-monitors than for low (2006) presented a perspective on the mo
self-monitors. tivations behind self-monitoring; they pro
pose that high self-monitors are motivated
to express and evoke positive affect in so
Balance o f Power in Romantic Relationships
cial situations and to regulate their affect
Oyamot, Fuglestad, and Snyder (in press) through impression management. The idea
examined balance of power and relation of monitoring the affective tone of an inter
ship quality in romantic relationships. They action is not antithetical to cultivating image
hypothesized that low self-monitors, out of and status; rather, it specifies a mechanism
a concern for cultivating intimate and trust to facilitate positive interactions and make
ing relationships, would perceive their rela favorable impressions. Indeed, research on
tionships as relatively egalitarian (i.e., the job performance suggests that high self
partners have equal influence) and that, for monitors are good at “getting along” and
them, balance of power would be a crucial “getting ahead” and that getting along facil
element in determining relationship qual itates getting ahead (e.g., Day & Schleicher,
ity (i.e., closeness, satisfaction, investment). 2 0 0 6 ; Flynn et al., 2006).
In contrast, they hypothesized that high In a study on the role of self-monitoring in
self-monitors would perceive greater power unstructured interactions, Ickes and Barnes
asymmetries in their relationships (i.e., one (1977) found that high self-monitors, relative
person having more influence) and might to low self-monitors, were more motivated
actually prefer hierarchical ones. Therefore, to make the interaction go well and report
although equal balance of power is generally ed being more self-conscious if the conver
associated with greater relationship quality sation did not go well. Furthermore, high
588 V I. S E L F -R E L A T E D D IS P O S IT IO N S
self-monitors also consciously develop self- have highly attuned sociometers in long
presentational plans and use clear scripts to term relationships. That is, the observed
enact appropriate self-presentation (Doug pattern of results in interaction studies
las, 1983; Jordan & Roloff, 1997). Ickes and could mean that low self-monitors are not
colleagues (2006) suggest that these findings concerned with superficial encounters and
are not simply due to a desire to facilitate a relationships, whereas high self-monitors
positive interaction but to the desire to culti are concerned with cultivating desired im
vate and maintain positive self-affect during ages in these brief encounters. Indeed, high
interaction. However, the primary motiva self-monitors do seem to benefit in the work
tion of high self-monitors may be to culti place from engaging in such encounters with
vate images and status, and the monitoring coworkers (Day & Schleicher, 2006).
of self-affect may serve as a guide to how Hoyle and Sowards (1993) have also dis
well that goal is being achieved. cussed the role of affect in self-monitoring. In
In support of the centrality-of-affect view, their view, high self-monitors should experi
high self-monitors tend to express and evoke ence positive affect to the extent that their be
more positive affect and less negative affect havior matches perceived standards of social
in dyadic interactions than low self-monitors appropriateness and negative affect to the ex
(Levine & Feldman, 1997; Lippa, 1978). In tent that their behavior does not correspond
terestingly, in the studies by Levine and Feld to these standards. On the other hand, low
man (1997), not only did high self-monitors self-monitors should experience positive af
express less negative affect and cultivate im fect to the extent that their behavior matches
ages of happiness, but they were also more their self-concepts and negative affect to the
successful in creating impressions of compe extent that it violates their self-concepts. Al
tence and likeability. As such, the expression ternatively, Gangestad and Snyder’s (2000)
and evocation of positive affect may be a reappraisal of self-monitoring suggests that
tool for the cultivation of desired images. high self-monitors should experience posi
Applying sociometer theory (Leary, tive affect to the extent that they are able to
Tambor, Terdal, & Downs, 1995) to self cultivate desired images (especially status-
monitoring, Ickes and colleagues (2006) related images), not merely to the extent that
suggested that high self-monitors possess their behavior is socially appropriate.
“sociometers” finely tuned to the effects of As pointed out by Ickes and colleagues
their self-presentations on others. Depend (2006), more research is needed to address
ing on the perceived effects of their expres these competing accounts of motivation and
sive behavior (acceptance and validation vs. affect in self-monitoring processes. Thus a
dismissal and rejection), high self-monitors sociometer account would suggest that the
will experience positive or negative affect. affective experiences of high self-monitors
On the other hand, low self-monitors’ “so should be more variable during social inter
ciometers” may be relatively insensitive to actions than those of low self-monitors. On
the effects of their expressive behaviors on the other hand, if the perspective of Hoyle
others. As noted by Ickes and colleagues, an and Sowards (1993) is correct, low and high
alternative account could be that low self self-monitors should be equally variable, but
monitors have sensitive “sociometers” in the affect of high self-monitors should vary
regard to self-presentation but that, instead with the appropriateness of their behavior
of responding to feedback regarding social (or their success in cultivating images and
appropriateness, they respond to feedback status), and the affect of low self-monitors
regarding self-verification. In this view, the should vary with the self-concept consistency
low self-monitor is not oblivious to or un of their behavior. Additionally, the relative
concerned about impression management, status of interaction partners should be cru
but he or she seeks to gain validation of his cially important to high self-monitors’ affec
or her own self-image and attempts to culti tive responses (e.g., being more responsive to
vate an image of genuineness. higher status partners), whereas the relative
Another interpretation of the self-monitoring closeness of interaction partners should be
and interaction literature is that low self critical to low self-monitors’ affective re
monitors have relatively insensitive sociom sponses (e.g., being more responsive with a
eters in b r ie f encounters with strangers but friend than a stranger).
39. S elf-M o n ito rin g 589
task groups and dyadic conflicts. J o u r n a l o f A p p lied and the regulation of social experience: A control-
P sy c h o lo g y , 91, 2 7 2 - 2 8 1 . process model. J o u r n a l o f S o c ia l a n d C lin ica l P sy
Flynn, F. J ., Reagans, R . E ., A m anatullah, E. T ., & c h o lo g y , 1 2, 2 8 0 - 3 0 6 .
Ames, D. R . (2 0 0 6 ). Helping one’s way to the top: Ickes, W ., & B arnes, R . D. (1 9 7 7 ). T he role of sex and
Self-m onitors achieve status by helping others and self-m onitoring in unstructured dyadic interactions.
knowing who helps whom. J o u r n a l o f P erson ality J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch olog y , 3S,
a n d S o c ia l P sy ch olog y , 9 1, 1 1 2 3 -1 1 3 7 . 3 1 5 -3 3 0 .
Fox, S. (1984). T h e m irr o r m a k e r s . New York: M o r Ickes, W ., Holloway, R ., Stinson, L. L ., Sc Hoodenpyle,
row. T. G. (2 0 0 6 ). Self-m onitoring in social interaction:
Fuglestad, P. T ., & Snyder, M . (2 0 0 5 , January). L o o k The centrality of self-affect. J o u r n a l o f P erson ality ,
in g g o o d a n d fee lin g g o o d : T h e r e la tio n sh ip o f s e lf 74, 6 5 9 - 6 8 4 .
m o n ito rin g to a ttr a c tiv en es s a n d self-e v a lu a tio n s . Ickes, W ., Reidhead, S., & Patterson, M . (1986). M a
Poster session presented at the annual meeting of chiavellianism and self-m onitoring: Different as
the Society for Personality and Social Psychology, “me” and “you.” S o c ia l C o g n itio n , 4 , 5 8 -7 4 .
New O rleans, LA. Jaw ah ar, I. M ., & M attsson, J . (2 0 0 5 ). Sexism and
Gangestad, S., & Snyder, M . (1985). “To carve nature beautyism effects in selection as a function of self
at its join ts”: O n the existence of discrete classes in m onitoring level of decision m aker. J o u r n a l o f A p
personality. P sy c h o lo g ic a l R e v iew , 9 2 , 3 1 7 -3 4 9 . p lie d P sy ch olog y , 9 0 , 5 6 3 - 5 7 3 .
Gangestad, S. W . (1 9 8 4). O n th e e t io lo g y o f in d iv id u a l Jon es, E. E ., & Baum eister, R . (1976). T he self-m onitor
d iffe r e n c e s in se lf-m o n ito r in g a n d e x p r es siv e self- looks at the ingratiator. J o u r n a l o f P erson ality , 4 4 ,
c o n tr o l: T estin g th e c a s e o f stron g g en e tic in flu 6 5 4 -6 7 4 .
en c e . Unpublished doctoral dissertation, University Jon es, M . (1993). Influence of self-m onitoring and dat
of M innesota. ing m otivations. J o u r n a l o f R e se a rc h in P erson ality ,
Gangestad, S. W ., & Snyder, M . (2 0 0 0 ). Self 27, 1 9 7 -2 0 6 .
m onitoring: Appraisal and reappraisal. P s y c h o lo g i Jord an , J . M ., & R oloff, M . E. (1 9 9 7 ). Planning skills
c a l B u lletin , 1 2 6 , 5 3 0 - 5 5 5 . and negotiator goal accom plishm ent. C o m m u n ic a
G onnerm an, M . E ., Jr., Parker, C . P., Lavine, H ., & tion R e se a rc h , 2 4 , 3 1 - 6 3 .
H uff, J. (2 0 0 0 ). T he relationship between self L akin , J . L ., & C hartrand , T. L. (2 0 0 3 ). Using noncon
discrepancies and affective states: T he m oderating scious behavioral m im icry to create affiliation and
roles of self-m onitoring and standpoints on the self. rapport. P s y c h o lo g ic a l S cien ce, 14, 3 3 4 -3 3 9 .
P erson ality a n d S o c ia l P sy ch o lo g y B u lletin , 2 6 , Langlois, J. H ., K alak an is, L ., Rubenstein, A. J .,
8 1 0 -8 1 9 . L arson, A ., H allam , M ., & Sm oot, M . (2 0 0 0 ).
G raeff, T. R . (1 9 9 6 ). Image congruence effects on M axim s or myths of beauty?: A m eta-analytic and
product evaluations: The role of self-m onitoring theoretical review. P s y c h o lo g ic a l B u lletin , 1 2 6 ,
and public/private consum ption. P sy c h o lo g y a n d 3 9 0 -4 2 3 .
M arketin g , 13, 4 8 1 - 4 9 9 . Lavine, H ., & Snyder, M . (2 0 0 0 ). Cognitive processes
G ray-Little, B ., & Burks, N. (1983). Power and satis and the functional m atching effect in persuasion:
faction in m arriage: A review and critique. P sy c h o Studies of personality and political behavior. In G.
lo g ic a l B u lletin , 9 3 , 5 1 3 - 5 3 8 . M . M aio & J. M . O lson (Eds.), W hy w e ev a lu a te :
G raziano, W ., Leone, C ., M usser, L ., & Lauten- F u n ctio n s o f a ttitu d es (pp. 9 7 -1 3 2 ). M ahw ah, N J:
schlager, G. (1 9 8 7 ). Self-m onitoring in children: A Erlbaum .
differential approach to social development. D e v e l Leary, M . R ., Tam bor, E. S., Terdal, S. K ., & Dow ns,
o p m e n t a l P sy c h o lo g y , 2 3 , 5 7 1 - 5 7 6 . D. L. (1995). Self-esteem as an interpersonal m oni
Gudykunst, W. B ., G ao, G ., N ishida, T ., Bond, M . tor: The sociom eter hypothesis. J o u r n a l o f P ers o n
H ., Kwok, L ., & W ang, G. (1989) A cross-cultural a lity a n d S o c ia l P sy c h o lo g y , 6 8 , 5 1 8 - 5 3 0 .
com parison of self-m onitoring. C o m m u n ic a tio n L ennox, R ., & W olfe, R . (1984). Revision of the Self-
R e se a rc h R e p o r ts , 6, 7 - 1 2 . M onitoring Scale. J o u r n a l o f P erson ality a n d S o cia l
Gudykunst, W. B ., Yang, S., & N ishida, T. (1987). P sy ch olog y , 4 6 , 1 3 4 9 -1 3 6 4 .
Cultural differences in self-consciousness and self Leone, C ., & H all, I. (2 0 0 3 ). Self-m onitoring, marital
m onitoring. C o m m u n ic a tio n R e se a rc h , 14, 7 - 3 6 . dissatisfaction, and relationship dissolution: Indi
H arris, M . J ., & R osenthal, R . (1986). Counselor and vidual differences in orientations to m arriage and
client personality as determ inants of counselor ex divorce. S e lf a n d Id en tity , 2 , 1 8 9 -2 0 2 .
pectancy effects. J o u r n a l o f P erso n a lity a n d S o c ia l Leone, C ., & H aw kins, L. B. (2 0 0 6 ). Self-m onitoring
P sy ch olog y , SO, 3 6 2 -3 6 9 . and close relationships. J o u r n a l o f P erson ality , 74,
H ofm ann, S. G. (2 0 0 6 ). The em otional consequences 7 3 9 -7 8 8 .
o f social pragmatism: T he psychophysiological co r Levine, S. P., 8c Feldm an, R . S. (1 9 9 7 ). Self-
relates of self-m onitoring. B io lo g ic a l P sy ch olog y , presentational goals, self-m onitoring, and nonver
73, 1 6 9 -1 7 4 . bal behavior. B a sic a n d A p p lied S o c ia l P sy ch olog y ,
H ogan, R . T. (1991). Personality and personality 19, 5 0 5 -5 1 8 .
measurement. In M . D. Dunnette & L. M . Hough Lippa, R . (1976). Expressive control and the leakage
(Eds.), H a n d b o o k o f in d u stria l a n d o r g a n iz a tio n a l of dispositional introversion-extraversion during
p s y c h o lo g y (2nd ed., Vol. 2 , pp. 8 7 3 -9 1 9 ). Palo role-played teaching. J o u r n a l o f P erson ality , 4 4 ,
A lto, CA: Consulting Psychologists. 5 4 1 -5 5 9 .
Holm es, J . G ., & Rem pel, J . K. (1989). Trust in close Lippa, R . (1978). Expressive control, expressive consis
relationships. In C . H endrick (Ed.), C lo s e r e la tio n tency, and the correspondence between expressive
sh ip s (Vol. 10, pp. 1 8 7 -2 2 0 ). London: Sage. behavior and personality. J o u r n a l o f P erson ality ,
Hoyle, R . H ., & Sow ards, B. A. (1993). Self-m onitoring 4 6 ,4 3 8 - 4 6 1 .
39. S elf-M o n ito rin g 591
Lippa, R ., 8c D onaldson, S. 1. (1990). Self-m onitoring ties: T h e p s y c h o lo g y o f self-m o n ito rin g . New York:
and idiographic measures of behavioral variability Freeman.
across interpersonal relationships. J o u r n a l o f P er Snyder, M ., Berscheid, E ., 8c G lick, P. (1985). Focusing
so n a lity , 5 8 , 4 6 5 - 4 7 9 . on the exterior and the interior: Two investigations
M arkus, H. R ., 8c K itayam a, S. (1991). Culture and of the initiation of personal relationships. J o u r n a l o f
the self: Im plications for cognition, em otion, and P erso n a lity a n d S o c ia l P sy ch olog y , 4 8 , 1 427-1439.
m otivation. P sy c h o lo g ic a l R e v iew , 9 8 , 2 2 4 - 2 5 3 . Snyder, M ., Berscheid, FI., 8c M atw ychuk, A. (1988).
M eehl, P. E. (1978). Theoretical risks and tabular as Orientations toward personnel selection: D if
terisks: Sir K arl, Sir Ronald, and the slow progress ferential reliance on appearance and personality.
of soft psychology. J o u r n a l o f C o n su ltin g a n d C lin i J o u r n a l o f P erson ality a n d S o c ia l P sy ch o lo g y , 5 4 ,
c a l P sy c h o lo g y , 4 6 , 8 0 6 - 8 3 4 . 9 7 2 -9 7 9 .
M ehra, A ., K ilduff, M ., 8c Brass, D. J . (2001). The so Snyder, M ., 8c Cantor, N . (1980). T hinking about our
cial netw orks of high and low self-m onitors: Im pli selves and others: Self-m onitoring and social know l
cations for w orkplace perform ance. A d m in istrativ e edge. J o u r n a l o f P erson ality a n d S o c ia l P sy ch olog y ,
S c ien c e Q u arterly , 4 6 , 1 2 1 -1 4 6 . 39, 2 2 2 - 2 3 4 .
M ill, J . (1984). High and low self-m onitoring individu Snyder, M ., 8c D eBono, K. G. (1985). Appeals to image
als: T heir decoding skills and em pathic expression. and claim s about quality: Understanding the psy
J o u r n a l o f P erson ality , 5 2 , 3 7 2 - 3 8 8 . chology o f advertising. J o u r n a l o f P erso n a lity a n d
M usser, L. M ., 8c Browne, B. (1991). Self-m onitoring S o c ia l P sy ch olog y , 49, 5 8 6 -5 9 7 .
in middle childhood: Personality and social co rre Snyder, M ., 8c Gangestad, S. (1982). Choosing social
lates. D e v e lo p m e n t a l P sy ch o lo g y , 27, 9 9 4 - 9 9 9 . situations: Two investigations of self-m onitoring
N elson, K. (1981). Individual differences in language processes. J o u r n a l o f P erson ality a n d S o c ia l P sy
development: Im plications for development and lan ch o lo g y , 4 3, 1 2 3 -1 3 5 .
guage. D e v e lo p m e n t a l P sy ch olog y , 17, 1 7 0 -1 8 7 . Snyder, M ., 8c Gangestad, S. (1986). On the nature of
N eto, F. (1993). Love styles and self-representations. self-m onitoring: M atters of assessm ent, m atters of
P erson ality a n d In d iv id u a l D iffe r e n c e s , 14, 7 9 5 - validity. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l
803. ogy, 51, 1 2 5 -1 3 9 .
N orris, S. L ., 8c Zw eigenhaft, R . L. (1999). Self Snyder, M ., Gangestad, S., 8c Sim pson, J . A. (1983).
m onitoring, trust, and com m itm ent in rom antic Choosing friends as activity partners: The role of
relationships. J o u r n a l o f S o c ia l P sy ch olog y , 139, self-m onitoring. J o u r n a l o f P erso n a lity a n d S o c ia l
2 1 5 -2 2 0 . P sy ch olog y , 4 5 , 1 0 6 1 -1 0 7 2 .
O yam ot, C . M ., Jr., Fuglestad, P. T ., 8c Snyder, M . (in Snyder, M ., 8c M onson, T. C . (1975). Persons, situa
press). Self-m onitoring and perceptions of balance tions, and the control of social behavior. J o u r n a l o f
of power in rom antic relationships: W ho influences P erso n a lity a n d S o c ia l P sy ch o lo g y , 3 2 , 6 3 7 - 6 4 4 .
whom, and with what effect? J o u r n a l o f S o c ia l a n d Snyder, M ., 8c Sim pson, J. A. (1984). Self-m onitoring
P erso n a l R e la tio n s h ip s. and dating relationships. J o u r n a l o f P erso n a lity a n d
Petty, R . E ., 8c Cacioppo, J. T. (1986). The elabora S o c ia l P sy ch olog y , 47, 1 2 8 1 -1 2 9 1 .
tion likelihood model of persuasion. In L. Berkow Snyder, M ., Sim pson, J . A ., 8c Gangestad, S. (1986).
itz (Ed.), A d v a n c es in e x p e r im e n ta l s o c ia l p s y c h o l Personality and sexual relations. J o u r n a l o f P ers o n
og y (Vol. 19, pp. 1 2 3 - 2 0 5 ). New York: Academic a lity a n d S o c ia l P sy ch olog y , 51, 1 8 1 -1 9 0 .
Press. Snyder, M ., 8c Sm ith, D. (1 9 8 4 , May). S elf-m o n ito r in g
Petty, R. E ., 8c Wegener, D. T. (1998). M atching versus a n d c o n c e p tio n s o f frie n d sh ip . Paper presented at
m ismatching attitude functions: Im plications for the annual meeting of the M idwestern Psychologi
scrutiny of persuasive messages. P erso n a lity a n d cal A ssociation, Chicago.
S o c ia l P sy c h o lo g y B u lletin , 2 4 , 2 2 7 - 2 4 0 . Snyder, M ., 8c Swann, W. B. (1976). W hen actions re
R a rick , D. L ., Soldow, G. F., 8c Geizer, R . S. (1976). flect attitudes: The politics of impression m anage
Self-m onitoring as a m ediator of conform ity. C e n ment. J o u r n a l o f P erso n a lity a n d S o c ia l P sy ch o l
tral S ta tes S p ee c h J o u r n a l, 27, 2 6 7 -2 7 1 . og y , 3 4 , 1 0 3 4 -1 0 4 2 .
Rubin, Z . (1980). C h ild ren 's frie n d sh ip s. Cambridge, Snyder, M ., 8c Tanke, E. D. (1976). Behavior and at
M A : Harvard University Press. titude: Some people are more consistent than others.
Sam pson, E. E. (1978). Personality and the location of J o u r n a l o f P erson ality , 4 4 , 5 0 1 -5 1 7 .
identity. J o u r n a l o f P erson ality , 4 6 , 5 5 2 - 5 6 8 . Spangenberg, E. R ., 8c Sprott, D. E. (2 0 0 6 ). Self
Snyder, M . (1974). Self-m onitoring of expressive be m onitoring and susceptibility to the influence of
havior. J o u r n a l o f P erson ality a n d S o c ia l P sy ch o l self-prophecy. J o u r n a l o f C o n s u m er R esea rc h , 3 2 ,
og y , 3 0 , 5 2 6 -5 3 7 . 5 5 0 -5 5 6 .
Snyder, M . (1976). Social perception and social cau W olfe, R ., L ennox, R ., 8c Cutler, B. L. (1986). G et
sation. In J . H. Harvey, W. J . Ickes, & R. F. Kidd ting along and getting ahead: Em pirical support
(Eds.), N ew d ire c tio n s in a ttr ib u tio n re se a r ch (Vol. for a theory of protective and acquisitive self
1). H illsdale, N J: Erlbaum. presentation. J o u r n a l o f P erso n a lity a n d S o c ia l P sy
Snyder, M . (1979). Self-m onitoring processes. In L. c h o lo g y , 4 4 , 1 0 6 9 -1 0 7 4 .
Berkowitz (Ed.), A d v a n ce s in e x p e r im e n ta l s o c ia l Z an n a, M . P., O lson, J. M ., 8c Fazio, R . H. (1980).
p s y c h o lo g y (Vol. 1 2 , pp. 8 5 -1 2 8 ). New York: A ca A ttitude-behavior consistency: An individual d if
demic Press. ference perspective. J o u r n a l o f P erson ality a n d S o
Snyder, M . (1 9 8 7 ). P u b lic a p p e a r a n c e s /p r iv a te re a li c ia l P sy ch o lo g y , 3 8 , 4 3 2 - 4 4 0 .
Author Index
592
A u th o r In d ex 593
Cervone, D., 5 Cla rk, L. A., 13, 3 6 , 37, 87, 130, 131, Co rzine, J . B „ 97, 100
Chai ken , S., I l l , 112 , 31 9, 321 1 3 4 , 1 3 6 , 13 8 , 140, 141, 148, 150 Cosmides, E., 33 , 49 6
Cham be rlain, K., 2 4 7 Cla rk, R. A., 3 8 2 , 3 8 3 , 3 8 4 , 411 Co sta, D. M . , 89
Cham be rli n, B., 3 0 9 Cla rk, R. I)., 54 Co sta, P. T „ Jr., 27, 3 0 , 31 , 3 6 , 3 8 , 4 0 ,
Cha mpi on, C., 199 Cla rk, T. G., 135, 136, 137 49, 1 00 , l i 4 , 117, 118, 13 0, 131, 132,
Ch an , I. S. Y„ 2 9 0 Clark- Car ter , D., 165 133, 134, 135, 1 3 8 , 139, 141, 149,
Ch an , J . , 2 9 0 Clarkin, J . F., 133 2 1 1 ,2 1 2 ,2 5 7 ,2 5 8 ,2 5 9 ,2 6 1 ,3 7 1
Ch an , W „ 4 6 8 , 4 6 9 , 4 7 5 , 4 7 6 Cleeland, C. S., 156 Costello, E „ 2 4 7
Chan g, c:., 2 5 0 , 321 Clifford, J. , 2 2 8 , 2 2 9 Cote, S., 376
Chan g, K. C., 4 8 5 , 4 8 8 Cloninger, B., 4 5 6 Cottier, E. B., 161
Chan g, E., 4 9 9 Cloninger, C. R „ 130, 4 5 6 , 4 5 9 , 4 6 2 Cottrell, C. A., 401
Chan g, S. H ., 4 4 6 (-lore, G. L., 39, 163 Co uc h m a n , C. E., 68
Chan g, W. C., 4 8 8 Co ates, D., 149 Couper, D. J . , 221
Chang-Schneider, C., 5 2 8 , 52 9 , 5 33 Cober, R., 3 76 C.'ourneya, K. S., 3 7 5 , 3 76
Chaplin, W. F., 9 Co ff m an , T. L., 3 5 9 Courtney, B. E., 13
C ha pm an , B. P., 137 Co gbu rn, H. E., 2 5 9 Co wa n , C. P., 69
Charles, S. T., 151 Cohen, A. R „ 3 1 8 , 319 Co wa n , P. A., 69
Ch arn g, H., 4 9 9 Cohen, J . , 135, 136 C o x , C. L „ 501
Ch art ra nd, T. I.., 121, 5 2 2 , 586 Cohen, R . J . , 3 9 1 C o x , E>. N „ 198
Chasiotis, A., 4 17 Cohen, S., 155, 1 5 6 , 2 1 9 , 4 2 0 Coy, K. C „ 4 9
C h atterjee, S., 321 Co hn, E. S., 29 3 Coyne, J . C:., 162, 164, 168 , 50 5
C ha vanon, \1. L., 34 Coie, j. D., 94 Cozzarelli, C „ 7 1 , 73
Cheavens, J . S., 2 4 7 Coke, J. S., 51 , 54 Craiger, |. P., 55
Cheek, ]. M „ 179, 18 3, 196 , 4 0 1 , 4 9 7 , Colb urn , T. A., 199 Cra in , T. 1.., 99
5 0 0 ,5 0 1 ,5 0 5 , 553,575 Colder, M . , 5 6 8 Cramer, D„ 4 4 7
Chen, E ., 180 Cole, P. M . , 53 Cramer, K. L., 1 1 7 , 5 1 6
Chen, G ., 3 7 5 , 376 Cole, S. W., 2 3 6 , 3 4 8 Cramer, K. M ., 2 2 8 , 4 9 9
Chen, H. C., 2 5 0 Coles, M . E., 183 Cramer, P., 4 8 2
Chen, S., 34 8 Collins, B., 2 7 7 Crane, R. J . , 2 1 3
Chen, V., 186 Collins, N. L., 6 8 , 69, 71 , 7 2 , 84 Craske, M „ 180
Chen, Y., 181, 51 4 , 5 23 Collins, P. F„ 3 4 , 132 Cratylus, 3 4 4
Chen, Y. R., 5 2 3 Colquitt, J. A., 376 Craven, R. G., 5 2 8 , 5 3 3 , 5 3 4 , 5 4 0
Cheney, S ., 5 3 1 , 5 5 6 Colvin, C. R., 6, 31, 38 , 1 4 8 , 2 4 4 , 247, Crawford, E. C „ 4 6 2
Cheng, C. M . , 58 6 260 Crawford, M . T „ 3 2 4
Cheng, H., 2 8 2 Comer, R., 291 Crawford, T. N ., 64
Cheng, R. W., 521 Co mpto n, W. M . , 161 Creed, A. T „ 4 9 9
Cherlin, A., 2 7 7 Conger, R. D„ 2 6 2 , 3 38 Creed, E., 165
Chernyshenko, O. S., 3 70 Connellv, S., 113 Crelia, R. A., 3 2 4
Chesney, M . A., 2 1 1 , 2 1 3 , 2 1 4 , 2 17 Conner,'iV(„ 3 7 5 , 3 7 6 , 4 4 9 C-ritchely, C., 3 19
Chester, N. I.., 4 1 0 , 421 Conne r, T. S., 2 0 , 163 Critel 1i, J . W., 5 5 4
Cheung, P., 3 8 8 Connolly, J. J . , 261 Crocker, J . , 14, 116, 5 3 5 , 5 3 8 , 5 4 0 , 5 5 6 ,
Cheung, P. C., 3 93 Co nn o r-S m it h , J. K., 136, 142 566, 567 , 571
Chew, C. H ., 481 C.’onrad, M . , 171 Cron bac h, E. J., 4
( d i c k e r i n g , S. A., 199 Conroy, D. E., 3 8 5 , 3 9 0 , 3 9 1 , 3 9 2 Cr ook s, V., 2 7 7
Chida, Y . , 2 2 1 Constable, R. T., 35 , 53 Cro pan za no, R. S., 155, 2 4 4
Child, I. I.., 121 Co nstans, J. E, 182 Cross, |. A., 55
Chiles, ]. A., 4 4 2 Co nst an tia n, C. A., 4 1 3 , 4 1 5 , 421 Cross, S. E „ 83 , 11 0, 115, 116, 117,
Chillag, K . , 4 4 9 Co nstantine, M . G ., 51 9 12 0 , 5 1 2 , 5 1 4 , 5 1 6 , 517, 5 1 8 , 519,
Chiodo, 1.. M . , 551 Constantinople, A., 113 5 2 0 ,522
Ch irib oga , D. A., 2 2 0 Conte, J . \1., 3^1 C'rotts, J . C „ 9 7
C hi ru m bol o, A., 3 4 9 Cont rad a, R. J . , 2 1 3 , 2 2 0 Crouppen, G. A., 196
Chiu, C ., 3 4 4 , 516 Conway, C., 3 3 7 Crowell, E A., 6 6 , 68
Ch rista!, R. E., 2 9 Conway, G. S., 121 Crowne/D. P., 4 1 3 , 4 4 1 , 4 4 4 , 4 4 5 ,
Christensen, A. J., 2 1 9 Conway, E. G., Il l , 36 2 4 8 3 ,584
Christensen, H ., 136 C o ok , D. R „ 194, 195 Cruet, I)., 5 0 5
Christensen, L., 99 , 100 C o ok , K. E., 5 36 Csikszentmihalyi, ,VE, 2 1 , 4 1 2 , 41 6,
Christensen, P. N ., 110 C o ok , () . , 2 3 0 421, 50 4 , 564
Christensen, T. C., 53 8 C o ok , T. D „ 50 Cubela Adoric, V., 2 9 2
Christiansen, N. D., 2 6 5 C o ok , W. W., 2 13 Cui xi a, 1.., 4 4 8
Christie, R., 93, 9 5 , 96 , 9 8 , 99, 102, Cooley, C. H „ 4 9 8 , 52 7, 5 3 4 Cuk rowi cz, K. C., 162
103, 1 0 4 , 2 7 8 , 3 0 1 , 3 1 1 Co on , H. M . , 115, 5 1 4 , 5 3 4 Gumming, S. P., 4 4 6
Christoffcrsen, D., 96 Cooper, H., 136, 142, 150, 21 3, 37 3, 5 4 0 Cum mings, E, 45 8
Chuah, S. C „ 371 Cooper, M . E., 14, 69, 71 , 74 , 75, 5 5 6 , Cunn ick, J . E., 3 3 7
Chun, W, Y„ 3 4 4 , 3 4 8 567 C un ni ng ham , M . R., 9 8 , 155
Church, M . A., 3 8 2 , 3 9 0 , 4 8 6 , 5 6 4 Cooper, M . M . , 3 3 7 Cur ran, P. J . , 2 0
Chusmir, L. H., 421 Coopersmith, S., 5 32 Curtis, R . C., 185
Chwalisz, K., 2 4 4 Copeland, J . , 168 Cut hbert, B. N., 53
Chymis, A., 155 Corbit, J . D., 5 7 Cutler, B. L., 5 8 5
Cialdini. R. B „ 55, 38 6 Cordova, M ., 571 Cutler, L., 196
Ciani, A. S. C., 2 6 9 Co ren, S., 358 Cutler, S. E., 2 4 4
C ia ro cc o , N. J . , 3 7 2 , 4 6 6 Corr, P. |., 33 , 3 6 , 4 0 , 2 8 2 Cu trona, C. E., 231
Ciar ro cc hi , J . W., 201 C:orral, S., 98
Civelek, A. C., 21 4 Correia, E, 2 8 9 , 2 9 2 , 2 9 3 D a c e y ,J . S., 4 3 6
Clancy, S. M ., 11 2 , 117 Correll, J., 556 Dadds, M . R., 4 4 6
Clark, A. E ., 150 Correv, B. L . , 2 9 1 D ’Agostino, P. R., 2 5 9
Clark, C. 1.., 68 Co rte, V., 95 D ahl, D. W „ 187
Clark, D. A., 4 1 9 , 5 0 4 Corteen, R. S., 361 Dah lst ro m, W. G ., 213
Clark, D. M „ 180, 181, 182 Co rte s, B. P., 5 3 0 D ah n ke , G ., 27 8
Cla rk, E., 27 8 Co rti n a, J . M . , 376 Dalai, A. K., 2 9 1 , 2 9 4
Clark, J. K „ 31 9 Corveleyn, |., 197 Da lbe rt, C., 2 8 8 , 2 8 9 , 2 9 0 , 2 9 1 , 2 9 2 ,
Clark, K. E ., 89 Coryell, W.‘ 162 2 9 3 , 2 9 4 , 2 95
596 A u th o r In d ex
Page numbers followed by f indicate figure, n indicate note, and t indicate table
612
Sub ject Index 613
Love, happiness and, 1 5 0 - 1 5 1 Miller Behavioral Style Scale (MB SS ), Narcissistic Personality Inventory (NPI)
Lying, Machiavellianism and, 99 483-484 contemporary models of, 5 5 0 - 5 5 5
M illon Clinical Mu ltiax ial Inventory, overview, 5 3 1 , 5 4 9 - 5 5 0 , 5 5 7
M ac h IV scale 2 4 8 ,549 self-esteem and, 5 56
Kiddie M a c h scale and, 102 Mimic ry, self-monitoring and, 58 6 Need for achievement, social behavior
overview, 95, 97, 104 Mindfulness, self-compassion and, a n d , 3 9 0 - 3 9 6 , 393/
Machiavellian Behavior Scale (Ma ch -B) , 5 6 2 ,5 7 1 Need for approval, 575
96-97 Mindfulness-based emotion-regulation Need for closure, 3 4 3 - 3 5 1
Machiavellian intelligence, 101 intervention, 170 group processes and, 3 4 8 - 3 5 0
Mac hiavellianism, 9 3 - 1 0 4 Mindfulness-based stress reduction individual differences related to, 3 4 4
ch aracter associated with, 9 3 - 9 7 ( M B S R ) , 571 integrative complexity and, 3 5 4 - 3 5 5
conceptual systems theory and, 3 56 Mi n n e so ta Multiphasic Inventory interpersonal processes and, 3 4 7 -
malevolence and, 9 7 - 1 0 0 (MMPI) 348
overview, 1 0 3 - 1 0 4 narcissism and, 5 4 9 , 5 5 3 - 5 5 4 intrapersonal processes and, 3 4 5 -
Sel f- Monitoring Scale and, 5 75 social desirability and, 4 4 2 347
M A F F measure, 4 1 7 - 4 1 8 Mo ni tor in g, psychological defenses and, openness and, 2 5 8 , 2 6 9
M a in -e ff ec t research, 7 - 8 483-484 overview, 3 4 3 - 3 4 4 , 3 5 0 - 3 5 1
Malevolence, Machiavellianism and, Mo n tg om e ry -A sb e rg Depression Rating self-esteem and, 5 6 6 - 5 6 7
97-100 Scale ( M A D R S ) , 162 situational determinants, 3 4 4
M an if es t Anxiety Scale, 4 8 3 M o o d , loneliness and, 2 2 8 urgency and permanence tendencies
M an ipu la tio n, 9 3 - 1 0 4 , 9 7 - 9 8 M o o d disorders of, 3 4 5
M a r l o w e - C r o w n e Social Desirability extraversion and, 3 5 , 4 0 Need for Closure Scale ( N F C S ) , 3 4 4
Scale, 4 4 4 , 4 4 5 , 5 8 4 introversion and, 29 Need for cognition, 3 1 8 - 3 2 6
Marriage neuroticism and, 136 applied areas, 3 2 4 - 3 2 5
conscientiousness and, 3 7 2 self-report measures and, 138 attitudes and persuasion and,
happiness and, 153 Morality 320-322
hostility and anger and, 2 2 0 - 2 2 1 Machiavellianism and, 9 8 - 9 9 conceptualization of, 3 1 8 - 3 1 9
loneliness and, 2 3 0 shame and guilt and, 1 9 4 - 1 9 5 , health and, 3 2 4 - 3 2 5
openness to experience and, 2 6 1 - 2 6 3 2 0 0 -2 0 1 ,2 0 4 integrative complexity and, 3 5 4 - 3 5 5
self-compassion and, 5 6 5 - 5 6 6 social desirability and, 4 4 4 interpersonal relationships a nd, 3 2 4
self-esteem and, 5 4 0 Motivation law and, 3 2 4 - 3 2 5
self-monitoring and, 581 affect intensity and, 2 4 9 measurement of, 3 1 8 - 3 1 9
sensation seeking and, 45 8 agreeableness and, 4 6 - 4 7 , 5 0 , 5 3 - 5 8 , openness and, 2 5 9 , 2 6 9
See a lso Relationships 57/ overview, 3 2 5 - 3 2 6
M a r x i s m , authoritarianism and, attac hment theory and, 63 social cognition and decision mak ing
299-300 authoritarianism and, 31 3 a n d ,322-324
Masc uline identity complexity and, 3 6 0 - 3 6 4 theories of judgment and, 3 1 9 - 3 2 0
compatibility principle and, 111 conscientiousness and, 3 7 4 - 3 7 5 Need for Cognition Scale, 3 1 9 - 3 2 0
overview, 110 embarrassability and, 186 Need to Belong Scale
personal traits and attributes and, interpersonal dependency and, 8 3 - 8 4 , cooperation and, 4 0 3
112-115 84/ fear of rejection and, 4 0 6
See a lso Gender identity justice and, 2 8 9 overview, 4 0 1 , 4 0 7
Mastery response, 3 8 7 - 3 8 8 loneliness and, 2 2 9 perceived rejection and, 4 0 5 - 4 0 6
Mastery-approach goals, 3 8 8 - 3 8 9 , 389/ Machiavellianism and, 9 3 - 9 4 social co nnection and, 4 0 2
Mastery-avoidance goals, 3 8 8 - 3 8 9 , 389/ neuroticism and, 136 Negative affectivity
Ma te rialis m, belonging motivation optimism and, 3 3 0 neuroticism and, 1 3 1 - 1 3 2 , 133
an d ,404 self-compassion and, 5 6 4 - 5 6 5 , psychological defenses and, 4 8 6 -
Maudsley Personality Questionnaire 570-571 487
(M P Q ) , 2 9 , 31/, 32/ self-construals and, 5 1 9 - 5 2 1 self-esteem and, 533
M a y e r - S a l o v e y -C a r u s o Emotional self-esteem and, 5 3 6 See a lso Affect
Intelligence Test ( M S C E I T ) , 2 4 9 self-monitoring and, 58 3 Negative cognitive triad, 164
Mea ni ng, sense of, 153, 5 6 3 - 5 6 4 shame and guilt and, 193 Negative emotionality, neuroticism
Me asurement modality, 14 social anxiety and shyness and, 183 a n d , 130
Measurement-of-mediation strategy, See a lso in dividu al m otiv ation al Negative temperament, neuroticism
18-19 d isp osition s a n d , 130
Measurements. See Assessment Motivation and opportunity as Negative thinking, psychological
measures determinants (M O D E ) model, 541 defenses and, 4 8 6 - 4 8 7
Me diational hypotheses, 1 8 - 1 9 , M S R measure, 4 1 7 - 4 1 8 N E O Five-Factor Inventory ( N E O FFI)
201-202 Multidimensional Personality agreeableness and, 49
Me ditation practices, 571 Questionnaire ( M P Q ) , 30 extraversion and, 3 0 , 3 8 , 39
Me mor ies, false, 32 3 impulsivity and, 132 neuroticism and, 135
M e m or y system, 6 5 - 6 6 , 5 1 7 - 5 1 8 neuroticism and, 13 0 , 130/, 131, 132 overview, 31/, 32/
M en tal health openness to experience and, 26 1 psychopathology and, 4 0
at tachment styles and, 7 3 - 7 6 overview, 31 /, 32/ self-compassion and, 5 69
Machiavellianism and, 9 5 - 9 6 Multifactorial theory of gender, 110 N E O Personality Inventory— Revised
narcissism and, 5 5 4 - 5 5 5 Multiprocess model, agreeableness and, ( N E O PI-R)
self-consciousness and, 5 0 4 - 5 0 6 53-54 cultural factors and, 2 6 7
See a lso Psychopathology dependency and, 1 3 3 - 1 3 4
M en tal processes, 2 3 1 - 2 3 3 . See also Narcissism, 5 4 7 - 5 5 8 emotional instability and, 1 3 2 - 1 3 3
Cognition assessment of, 5 5 7 - 5 5 8 hostility and anger and, 2 1 2 - 2 1 3
Men tal representations. See co ntemporary models of, 5 5 0 - 5 5 3 impulsivity and, 132
Representations Machiavellianism and, 1 0 0 - 1 0 1 neuroticism and, 130/, 13 2 , 1 3 2 - 1 3 3 ,
Me tac og nit ion , 3 2 1 - 3 2 2 , 53 6 mental health and, 5 5 4 - 5 5 5 1 3 3 - 1 3 4 , 1 3 4 - 1 3 5 , 138
M e th od s o f study overview, 5 2 7 - 5 2 8 , 547, 5 4 8 - 5 5 0 , 558 openness and, 2 5 8 , 2 6 7 , 269m
data analysis and, 2 1 - 2 2 self-compassion and, 5 6 6 - 5 6 7 , 5 6 8 overview, 130
general versus specific measurement, self-esteem and, 5 2 7 - 5 2 8 , 531 Nervous system, 2 3 7 , 2 4 5 - 2 4 6
13-14 self-regulatory models, 5 5 3 - 5 5 8 Netherlands Twin Register Study, 2 2 9
methodological strategies, 1 5- 21 social desirability and, 4 4 4 N euroendocrine functioning, loneliness
multiple measures and, 1 4 - 1 5 theoretical approaches, 5 4 7 - 5 4 8 a n d ,23 5 -2 3 6
overview, 12, 2 2 Narcissistic personality- disorder, Neurological functioning, see Brain
trait versus state measurement, 1 2 - 1 3 549-560 functioning/structure
620 Sub ject In d ex