Professional Documents
Culture Documents
Which Are The Basic Meaning Dimensions o
Which Are The Basic Meaning Dimensions o
Which Are The Basic Meaning Dimensions o
a r t i c l e i n f o a b s t r a c t
Article history: We investigated the basic meaning dimensions of overt interpersonal behavior in an observational multi-
Received 16 May 2011 rater multi-situation study. The interpersonal behavior of 99 participants in 17 different interpersonal
Received in revised form 30 July 2011 role-plays was videotaped and judged by three independent observers using 35 adjective-pairs. Principal
Accepted 3 August 2011
Components Analysis yielded three factors with Eigenvalues above chance level. The first two factors
Available online 10 September 2011
were easily identifiable as Agency and Communion, which represent the fundamental dimensions in inter-
personal theory (Wiggins, 1991). Unexpectedly, we also found a third, weaker factor pertaining to Emo-
Keywords:
tional Stability. Factor coefficients and factor scores of all three factors showed acceptable to good
Big Five
Agency
stability across observers and role-plays. Our study is the first to demonstrate the appropriateness of
Communion the Agency/Communion framework with observational data.
Emotional Stability Ó 2011 Elsevier Ltd. All rights reserved.
Neuroticism
Behavior
Observation
0191-8869/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.paid.2011.08.003
D. Leising, W. Bleidorn / Personality and Individual Differences 51 (2011) 986–990 987
people would use spontaneously when they are asked to judge of other measures. In the present paper, we will only refer to
other people’s overt, momentary interpersonal behavior. The goal assessments of the participants’ overt interpersonal behavior. Each
of the present study was to answer this question. participant took part in 17 dyadic role-plays (‘‘standard interaction
Borkenau and Ostendorf (1998) demonstrated that the Five Fac- tasks’’), to assess his or her habitual ways of dealing with challeng-
tor Model of personality may not only be used to conceptualize sta- ing interpersonal situations. At the beginning of each role-play, the
ble individual differences between people, but also changes within participant received a written instruction containing an outline of a
individual persons (= states). For demonstrating this, they used ‘‘p- hypothetical situation and a request to demonstrate some critical
factoring’’, which means that the correlations of items across sev- target behavior. For example, in one role-play the participants
eral measurement occasions (= ‘‘cases’’) were separately factored were asked to imagine that they had attended a very good party
for each person. However, Borkenau and Ostendorf used self-report the evening before, and now incidentally met the host of that party
data only: Their participants reported their own states in terms of again. Their task was to thank and commend the host for the great
adjectives for 90 consecutive days. The idea that states and traits party.
may be described in terms of the same items has recently gained We used the following interaction tasks: (1) initiate a conversa-
some new popularity within the ‘‘personality states’’-approach tion, (2) apologize to someone, (3) thank someone, (4) assert a
(Bleidorn, 2009; Fleeson, 2001). For example, Fleeson (2001) con- claim, (5) confront someone, (6) present yourself positively, (7)
ceptualized personality in terms of ‘‘density distributions of be self-critical, (8) show empathy, (9) reprimand someone, (10)
states’’: As a person’s states accumulate over time, they form a dis- convince someone of doing something, (11) give instructions,
tribution. The mean of that distribution may be used as an estimate (12) pay a compliment, (13) ask for emotional support, (14)
of the person’s trait level, whereas measures of dispersion (e.g., the encourage someone, (15) bargain with someone, (16) express your
standard deviation) may be used to assess how much the person’s affection for someone, and (17) ask for instrumental support. The
states vary across time and situations. However, like Borkenau and creation and selection of the role-plays was partly based on the
Ostendorf (1998), Fleeson also used self-report measures (i.e., theoretical models we referred to above, and on experiences with
experience sampling) only. interpersonal role-plays that are reported in previous studies
We emphasize the issue of perspective (self- versus other-rat- (e.g., Leising, Sporberg, & Rehbein, 2006; Leising & Müller-Plath,
ings), because it is likely to make a difference in terms of the 2009). More detailed descriptions of the individual role-plays
dimensionality of person descriptions. Judgments of the same tar- may be obtained from the first author.
get person from different perspectives (e.g., from within versus In the role-plays, the participants interacted with one of three
from outside) may differ systematically with regard to the kind confederates (two female, one male), who were advanced psychol-
and amount of information on which they are based (cf. Beer & Wat- ogy students. Participants were randomly paired with the confed-
son, 2008; Vazire, 2010). Most important, because more internal erates, as the orders in which participants and confederates signed
phenomena like thoughts, intentions and expectations are difficult up for time slots in the study were independent of each other. The
or impossible to observe, they are unlikely to form separate factors only restriction was that the male confederate would be the inter-
when a person is judged from outside, by strangers. Thus, if we are action partner for about half of the sample. As a consequence of the
to judge only the overt (i.e., visible, audible), momentary behaviors random pairing, participant sex and confederate sex were uncorre-
of people who are interacting with others, and we have no prior lated. It was ensured that confederates and participants were unac-
information about those people, and no opportunity to ask them quainted with each other. For each participant, the order in which
about their inner experiences, it is likely that the number of mean- the individual role-plays were presented was randomized. A par-
ing dimensions we use in our judgments will be rather limited. The ticipant interacted with the same confederate in all role-plays,
intriguing question that remains is: How many, and which, dimen- and the confederates gave scripted answers, in order to make the
sions will it be? To answer this question, we videotaped target per- situations as similar as possible for all participants.
sons in several interpersonal role-plays, let unacquainted The participants’ interpersonal behavior in the role-plays was
observers judge the targets’ behavior by means of natural language videotaped and then rated from video by three female observers,
terms, and then applied exploratory factor analysis to the obser- all of whom were grad students in psychology. For their ratings,
ver-ratings. the observers used 35 pairs of adjectives, which always appeared
in the same order (reported in Table 1). The observers watched
each video-clip once, judged the respective participant’s behavior
2. Method
by means of the first 18 items, then watched the video-clip a sec-
ond time, and then judged the participant’s behavior by means of
2.1. Sample
the remaining 17 items. By using this procedure, we hoped to max-
imize the amount of actual behavioral information that would be
The study was carried out at a mid-size university in the East of
considered by the observers as a base for their judgments. It was
Germany. We used-an ad hoc sample of ninety-nine research vol-
ensured that the observers had never met the participants they
unteers (66 female) with a mean age of 23.0 years (SD = 3.8,
rated before. The observers worked independently, and did not
range = 18–36). About half of the participants were university stu-
communicate with each other about their judgments. Furthermore,
dents, whereas the remaining participants came from the commu-
the observers were not aware of the research questions of the
nity. Participation was denied to persons who were undergoing
study, and did not know how the items of the measure (see below)
psychiatric treatment, in order to protect them from the stress that
related to the different dimensional models.
may be caused by participating in the somewhat challenging inter-
personal role-plays (see below). All participants received a finan-
2.3. Measure
cial compensation of 15 Euro. The study was advertised online
and by public notice on campus.
The Minimum Redundancy Scales (Ostendorf, 1990) consist of
adjective pairs that were carefully selected to assess the Big Five per-
2.2. Procedure sonality factors with as little semantic overlap as possible. Its five-
factor structure has repeatedly been demonstrated with large sets
The data were collected in the course of a larger research pro- of self-ratings of personality. In the full version of the measure, each
ject, in which the same participants were also assessed by means factor is assessed by nine pairs of adjectives (e.g., talkative–silent for
988 D. Leising, W. Bleidorn / Personality and Individual Differences 51 (2011) 986–990
Table 1
Principal components analysis of ratings of overt interpersonal behavior (averaged across observers and situations).
Note: Original items from the Minimum Redundancy Scales (MRS). The ‘‘Item’’ column contains the order in which the items were presented.
In the table, items were re-ordered to facilitate factor interpretation. E = Extraversion, ES = Emotional Stability, C = Conscientiousness, A = Agreeableness, O = Openness.
*
Additional items (not included in MRS). Boldface: Highest loading per item.
Extraversion (E); softhearted–ruthless for Agreeableness (A); secure– 3.1. Inter-rater agreement
insecure for Emotional Stability (ES); creative–uncreative for Open-
ness (O); thorough–careless for Conscientiousness (C)). In the present Separately for each of the 17 situations and each of the 35 items,
study, the observers used an abridged version, due to limited re- we computed the agreement (a) between the three observers. The
sources. This abridged version (MRS-30) comprises 30 items only, resulting 595 coefficients (17 situations 35 items) had a mean of
but has been shown to still provide good discriminant validity for .61 and a standard deviation of .13 (range: .13 to .87). Eighty per-
the factors, and good internal consistencies for the scales (Schallber- cent of these coefficients were .50 or higher. When aggregating
ger & Venetz, 1999). One Openness item (artistic–unartistic) was across the 17 situations before computing inter-rater agreement,
omitted because it seemed irrelevant for the assessment of interper- the average a was .80 with a standard deviation of .07 (range:
sonal behavior. Because the Dominance dimension of the Interper- .58–.90). Eighty percent of these coefficients were .76 or higher.
sonal Circumplex model is not sufficiently represented within the Thus, at least at a higher level of aggregation, the agreement be-
Minimum Redundancy Scales, we added another six items from tween the observers was good.
Ostendorf’s (1990) impressive set of adjective pairs to assess this do-
main as well (i.e., dominant–submissive, timid–assertive, determined–
slack, proud–unassuming, independent–dependent, self-sufficient-con- 3.2. Factor structure of aggregated observer-ratings
forming). For each item, the observers indicated which of the two
adjectives described the interpersonal behavior of a target person In order to determine the factor structure of the observer-rat-
better, by means of a six-point bipolar rating scale. ings, we conducted Principal Components Analyses (PCA) with
Varimax rotation. The number of factors to be retained was deter-
mined by parallel analysis (Horn, 1965). Fig. 1 displays the Scree
3. Results plot for a PCA with observer-ratings that were averaged across
observers and situations. The figure also contains the 95 percent
In the following, we will report on the agreement between observ- confidence limit for Eigenvalues that would be expected when
ers, as well as on the stability of values (factor coefficients, factor using a random data matrix with the same numbers of cases and
scores) across situations, and across observers. These different kinds variables (= parallel analysis, 1000 trials). By comparing the empir-
of associations are traditionally expressed by means of different indices ical Eigenvalues with the confidence limit, it is possible to identify
(i.e., Intraclass Correlations (ICC) vs. Cronbach’s a). However, because the factors whose Eigenvalues exceed chance level. The parallel
a and ICC (2, k), which is the appropriate measure of inter-rater analysis unequivocally suggested retaining three factors. Thus, as
agreement in the present study, yield identical results when applied expected, the factor structure of the item set was simpler when
to the same data, we will only use a, for the sake of simplicity. the items were used for judging people’s overt interpersonal
D. Leising, W. Bleidorn / Personality and Individual Differences 51 (2011) 986–990 989
study suggests that Emotional Stability may be more interpersonal Beer, A., & Watson, D. (2008). Asymmetry in judgments of personality: Others are
less differentiated than the self. Journal of Personality, 76(3), 535–559. doi:
than is often assumed: The observers were obviously able to judge
10.1111/j.1467-6494.2008.00495.x.
some of the participants’ experiences relating to this domain (e.g., Bleidorn, W. (2009). Linking personality states, current social roles and major life
‘‘robust–vulnerable’’, ‘‘relaxed–tense’’) from outside, with good goals. European Journal of Personality, 23(6), 509–530. doi: 10.1002/per.731.
agreement. In the context of our interpersonal role-plays, we think Borkenau, P., & Ostendorf, F. (1998). The Big Five as states: How useful is the Five-
Factor Model to describe intra-individual variations over time? Journal of
that the Emotional Stability factor we found may reflect (low) so- Research in Personality, 32(2), 202–221. doi: 10.1006/jrpe.1997.2206.
cial anxiety. Fleeson, W. (2001). Toward a structure- and process-integrated view of personality:
We obtained three factors with a measure whose five-factor Traits as density distributions of states. Journal of Personality and Social
Psychology, 80(6), 1011–1027. doi: 10.1037//0022-3514.80.6.1011.
structure for self-assessments of personality is well-established Gurtman, M. B., & Balakrishnan, J. D. (1998). Circular measurement redux: The
(cf. Ostendorf, 1990). The most obvious reason for this lies in the analysis and interpretation of interpersonal circle profiles. Clinical Psychology:
lower amount of information to which our observers had access Science and Practice, 5, 344–360. doi: 10.1111/j.1468-2850.1998.tb00154.x.
Hogan, R. (1983). A socioanalytic theory of personality. In M. M. Page (Ed.), 1982
(cf. Beer & Watson, 2008; Vazire, 2010). First, they had never met Nebraska symposium on motivation (pp. 55–89). Lincoln: University of Nebraska
the targets before, and only saw them interacting with another Press.
person for about 20 min altogether. Given this limited amount of Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis.
Psychometrika, 30(2), 179–185. doi: 10.1007/BF02289447.
behavioral information, one would not expect a highly differenti- John, O. P., Naumann, L. P., & Soto, C. J. (2008). Paradigm shift to the integrative Big
ated pattern of correlations between items. Second, the observers Five trait taxonomy: History, measurement, and conceptual issues. In O. P. John,
were unable to communicate with the targets to obtain more de- R. W. Robins, & L. A. Pervin (Eds.), Handbook of personality: Theory and research
(3rd ed., pp. 114–158). New York, NY: Guilford Press.
tailed information regarding (e.g.) their thoughts, values, and aspi-
Kiesler, D. J. (1983). The 1982 Interpersonal Circle: A taxonomy for
rations. The fact that we did not find separate Conscientiousness complementarity in human transactions. Psychological Review, 90(3), 185–214.
and Openness factors is likely to reflect the lack of this kind of doi: 10.1037/0033-295X.90.3.185.
information. It may be concluded that these two are the least inter- Leary, T. (1957). Interpersonal diagnosis of personality. New York: Ronald Press.
Leising, D., & Müller-Plath, G. (2009). Person-situation integration in research on
personal of the Big Five factors. personality problems. Journal of Research in Personality, 43(2), 218–227. doi:
In explaining the simpler factor structure that we found, one 10.1016/j.jrp.2009.01.017.
may also argue that motivational reasons may have played a role: Leising, D., Sporberg, D., & Rehbein, D. (2006). Characteristic interpersonal behavior
in dependent and avoidant personality disorder can be observed within very
The observers watched each of 99 targets in each of 17 situations short interaction sequences. Journal of Personality Disorders, 20(4), 319–330.
(twice), and were asked to rate each target by means of 35 items, 10.1521/pedi.2006.20.4.319.
constituting a total amount of (99 17 35 =) 58.905 ratings per McCrae, R. R., & Costa, P. T. (1989). The structure of interpersonal traits: Wiggins’
circumplex and the five-factor model. Journal of Personality and Social
observer. Even though we had great trust in the conscientiousness Psychology, 56(4), 586–595. doi:10.1037/0022-3514.56.4.586.
of our observers, it would be conceivable that their ratings became McCrae, R. R., & John, O. P. (1992). An introduction to the five-factor model and its
somewhat stereotypical (i.e., less differentiated), because of the applications. Journal of Personality, 60(2), 175–215. doi:10.1111/j.1467-
6494.1992.tb00970.x.
sheer amount of ratings they were asked to deliver. However, if a Ostendorf, F. (1990). Sprache und Persönlichkeitsstruktur. Zur Validität des Fünf-
lack of motivation was the reason for the simpler factor structure Faktoren-Modells der Persönlichkeit [Language and personality structure. On the
of the observer-ratings, it would be hard to explain why the factors validity of the Five Factor Model of personality]. Regensburg, Germany: Verlag S.
Roderer.
that did emerge were so well replicable across observers. More-
Schallberger, U., & Venetz, M. (1999). Kurzversionen des MRS-Inventars von Ostendorf
over, the first two factors that we found were exactly the ones that (1990) zur Erfassung der fünf ‘‘großen’’ Persönlichkeitsfakoren [Brief versions of
Wiggins (1991) suggested, even though the observers were una- Ostendorf’s (1990) MRS inventory for assessing the five, big personality factors].
ware of Wiggins’s theory. Therefore, we think it is safe to say that Switzerland: University of Zürich.
Vazire, S. (2010). Who knows what about a person? The self-other knowledge
the simpler factor structure of the observer-ratings does not asymmetry (SOKA) model. Journal of Personality and Social Psychology, 98(2),
merely reflect an effect of motivational difficulties, but the actual 281–300. doi:10.1037/a0017908.
meaning dimensions of observable interpersonal behavior. Watson, D., & Clark, L. A. (1984). Negative affectivity: The disposition to experience
aversive emotional states. Psychological Bulletin, 96(3), 465–490. doi:10.1037/
0033-2909.96.3.465.
Wiggins, J. S. (1979). A psychological taxonomy of trait-descriptive terms: The
interpersonal domain. Journal of Personality and Social Psychology, 37(3),
References 395–412. doi:10.1037/0022-3514.37.3.395.
Wiggins, J. S. (1991). Agency and communion as conceptual coordinates for the
Abele-Brehm, A., & Wojciske, B. (2007). Agency and Communion from the understanding and measurement of interpersonal behavior. In W. Grove & D.
perspective of self versus others. Journal of Personality and Social Psychology, Cicchetti (Eds.), Thinking clearly about psychology: Essays in honor of Paul E. Meehl
93(5), 751–763. doi: 10.1037/0022-3514.93.5.751. (Vol. 2., pp. 89–113). Minneapolis, MN: University of Minnesota Press.
Bakan, D. (1966). The duality of human existence. Chicago, IL: Rand McNally &
Company.