Ideals in Intimate Relationships

Journal of Personality and Social Psychology Copyright 1999 by the American Psychological Association, Inc.
1999, Vol. 76, No. 1, 72-89 0022-3514/99/S3.00
Ideals in Intimate Relationships
Garth J. O. Fletcher Jeffry A. Simpson

University of Canterbury Texas A&M University
Geoff Thomas and Louise Giles

University of Canterbury
This research examined lay relationship and partner ideals in romantic relationships from both a
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
social-cognitive and an evolutionary perspective. Studies 1 and 2 revealed that the qualities of an ideal
This document is copyrighted by the American Psychological Association or one of its allied publishers.
partner were represented by 3 factors (partner warmth-trustworthiness, vitality-attractiveness, and

status-resources), whereas the qualities of an ideal relationship were represented by 2 factors (relation-
ship intimacy-loyalty and passion). A confirmatory factor analysis in Study 3 replicated these factor
structures but found considerable overlap across the partner and relationship dimensions. Studies 4 and 5
produced convergent and discriminant validity evidence for all 5 factors. Study 6 indicated that the higher
the consistency between the ideals and related assessments of the current partner and relationship, the
more positively the current relationship was evaluated.
How do people know whether they are in a good or a bad ship ideals, guided by past research and theorizing drawn from
intimate relationship? On what basis do people decide whether to both social-cognitive and evolutionary approaches.
become more involved, live together, get married, or look for
another mate? These are some of the most complex and difficult Structure and Content of Partner and Relationship Ideals
questions that relationship researchers grapple with, and a variety
of theories have been developed to address them (see Fletcher & A Social-Cognitive Approach
Fitness, 1996). One answer to such questions is that judgments or
decisions concerning a particular relationship should be based, at From a social-cognitive standpoint, partner and relationship
least in part, on the consistency between general relationship ideals will include chronically accessible knowledge structures
standards or expectations, on the one hand, and perceptions of the that are likely to predate— and be causally related to—judgments
current relationship, on the other (e.g., see Fletcher & Thomas, and decisions made in ongoing relationships.
1996; Rusbult, Onizuka, & Lipkus, 1993; Sternberg & Barnes, There are several reasons why ideals should play a prominent
1985). role in ongoing relationships. First, intimate relationships are very
This idea is hardly new, being originally formulated by Thibaut important in many people's lives. Hence, it is hardly surprising
and Kelley (1959) (as part of interdependence theory) as the that relationships are the subject of considerable lay theorizing and
contrast between what people believe they deserve in a relation- cognitive work (both conscious and unconscious) at both the
ship (comparison level) and the perceived level of awards derived individual and cultural levels (see Berscheid, 1994; Fletcher &
from the relationship (outcomes). According to interdependence Thomas, 1996). Second, there is no shortage of material from
theory, comparison levels reflect the average amount of reward which people can develop their ideal standards. In addition to
value that can be obtained from relationships. However, recent personal experience and observation of other people's relation-
ships, individuals are subjected, on a daily basis, to voluminous
research and theorizing suggest that comparisons between prior
doses of relationship-oriented information (at least in developed
standards and perceptions of current relationships are likely to be
countries) via TV, novels, films, books, plays, and so forth. Third,
made on content-rich dimensions involving specific ideal stan-
ideals are appropriate knowledge structures to serve as standards
dards rather than on global dimensions reflecting general expec-
against which perceptions of the relationship and partner can be
tations of the rewards available in relationships. The present
gauged, subsequently influencing relationship evaluations. Fourth,
research examined the structure and function of romantic relation-
ideals are located in the right kind of cognitive "niche" to exert
considerable influence over current relationship cognition and
behavior. This last proposition is based on the notion that stored
Garth J. O. Fletcher, Geoff Thomas, and Louise Giles, Psychology relationship-relevant knowledge constructs tend to involve three
Department, University of Canterbury, Christchurch, New Zealand; Jeffry interlocking domains: the self, the partner, and the relationship
A. Simpson, Department of Psychology, Texas A&M University. (see Baldwin, 1992; Fletcher & Thomas, 1996). Figure 1 depicts
Correspondence concerning this article should be addressed to Garth the relations among these categories and provides examples of
J. O. Fletcher, Psychology Department, University of Canterbury, Christ-
beliefs (including ideals) that appropriately fit into each one.
church, Private Bag 4800, New Zealand. Electronic mail may be sent to
g.fletcher@psyc.canterbury.ac.nz. The major point we wish to illustrate in Figure 1 is that the
72
IDEALS IN INTIMATE RELATIONSHIPS 73
studies, Rusbult et al. (1993) used a combination of free-response

data and multidimensional scaling to investigate relationship ide-
als. They found two dimensions: (a) superficial versus intimate and
(b) romantic-traditional versus nontraditional. Although this re-
search provided a valuable beginning to this topic, it had some
drawbacks. First, the research was not designed to produce a
standardized set of ideal scales that could be used by other re-
searchers. Second, it did not distinguish between relationship-
oriented items (e.g., relationship equality) and items specifically
pertaining to the partner (e.g., good looks). As already noted, we
wanted to determine to what extent relationship-oriented and
partner-oriented measures are independent constructs.
Although social-cognitive research does not specify the content
of partner or relationship ideals, a plethora of research has exam-
ined other kinds of relationship-level knowledge structures, includ-

ing the concept of love (Aron & Westbay, 1996; Fehr & Russell,
1991), the perceived causes of relationship success (Fletcher &
Examples of Cognitions Kininmonth, 1992), and the concept of a "good" relationship
(Hassebrauck, 1997). Such knowledge structures, however, differ
a. I am intelligent
b. Relationships fail without good communication from ideals in certain crucial respects. Research examining the
c. Men are aggressive meaning and structure of commonly held concepts such as love has
d. Relationships work well when one partner is dominant typically used a prototype approach in which exemplars or lists of
e. I want an exciting relationship features are treated as prototypical if they are either similar to the
f. I want an honest relationship with a partner I can trust modal (or average) values on particular dimensions, or similar to
g. I am suited to someone who is sporting and athletic
members of the same family of categories (see Aron & Westbay,
Figure 1. General model of relationship cognition.
1996; Fehr & Russell, 1991). Ideals, in contrast, involve the
positive end of evaluative dimensions rather than the average or
mode (see Barsalou, 1985).
cognitions that represent ideals are most naturally located in the Constructs such as the perceived causes of relationship success
areas that overlap self and relationship-partner (Areas e, f, and g). (Fletcher & Kininmonth, 1992) and the concept of a good rela-
Ideals are paradigmatic examples of relationship-level knowledge tionship (Hassebrauck, 1997) are perhaps more similar to ideals
structures that represent overlapping yet potentially nonredundant than general concepts of love or commitment. However, partici-
constructs typically encompassing elements of the self, the poten- pants in related past research did not provide ratings of their own
tial partner, and the potential relationship. In short, they specify a ideal standards. And, even though ideals and conceptions of a
set of expectations, hopes, or standards that are truly relational in "successful" relationship might be related, they could well be
character. As a consequence, ideal standards should often be divergent. For example, a person may believe that greater passion
pressed into psychological service in specific relationship settings. usually produces successful relationships; however, this individual
This model suggests that commonly held ideals regarding po- might prefer a relatively lackluster and passionless relationship if
tential partners and relationships may, to some extent, be stored he or she has a low sex drive. Thus, personally held ideals should
separately and represented in lay cognitive schemas. For example, be more firmly connected to the self than is the case for more
a man's ideal partner might be a film star look-alike, with the general beliefs or attitudes concerning relationships (see Figure 1).
accent on passion, whereas his relationship ideal may emphasize Although few studies have measured ideals directly, one recur-
companionship and stability. Given this possibility, we assessed rent theme in past research has been the centrality of intimacy or
partner-based and relationship-based ideals separately. Nonethe- closeness in lay relationship cognition (see Fletcher & Fitness,
less, we expected that partner and relationship ideals would over- 1996). Accordingly, we predicted that items associated with inti-
lap in the manner shown in Figure 1. Indeed, these two ideal macy would constitute one basic dimension underlying people's
categories are likely to strongly cohere because people should seek ideals. More specific predictions about the content of partner ideals
out partners who are capable of helping them achieve their ideal can be derived from theory and research from an evolutionary
relationships. For example, if an individual views laughter and perspective, to which we now turn.
humor as central components of an ideal relationship, this person
should rate a sense of humor highly in his or her ideal mate. Given
the paucity of research on this topic, we were uncertain of the An Evolutionary Perspective
degree to which partner and relationship ideals would form distinct Research from an evolutionary perspective suggests that two or
constructs. three relatively stable, semi-independent dimensions should under-
A social-cognitive approach offers some guidelines about the lie conceptions of ideal partners. Simpson and Gangestad (1992)
general structure of partner and relationship ideals; however, it is had people rate the extent to which 15 common mate attributes
not particularly illuminating about the specific content of the (e.g., physical attractiveness, kindness, loyalty, and social status)
categories. In fact, only one previous investigation has explored affected their selection of a prospective romantic partner. Factor
the content of intimate relationship ideals. In a series of important analyses revealed two factors within both sexes. The first factor
74 FLETCHER, SIMPSON, THOMAS, AND GILES
was composed of attributes known to foster relationship closeness provided only weak or indirect support for this last hypothesis,
and intimacy (e.g., kindness, responsibility, loyalty, and qualities because many of the items (attributes) used in previous mate-
of a good parent). The second factor contained attributes pertaining choice studies have been selected on the basis of a particular
to the partner's attractiveness and social visibility (e.g., physical theoretical perspective. If a set of items chosen to assess two ideal
attractiveness, financial resources, and social status). factors (say, intimacy and attractiveness) is factor analyzed, then it
Recent extensions of this work, however, have indicated that the is hardly surprising that these same two factors will emerge. Past
second factor (attractiveness and social visibility) may contain two research has shown that individuals can (and do) distinguish be-
theoretically distinct components. Gangestad and Simpson (1996) tween different ideal categories, but it has not addressed whether
found that indicators of an individual's health (e.g., his or her such constructs represent the most central and accessible ideal
physical attractiveness, physical fitness, and health history) do not categories in folk psychology.
correlate highly with markers of social prominence and resources In the present research, we tested the hypothesis that the ideal
(e.g., his or her social status, social visibility, and financial re- categories we have postulated are centrally represented in lan-
sources). This new evidence suggests that three major dimensions guage and the folk psychological lexicon. We did so by inductively
may define what people use as standards to evaluate ideal part- deriving items that assessed attributes of both ideal partners and
ners; the prospective partners' capacity for intimacy and commit- ideal relationships from free-response protocols that were subse-
ment, their attractiveness and general health, and their social quently factor analyzed.
status-resources.
These three dimensions also make theoretical sense in light of
recent models of mating strategies in humans. Each dimension The Functions of Ideals: Links Among Ideals,
represents a different route to obtaining a mate and promoting Relationship Perceptions, and Relationship Evaluations
one's own reproductive fitness (see Buss & Schmitt, 1993). Past
research on mate selection (Simpson & Gangestad, 1992) and To predict the probable connections among ideals, perceptions,
mating strategies (Buss & Schmitt, 1993) has suggested that all and relationship evaluations, it is helpful to consider the functions
three domains should be germane to successful mating. By being that ideals might play. Drawing on social-cognitive theories about
attentive to a partner's capacity for intimacy and commitment, for the role of ideals in relation to the self (e.g., see Higgins, 1987,
example, an individual increases his or her chances of finding a 1989), we postulate that relationship ideals serve two basic kinds
cooperative, committed partner who is likely to be a good parent. of function: evaluative and regulatory. As applied to relationship
By focusing on attractiveness and health, an individual is more contexts, the proposed pivotal causal factor is the consistency (or
likely to acquire a mate who is younger, healthier, and perhaps discrepancy) between chronically accessible ideals and percep-
more fertile (at least in the case of men choosing women). Finally, tions "of the relationship or partner. The magnitude of the consis-
by considering a partner's resources and status, individuals should tency between ideals and perceptions, in turn, gives the individual
be more likely to obtain a mate who can ascend social hierarchies valuable information that can be used to (a) evaluate the relation-
and form coalitions with other people who have—or can acquire— ship and partner (e.g., assess the appropriateness of a potential or
valued social status or other resources (Gangestad & Simpson, current mate or relationship) and (b) regulate the relationship (e.g.,
1996). predicting and controlling the relationship and the partner).
In light of current theory and previous research, it is less clear If individuals use the consistency between their ideals and
what the factor structure of relationship ideals should be like. perceptions to evaluate the partner or relationship (the evaluative
Given the importance and centrality of intimacy in lay relationship function), then it follows that, to the extent that perceptions of
theories and knowledge structures (see Fletcher & Thomas, 1996), current partners and relationships are more consistent with ideals,
we predicted that one of the relationship ideal factors would evaluations of partners and relationships should be more positive.
probably capture a partner's capacity for intimacy and commit- We tested this prediction in the current research.
ment (similar to one of the partner-based factors). However, the A second major postulate (also derived from self-discrepancy
remaining two partner-based factors (attractiveness-health and theory) is that we would expect people to be motivated to maxi-
status-resources) do not have obvious conceptual parallels at the mize the consistency between ideals and perceptions of the partner
relationship level. Nevertheless, we expected that there would be or relationship, as a function of the general motivation to achieve
substantive correspondence between the partner and relationship- positive views of their relationships and partners. Large discrep-
level factors, given that individuals should prefer ideal partners ancies between chronically accessible ideals and perceptions will
who would facilitate the development of their ideal relationships. mean that important relationship goals and outcomes have failed to
This research extends previous work in three ways. First, past materialize. Large discrepancies should also produce strong feel-
research has not analyzed or considered separately the role of ings of discouragement, dejection, or dissatisfaction with the cur-
partner versus relationship ideals. Second, as indicated earlier, we rent partner or relationship (see Higgins, 1987). Negative emotions
wanted to test more definitively whether partner-based ideals con- resulting from large ideal-perception discrepancies, in turn, should
tain three dimensions rather than just two. Third, if evolutionary motivate individuals to do one of four things. First, individuals
theories are correct, the structure and content of ideals should be might leave the relationship. Second, individuals might engineer a
represented both in language and in accessible lay cognitive con- change in their relationships over time to produce relationships and
structs. Indeed, many evolutionary psychologists have argued that partners that are closer to their ideals (see Murray, Holmes, &
evolutionary processes in ancestral environments endowed hu- Griffin, 1996). Third, individuals might simply change their views
mans with highly domain-specific cognitive mechanisms and be- of the current partner or relationship to bring them more into line
liefs (Tooby & Cosmides, 1992). Previous research, however, has with their ideal standards. Fourth, individuals could adjust their
ideal standards to match their perceptions of the current partner or ships.. The methods and procedures adopted in both this study and
relationship. Study 2 were based, in part, on the research of Fehr and her
Note that the first two actions just described (leaving or chang- colleagues (Fehr, 1998; Fehr & Russell, 1991).
ing the relationship) represent attempts to regulate the relationship,
whereas the latter two cognitive moves (altering ideals or percep- Method
tions) are linked to the goal of evaluating the relationship. In
Participants. Participants were 100 undergraduate students attending
general, however, it is clear that there should be strong psycho-
the University of Canterbury (50 men and 50 women). The mean age of the
logical forces bringing chronically accessible ideals into line with sample was 21.42 years (SD = 4.63).
relationship and partner perceptions. Yet another factor that should Procedure. In one condition, 25 men and 25 women were asked to
produce the same push toward consistency is that chronically "build a mental picture of your IDEAL partner in a dating or marital
accessible ideals might operate like schemas, filling in gaps and relationship. Using words or phrases, describe the important characteristics
facilitating the development of specific person and relationship of the IDEAL partner. Use one slip of paper for each characteristic." In the
models. The upshot is that, in the current study, we predicted that second condition (also comprising 25 men and 25 women), the same
the importance attached to particular ideal dimensions should be instructions were used, except that participants rated the ideal relationship
positively related to perceptions of the partner and relationship on rather than the ideal partner. Previous research by Fletcher and Kininmonth
the same dimensions. (1992) found strong similarities in the perceived causes of successful
dating and marital relationships. Thus, we used the phrase dating or
However, ideals are not the only kind of benchmark against marital to make it clear to participants that we were interested in stable,
which perceptions of the relationship or partner can be compared. romantic, heterosexual relationships. No time limit was imposed on the
Perceptions of ongoing relationships and partners might also be task, and participants had access to an unlimited supply of slips of paper.
routinely compared with alternative partners or relationships per- All participants completed the task within 15 min.
ceived as readily available (Thibaut & Kelley, 1959). If alternative
romantic partners perceived as superior to the current partner are Results and Discussion
indeed available, this may produce lower levels of relationship
satisfaction or commitment. Consistent with this hypothesis, there Not surprisingly, participants occasionally described aspects of
is evidence that people who rate more highly the attractiveness of the ideal partner when describing the ideal relationship, and vice
available alternatives tend to be less committed to their current versa. Items that were clearly in the wrong category were dis-
relationships (Johnson & Rusbult, 1989). Accordingly, when test- carded (e.g., describing an ideal relationship as one in which the
ing the hypothesis that greater consistency between ideals and partner was tall, dark, and handsome or describing an ideal partner
perceptions of the current partner or relationship should be as one in which the relationship was passionate or exciting). Using
associated with more positive relationship evaluations, we con- this procedure, we discarded a mean of 1.40 items from the ideal
trolled for the perceived availability of alternative partners and partner category and a mean of 0.58 items from the ideal relation-
relationships. ship category. Two raters performed this task, and they agreed on
96% of these classifications (K = .92). Differences were resolved
by discussion. Of the remaining items, participants provided a
Overview mean of 10.36 ideal partner descriptions and 9.68 ideal relation-
This research had two broad aims and predictions. First, we ship descriptions.
wanted to identify the content and structure of partner and rela- Partner and relationship ideal items were coded separately. The
tionship ideals and to develop reliable scales that measured these descriptions were sorted into categories based on the exact word-
ideals. Second, we wished to gather evidence concerning how ing used by the respondents but omitting qualifying terms and
ideals and perceptions of specific partners and relationships are treating very similar adjectives as synonymous. For example, very
related to one another and to perceptions of relationship quality. assertive was coded as assertive, and good body, wonderful body,
We conducted six studies. In Study 1, participants generated a and excellent body were all coded as good body. The goal was to
list of attributes that described both ideal partners and ideal ro- reduce inferential or higher order coding to a minimum. Categories
mantic relationships. Study 2 was an exploratory factor-analytic that included descriptions provided by at least 3 participants (at
examination of the importance ratings of the partner and relation- least 6% of each sample) were used in Study 2. Two coders
ship items generated in Study 1. Study 3 used confirmatory factor performed these tasks independently and agreed on 96% of the
analysis (CFA) to replicate the results of and test a factorial model categorizations (K = .96). Differences were resolved by discus-
suggested by Study 2. Study 4 provided convergent and discrimi- sion. The descriptive words or phrases (n = 78) are presented in
nant validation evidence for the factors obtained in Study 2. Tables 1 and 2, along with the percentages of individuals in each
Study 5 tested the hypothesis that particular ideals and related sample who included at least one mention of each category. The
partner or relationship perceptions should be positively correlated. number of items produced was analyzed in a 2 (sex) X 2 (partner
Finally, Study 6 tested the hypothesis that relationship evaluations vs. relationship) analysis of variance (ANOVA). This analysis
should be more positive to the extent that perceptions of the revealed no significant main effects and no interactions, although
current relationship and partner are more consistent with relevant the tendency for women to generate more items (M = 10.74) than
ideals. men (M = 9.30) was marginally significant (p < .06).
Study 1 Study 2
The aim of Study 1 was to create a list of typical or common Study 2 examined the relations among and between the ideal
characteristics of both ideal partners and ideal romantic relation- partner and the ideal relationship items. The descriptive phrases
Table 1
Factor Loadings of Factor Analyses on the Importance Ratings of Partner Ideals
Rate of Importance Factor 1 Factor 2 Factor 3

mention rating (Warmth- (Vitality- (Status-
Variable (%; Study 1) (Study 2) Trustworthiness) Attractiveness) Resources)
Understanding 42 5.99 .87 -.18 .04

Supportive 12 5.95 .85 -.15 .02
Considerate 52 5.93 .83 -.11 .03
Kind 14 5.87 .81 .00 .00
Good listener 08 5.74 .75 -.04 .11
Sensitive 28 5.75 .75 .01 .08
Trustworthy 34 6.45 .72 .01 -.06
Warm 14 5.66 .71 -.03 -.04
Affectionate 18 5.93 .69 -.02 .08
Reliable 16 5.81 .68 .03 .15
Friendly 18 5.87 .67 .10 -.05

Communicative 48 6.08 .62 .14 -.13

Honest 36 6.36 .58 .07 -.14
Mature 08 5.47 .53 .11 .02
Stable 18 5.47 .52 .10 .20
Romantic 18 5.26 .46 .08 .21
Broad-minded 20 5.77 .43 .33 -.31
Easygoing 26 5.57 .42 .27 -.11
Self-aware 10 5.43 .41 .20 -.18
Generous 12 5.04 .35 .20 .15
Deals well with criticism 48 5.29 .34 .33 -.11
Likes children 10 4.43 .32 -.13 .29
Adventurous 06 5.31 -.01 .75 -.18
Nice body 18 4.82 -.20 .68 .30
Outgoing 52 5.15 .09 .65 .03
Sexy 16 5.21 -.06 .65 .16
Attractive 92 5.20 -.13 .62 .26
Good lover 18 5.48 .02 .59 .13
Active lifestyle 06 5.25 --.09 .58 .09
Sporty and athletic 36 4.34 -.22 .55 .29
Confident 16 5.39 .15 .52 -.03
Independent 34 5.50 .23 .48 -.22
Ambitious 36 4.83 .06 .47 .28
Interesting 06 6.11 .39 .46 -.11
Spontaneous 06 5.17 .16 .43 -.27
Good fun 16 6.00 .42 .43 -.08
Good sense of humor 68 5.94 .36 .42 -.11
Assertive 12 4.88 .26 .41 .08
Creative 08 4.65 .29 .35 -.04
Intelligent 84 5.54 .20 .34 -.13
Good job 06 3.85 .04 .21 .75
Financially secure 12 3.89 .10 .08 .73
Nice house or apartment 08 2.79 -.07 .16 .69
Appropriate ethnicity 06 2.56 -.10 .03 .61
Successful 08 4.31 .14 .32 .58
Dresses well 10 4.25 -.02 .30 .56
Appropriate age 06 3.83 -.05 .24 .42
Religious beliefs 06 2.44 .07 -.14 .30
Does not smoke 06 5.08 .08 -.04 .24
Note. N = 50 (Study 1); N = 320 (Study 2). Factor loadings of .40 and higher are in boldface type.
and words generated in Study 1 were rated by a large sample of but we did anticipate that an intimacy-commitment set of items
undergraduate students in terms of their importance for their own would constitute one of the relationship factors.
ideal partners and ideal relationships. On the basis of evolutionary
theory (Gangestad & Simpson, 1996) and research (Simpson & Method
Gangestad, 1992), we expected that three factors would underlie
Participants. Three hundred twenty undergraduate students at the Uni-
the ideal partner items, reflecting an ideal partner's capacity for versity of Canterbury completed this phase of the research (165 women and
intimacy and commitment, his or her attractiveness and general 155 men). Of the sample, 147 were not involved in a heterosexual rela-
health, and his or her social status and resources. We had no tionship, 127 were dating, 34 were living together, and 12 were married.
specific predictions about the number of relationship ideal factors, The mean age of the sample was 22.56 years (SD = 4.92). As a means of
Table 2
Factor Loadings of Factor Analyses on the Importance Ratings of Relationship Ideals
Rate of mention Importance rating Factor 1 Factor 2

Variable (%; Study 1) (Study 2) (Intimacy-Loyalty) (Passion)
Honest 48 6.39 .80 -.06

Commitment 16 5.76 .78 -.22
Caring 36 5.99 .78 -.01
Trusting 26 6.36 .77 .02
Support 26 5.86 .74 .07
Respect 24 6.13 .69 .07
Understanding 36 5.90 .67 .13
Friendship 24 6.39 .66 .08
Good communication 40 6.31 .66 .05
Loyalty 08 6.10 .60 .11
Stability 16 5.30 .54 .06
Monogamous 14 6.07 .50 -.11

In love 40 5.90 .50 -.04

Affectionate 16 5.77 .47 .20
Equality 12 5.92 .45 .26
Sharing 18 5.63 .44 .26
Acceptance 06 5.60 .41 .29
Compromise 16 5.36 .36 .34
Exciting 28 5.33 -.15 .82
Challenging 10 4.71 -.21 .69
Humorous 16 5.68 .03 .66
Fun 28 5.87 .15 .63
Independence 32 5.63 -.02 .57
Passionate 32 5.42 .05 .54
Intellectual equality 24 5.01 .05 .52
Relaxed 12 5.63 .15 .50
Similar personalities 30 4.47 .06 .40
Romantic 12 5.18 .21 .40
Confronts conflict 08 5.13 .25 .38
Similar interests 32 5.09 .22 .33
Note. N = 50 (Study 1); N = 320 (Study 2). Factor loadings of .40 and higher are in boldface type.
establishing test-retest reliability, an independent sample of 42 participants value scree test and the amount of variance explained (42.1%).
(17 men and 25 women) completed the scales twice, with a 3-week gap Moreover, the factor loadings were clean, interpretable, and con-
between testing sessions. Of this latter sample, 22 participants were not sistent with our prior theorizing (eigenvalues above 1.0 were as
involved in a heterosexual relationship, 10 were dating, 7 were living follows: 12.5, 5.2, 3.0, 2.0, 1.9, 1.6, 1.3, 1.2, 1.1, and 1.1). For the
together, and 3 were married. The mean age of this sample was 22.98 years
ideal relationship items, we had no particular theoretical expecta-
(SD = 4.74).
tions, but a two-factor solution was suggested by a scree test
Procedure. Participants were asked to rate the ideal partner items
(shown in Table 1) "in terms of the importance that each item has in
(eigenvalues above 1.0 for this analysis were 9.6, 2.5, 1.7, 1.5, 1.3,
describing your IDEAL PARTNER in a close relationship (dating, living and 1.1; variance explained with two factors: 40.2%). In addition,
together, or married)." Each item was answered on a 7-point Likert scale the factor loadings were clearly interpretable and clean.
ranging from very unimportant (1) to very important (7). The same format The loadings from the factor pattern matrices for both factor
was used for the ideal relationship items (see Table 2), with participants solutions are reported in Tables 1 and 2, with loadings of .40 or
being asked to rate each item "in terms of the importance that each item has higher highlighted. The interfactor correlations for the partner
in describing your IDEAL CLOSE RELATIONSHIP (dating, living to- ideal factors were .34 between Factor 1 and Factor 2, —.09
gether, or married)." The items were randomly mixed within each ques-
between Factor 1 and Factor 3, and .10 between Factor 2 and
tionnaire type and presented in one fixed order, and the order in which the
Factor 3. For the two relationship ideal factors, the interfactor
ideal relationship and ideal partner scales were completed was counterbal-
anced within each sex. Order of scale completion did not influence the correlation was .46.
results and is not discussed further. The interpretation of these factors seems relatively straightfor-
ward. Consistent with expectations, three factors defined the part-
ner ideals: (a) personal characteristics that appeared particularly
Results and Discussion
relevant to the development of intimacy and loyalty (labeled
Factor analyses. We first conducted two exploratory factor Partner Warmth-Trustworthiness); (b) personality and appearance
analyses (principal-components analyses with oblique Harris- characteristics related to how attractive, energetic, and healthy the
Kaiser rotations) on the items contained in the ideal partner and the person was (labeled Partner Vitality-Attractiveness); and (c) char-
ideal relationship scales. For the ideal partner items, as expected, acteristics related to the social status and resources the person
a three-factor solution provided a good fit in terms of the eigen- possessed (labeled Partner Status-Resources). The relationship
Table 3
Ideal Scale Internal Reliabilities, Test-Retest Reliabilities, and Means
Men Women
Internal Test—retest
Factor reliability reliability M SD M SD
Partner Warmth-Trustworthiness .93 .75 5.97 0.63 5.66 0.76

Partner Vitality-Attractiveness .88 .86 5.27 0.72 5.31 0.74
Partner Status—Resources .82 .85 3.72 1.07 3.58 1.14
Relationship Intimacy-Loyalty .91 .76 6.17 0.60 5.76 0.74
Relationship Passion .79 .82 5.22 0.77 5.13 0.82
Note. Means and standard deviations are expressed in terms of scores on a 7-point scale. N = 320 (Study 2).
ideal items formed two factors: (a) the importance of intimacy and Correlations among ideal scale scores. Correlations calcu-
stability in a relationship (labeled Relationship Intimacy-Loyalty) lated among the ideal scales are shown in Table 4. As expected, the
and (b) how passionate and exciting the relationship was (labeled highest correlations were obtained between the relationship and
Relationship Passion). partner ideal dimensions that measured similar constructs. People
Internal and test-retest reliabilities. Table 3 shows the inter- who valued more highly an ideal partner who was warm and
nal reliability coefficients (Cronbach alphas) and the test-retest trustworthy also rated the presence of intimacy and loyalty in an
reliabilities for. each scale. Each scale was constructed by unit ideal relationship as more important. In addition, people who
weighting the items that loaded .40 or greater on each factor. The placed more importance on vitality and attractiveness in a potential
internal reliabilities and test-retest reliabilities for each scale were mate also rated a passionate relationship as more important. Part-
all acceptably high and positive. ner Status-Resources correlated less highly with the other ideal
Mean differences in ideal scale scores. The mean importance scales. However, those valuing a partner who possessed high status
ratings for each ideal category are shown in Table 3." The partner and resources also rated a passionate relationship and obtaining a
ideal scores were initially analyzed in a 2 (sex) X 3 (ideal cate- partner high in vitality and attractiveness as more important. This
gory) ANOVA with the second variable as a repeated measure. pattern of correlations was very similar when calculated within
The analysis revealed significant main effects for sex, F(l, 318) = groups according to sex or relationship status (currently involved
3.91, p < .05, and for ideal category, F(2, 636) = 756.62, p < vs. not Involved in a romantic relationship).
.001. However, these two main effects were qualified by a signif- To take into account possible method variance and to clarify the
icant interaction between sex and ideal category, F(2, 636) = 4.41, correlations across partner and relationship ideals, we calculated
p < .05. The relationship ideal scores were also initially analyzed residuals for each partner ideal category while partialing out the
in a 2 (sex) X 2 (ideal category) ANOVA; this analysis revealed remaining partner ideal categories. We also calculated residuals for
significant main effects for sex, F(l, 318) = 13.31, p < .001, and each of the relationship categories while partialing out the remain-
for ideal category F(l, 318) = 380.30, p < .001. However, as ing relationship ideal category (see Aron & Westbay, 1996). We
previously, these two main effects were qualified by a significant then correlated these residuals across the partner and relationship
interaction between sex and ideal category, F(l, 318) = 15.66, p < categories. The results showed that the convergent correlations
.001. 2 remained high, positive, and significant (Partner Warmth-
To clarify the results, we carried out post hoc contrasts between Trustworthiness and Relationship Intimacy-Loyalty, r = .69, and
group means using t tests with alphas set at p < .01 (using a Partner Vitality-Attractiveness and Relationship Passion, r = .45),
Bonferroni adjustment). These contrasts showed that all three whereas the discriminant correlations across categories were closer
partner ideal categories were significantly different from one an- to zero and became negative (Partner Warmth-Trustworthiness
other for both men and women at thep < .001 level (see Table 3). and Relationship Passion, r = —.12, and Partner Vitality-
For both sexes, warmth-trustworthiness was rated most highly, Attractiveness and Relationship Intimacy-Loyalty, r = —.29).
followed by vitality-attractiveness; status-resources was rated as In summary, these results suggest that people's ideal partners
the least important ideal category. The relationship ideals also and ideal relationships have a relatively clear structure generally
revealed a consistent pattern, with intimacy-loyalty rated signifi- consistent with our predictions. Specifically, the partner items
cantly higher than passion for both men and women at the p < .001 reduced to three factors labeled Warmth-Trustworthiness,
level. Comparisons across sex of rater for the five ideal categories Vitality-Attractiveness, and Status-Resources. The relationship
(see Table 3) revealed that men rated both partner warmth- items produced two factors labeled Intimacy-Loyalty and Passion.
trustworthiness and relationship intimacy-loyalty as significantly All five ideal scales had adequate psychometric properties,
more important than did women (p < .001). These results were
unchanged when the same comparisons across sex were calculated
with the responses to the remaining ideal factors partialed out, 1
Mean scores for the individual items are shown in Tables 1 and 2.
which shows that these sex differences were not simply a product 2
A series of ANOVAs was also conducted, adding the variable of those
of the male ratings being generally higher than the female ratings.
currently involved in a romantic relationship (n = 178) versus those who
There were no significant sex differences for the remaining three
were not (n = 152). No significant main effects or interactions were found
ideal categories. for this variable.
Table 4
Correlations Among Partner and Relationship Ideal Scale Scores
Factor 1
1. Partner Warmth-Trustworthiness .49* .10 .80*** 49***

2. Partner Vitality-Attractiveness .39*** .32*** .60***
3. Partner Status-Resources — .06 .17*
4. Relationship Intimacy-Loyalty .50***
5. Relationship Passion
Note. N = 320 (Study 2).

*p<.05. ***/><.001.
including good internal consistency and good test-retest reliability. Results

Moreover, our expectation that partner and relationship ideals

would form semi-independent, overlapping constructs was con- To reduce the number of variables for the CFA, we divided the
firmed. Specifically, and most obvious, individuals who rated their items measuring each of the five ideal factors into three equal
ideal relationship as high on either intimacy-loyalty or passion groups and then summed to produce three observed variables that
wanted partners who possessed attributes that should foster the loaded on each of the ideal factors.3 All CFAs were completed via
development of such relationships. The size of the correlations version 5.5 of EQS for Windows (Bentler & Wu, 1995).4 Because
across the partner and relationship ideals suggests that they are not there was some evidence of negative skew and multivariate kur-
entirely distinct constructs. This possibility was tested further in tosis, we used robust measures of fit, in accordance with the advice
Study 3. of Bentler and his colleagues (see Bentler, 1993; Chou, Bentler, &
Satorra, 1991). The main index of fit used was the comparative fit
Study 3 index (CFI), which is not adversely affected by sample size and is
considered to indicate a good fit at a level above .90.
Study 3 had two objectives. First, given the results of Study 2, Model comparisons. The first model we tested treated each
we wished to test different models of the general factor structure observed variable as loading on a single factor, reflecting the
underlying both the partner and the relationship ideal scales using tendency for participants to rate all ideal items across all categories
CFA. Second, we wanted to replicate the factor structure found in in a similar fashion. This analysis revealed a poor fit (CFI = .58).
Study 2 using CFA on an independent sample. For the second model, we loaded each set of three items onto the
Study 2 revealed that some of the ideal relationship scales were five relevant higher order latent variables representing each ideal
heavily correlated with some of the ideal partner scales. Specifi- factor (but leaving the factors as uncorrelated). This model pro-
cally, the partner ideal of warmth-trustworthiness was strongly duced a somewhat better but still weak fit (CFI = .66). In the third
and positively correlated with the relationship ideal of intimacy- model, we loaded each of the three first-order partner ideal factors
loyalty, the relationship passion ideal dimension was strongly and on one higher order factor representing partner ideals, and we also
positively correlated with the vitality-attractiveness ideal partner loaded the two first-order relationship ideal factors on a single
dimension, and the partner ideal status-resources dimension was higher order factor representing relationship ideals. This model
weakly to moderately correlated with the partner vitality-
produced an improved fit, but the value was still well below .90
attractiveness and relationship passion ideals. Therefore, we tested
(CFI = .76).
several competing models in Study 3. We predicted that the
The fifth model, shown in Figure 2, produced a much superior
best-fitting model would consist of five lower order factors (the
fit to all of the prior models tested (CFI = .90), Satorra-Bentler
three ideal partner and the two ideal relationship factors, described
in Study 2) but that these lower order factors would load on two scaled ^ C 8 4 . N = 181) = 210, p < .001. Moreover, both the
higher order factors (Warmth-Loyalty and Vitality-Status- standardized lower order and higher order paths were high, sig-
Passion), as depicted in Figure 2. nificant at the/7 < .05 level, and positive (ranging from .59 to .99).
It should be noted that when we repeated this model but dispensed
with the lower order factors so that the two sets of items loaded
Method
Participants. One hundred eighty-one undergraduate students enrolled
3
at the University of Canterbury completed this phase of the research (103 So that each summed item was of roughly equivalent internal reliability
women and 78 men). Of the sample, 154 were dating, 22 were living within each ideal category, each set of three items was produced by
together, and 5 were married. Participants had been involved in heterosex- combining items with the highest and lowest loadings from within each
ual relationships for a mean of 24.20 weeks (SD = 24.32 weeks). The mean ideal category. For example (see Table 1), the three items for partner and
age of the sample was 21.45 years (SD = 4.62). status ideals were summed products of (a) good job and appropriate age,
Procedure. Participants completed the partner and relationship ideal (b) financially secure and dresses well, and (c) nice house or apartment,
scales described in Study 2. As in Study 2, the order of scale completion appropriate ethnicity, and successful.
4
was counterbalanced within sex. Order of scale completion did not influ- The correlation matrix for all of the items in this CFA is available from
ence the results and is not discussed further. Garth J. O. Fletcher.
80
-
•
' Relations Relationshi

—
*
Intimacy-L Passion
/IN
f i\
~~-
O
ON
o
00
/ 2 £> \-—-^ °°
OS
FLETCHER, SIMPSON, THOMAS, AND GILES
.
^
>
>
en
>
>
Si
SJ
(4-1
O
i
9
o
•a
u
eo
•y
_o
13
£•
[at seales, order and second- erl atent factors. Al loadings a
ii
-4->
c II
•a
•
in
O *3
o
1>
•H
a en>-,
£2-
directly onto the two higher order factors, the fit again became a high order of replication across two independent samples and
unsatisfactory (CFI = .69). across sex. These results suggest that the five ideal factors exist as
As noted by Marsh and Hocevar (1985), higher order models of semi-independent structures that nevertheless may be represented
the sort shown in Figure 2 attempt to explain the covariation in terms of two overarching sets of ideals.
among the first-order factors in a parsimonious way. Conse- It is also important to note that our comparisons tended to rule
quently, the fit of a model with one or more higher order factors out alternative models, including (a) a model in which all of the
cannot be greater than the fit of the corresponding first-order ideals loaded on one factor, (b) a model in which the five ideal
model in which all of the first-order factors are allowed to be factors were uncorrelated, (c) a model in which the partner ideals
intercorrelated. Thus, the most rigorous test of the model shown in loaded on an one overall partner factor and the relationship ideals
Figure 2 is to compare the fit of this model with one in which all loaded on one overarching relationship factor, and (d) a model in
of the first-order factors are intercorrelated. This latter model (with which the ideals loaded directly onto the two higher order factors,
all first-order factors intercorrelated) produced a CFI of .90, which omitting the five lower order factors.
was the same as the fit attained by the higher order model in
Figure 2. A precise measure of the disparity of fit between the
Study 4
higher order and lower order models is the target coefficient

developed by Marsh and Hocevar (1985). The target coefficient is Study 4 was designed to test the convergent and discriminant
the ratio of the chi-square value of the first-order model to the validity of the ideal scales. We considered convergent and dis-
chi-square value of the second-order model. Marsh and Hocevar criminant validity in relation to two classes of variables: general
suggested that a target coefficient above .90 shows that the higher beliefs or traits relevant to relationships in general and judgments
order model effectively accounts for the covariance among the about specific relationship outcomes such as relationship-quality
first-order factors. The target coefficient for the higher order model evaluations (see Fletcher & Thomas, 1996).
shown in Figure 1 was .94, suggesting that an excellent fit was
obtained.
Convergent Validity
Factorial replication of higher order model. To determine
whether the factor structure and loadings from Study 3 (shown in Generally speaking, the beliefs that people have about the
Figure 2) replicated the same structure reported in Study 2, we causes of relationship success should be related to the kinds of
conducted a multiple-sample CFA (Bryne, 1994; Marsh & Hoce- relationship and partner ideals they possess. Using constructs
var, 1985). We first tested for fit across the two samples with no assessed by the Relationship Beliefs Scale (Fletcher & Kinin-
equality constraints (total N = 501). Next, all first-order and month, 1992), we hypothesized that (a) stronger beliefs about the
second-order factor loadings across the two samples were con- importance of intimacy in producing successful relationships
strained to be equal. If the factor loadings replicate perfectly across should predict more importance being given to partner warmth-
the two samples, there should be no difference in fit between the trustworthiness and relationship intimacy-loyalty ideals, (b) stron-
two analyses. The results showed a high degree of replication, with ger beliefs about the importance of passion in producing successful
no change in the CFI and a nonsignificant increase in the chi- relationships should predict more importance being placed on the
square value ^(15) = 19, p > .20, in the constrained model partner vitality-attractiveness and relationship passion ideals, and
relative to the model in which all factor loadings (shown in Figure (c) stronger beliefs about the importance of external factors affect-
2) were free to vary across both samples. ing relationships (e.g., financial problems and children) should
The same procedures were used to test whether the same higher predict more importance being assigned to the partner status-
order model replicated across sex. To accomplish this, we com- resources ideal.
bined Study 2 and study 3 samples to produce a sample of men The other measure used to test convergent validity tapped indi-
(n = 233) and women (n = 268). The results showed a high degree vidual differences in sociosexual orientation, as assessed by the
of replication across groups of men and women, with no change in Sociosexual Orientation Inventory (Simpson & Gangestad, 1991).
the CFI and a nonsignificant increase in the chi-square value, Sociosexuality refers to individual differences in willingness to
^(15) = 15, p > .30, in the constrained model relative to the have sex in the absence of closeness, commitment, and strong
model in which all factor loadings (shown in Figure 2) were set emotional bonding. Simpson and Gangestad (1991) have docu-
free across both samples. An inspection of the models for both the mented that the more individuals exhibit an "unrestricted" socio-
male and female samples revealed patterns of lower order and sexual orientation (e.g., a willingness to engage in sex without
higher order loadings that were very similar to those shown in intimacy), the less likely they are to have relationships character-
Figure 2. ized by high levels of commitment, stability, and intimacy. Thus,
we predicted that the more strongly individuals exhibited an un-
restricted sociosexual orientation, the less importance they would
Discussion place on the partner warmth-trustworthiness and relationship
These results both replicate and extend the results from Study 2. intimacy-loyalty ideals.
They show that the subscales assessing each ideal factor are We also predicted that relationship length and evaluations of
reliable and internally consistent but that they also load on two relationship quality would be positively related to the ideals of
higher order factors running across both the ideal partner and the partner warmth-trustworthiness and relationship intimacy-loyalty.
ideal relationship. We labeled these two higher order factors People who strongly endorse ideals that stress the maintenance of
Warmth-Loyalty and Vitality-Status-Passion. Moreover, the fac- close and loyal relationships with warm and trustworthy partners
torial structure (both lower order and higher order paths) showed should work especially hard at maintaining and nurturing their
own relationships, and consequently they should be involved in than .40 were not included in the scales (see Tables 1 and 2). Ideals scale
longer term and more satisfying relationships. These predictions scores were computed for each participant by summing the items that
are consistent with previous research using the Relationship Be- compose each scale (Partner Warmth-Trustworthiness, Partner Vitality-
liefs Scale, which has shown that the more strongly people believe Attractiveness, Partner Status-Resources, Relationship Intimacy-Loyalty
and Relationship Passion).
that intimacy determines relationship success, the more positively
Participants also completed three additional scales. The first measure
they evaluate their relationships (Fletcher & Kininmonth, 1992;
was the Relationship Beliefs Scale (Fletcher & Kininmonth, 1992), which
Fletcher, Rosanowski, & Fitness, 1994). In contrast, other catego- assesses beliefs about the causes of relationship success in general. This
ries of relationship beliefs (e.g., the importance of passion) are not instrument has four subscales that assess the strength of beliefs about
as strongly and directly related to relationship evaluations. intimacy (e.g., "the best relationships depend on being absolutely loyal to
one another"), passion (e.g., "Without good sex, relationships do not
Discriminant Validity survive"), external factors (e.g., "Financial problems wreck relationships"),
and independence (e.g., "It is essential for partners to remain individuals no
In addition, we predicted several null correlations between the matter how close they are"). There is good evidence for the reliability and
ideal scales and (a) the Relationship Beliefs Scale, (b) the Socio- validity of these subscales (Fletcher & Kininmonth, 1992; Fletcher et al.,
1994).
sexual Orientation Inventory, and (c) the length and general eval-
uations of relationships. For example, we expected that neither the A six-item scale designed for use with either married or unmarried
partner warmth-trustworthiness nor the relationship intimacy- individuals was used to assess relationship quality (see Fletcher & Fitness,
loyalty ideal would correlate significantly with the strength of 1990). This scale assesses the amount of love, happiness, satisfaction,
stability, conflict, and commitment that exists in the current relationship.
beliefs about the importance of passion or the role of external
The scale has good internal reliability, convergent validity, and predictive
factors in producing relationship success (as measured by the validity (Fletcher, Fitness, & Blampied, 1990; Fletcher & Kininmonth,
Relationship Beliefs Scale). 1992). Finally, participants completed the Sociosexual Orientation Inven-
As a means of testing the possibility that responses to the ideal tory, which also has good reliability and validity (Simpson & Gangestad,
scales represent a tendency to express socially desirable responses, 1991). Scales were presented in a different random order for each
participants also completed the Crowne-Marlowe Social Desirabil- participant.
ity Scale (Crowne & Marlowe, 1964). We predicted that the ideal
ratings would not be significantly correlated with participants' Results and Discussion
social desirability scores.
The pattern of correlations obtained is shown in Table 5. Rela-
tionship length was transformed by means of a log transformation,
Method
to correct for a nonnormal distribution, before the correlations
Participants. Fortyjone women and 35 men at the University of Can- were calculated. As can be seen, 9 of the 11 predicted convergent
terbury, all of whom were involved in heterosexual relationships, com- correlations were statistically significant. In particular, people who
pleted all of the scales just mentioned. Of this sample, 58 participants were held stronger beliefs about the importance of intimacy in produc-
dating, 15 were living together, and 3 were married. Participants reported
ing successful relationships rated partner warmth-trustworthiness
how long they had been in their current relationship (in years and months).
and relationship intimacy-loyalty ideals as more important.
The mean length of relationship was 21.80 months (SD = 25.10 months).
The mean age of participants was 22.10 years (SD = 5.10). Participants Those who rated passion more highly as a cause of relationship
were each paid $10 for taking part. success gave higher importance ratings to the partner vitality-
Procedure. The relationship and partner ideal scales were completed in attractiveness and relationship passion ideals. Participants who
the same format described in Study 2. Items with factor loadings of less rated external factors as more important in influencing relationship
Table 5
Convergent and Discriminant Correlations Between Partner and Relationship Ideals and Other
Measures
Partner ideal Relationship ideal
Warmth- Vitality- Status- Intimacy-

Scale trustworthiness attractiveness resources loyalty Passion
Social desirability .13 .00 .21 .07 -.11

Relationship quality .16 .14 .16 .24* .26*
Relationship length .20 -.10 .12 .29* -.10
Sociosexual orientation -.32** .10 -.17 -.30** .07
Relationship beliefs
Intimacy .35** .14 .20 .38** .08
Passion .17 .44** .26* .09 .38**
External .10 .02 .46** .17 -.01
Independence .16 .21 .21 .19 .13
Note. Correlations for which significant convergent correlations were predicted are shown in boldface. N = 76
(Study 4).
*jt><.05. **/><.01.
success placed more importance on the partner status-resources In addition, participants completed an adaptation of the relationship
ideal. Participants who were more unrestricted in sociosexuality and partner ideal scales that assessed judgments of ideal-related behav-
rated the partner warmth-trustworthiness and relationship iors in their current relationships. The scale asked participants to "rate
intimacy-loyalty ideals as less important. Finally, participants who each item in terms of how accurately each one describes your ACTUAL
CURRENT CLOSE RELATIONSHIP (dating, living together, or mar-
reported higher relationship quality and who were involved in
ried)." The items rated were identical to those listed in the relationship
longer relationships placed more importance on the relationship
ideal scale (see Table 2), with each item being answered on a 7-point
intimacy-loyalty ideal. These convergent correlations ranged from Likert scale (1 = not at all like my relationship, 7 = very much like my
.24 to .46, suggesting that the constructs being measured are relationship). The partner perception scale asked participants to "rate
similar but not identical.5 each item in terms of how accurately each term describes your AC-
In terms of discriminant validity, only 2 of the 29 correlations TUAL PARTNER in your CURRENT relationship (dating, living to-
for which we predicted nonsignificant effects were significant; gether, or married)." The items listed were identical to those described
namely, greater relationship quality was associated with higher in the partner ideal scale (see Table 1), with each item being accom-
panied by a 7-point Likert scale (1 = not at all like my partner, 1 = very
ratings on the relationship passion ideal, and higher ratings on the
much like my partner). The three scales were completed in counterbal-
partner status-resources ideal were associated with more impor-
anced order within each sex.
tance being placed on passion in producing successful relation-

ships. The occurrence of 2 significant correlations is approxi-

mately what would be expected by chance, although the findings
Results and Discussion
do provocatively suggest that passion ideals and beliefs might have
a more general influence than we expected. The ratings were first summed to produce total scores for each
To summarize, the entire pattern of convergent and discriminant of the five scales in terms of both the importance of each ideal and
correlations provides support for the construct validity of the ideal the corresponding perceptions of the current partner and relation-
scales. Moreover, as expected, the pattern of correlations for the ship. The internal reliabilities for each of the five subscales of the
partner ideal of warmth-trustworthiness was very similar to the partner and relationship perceptions ranged from .76 to .93 (similar
pattern for the relationship ideal of intimacy-loyalty, and the same to those attained for the ideal subscales).
was true for the pattern of correlations involving both the partner Correlations among relationship quality, the ideal ratings, and
ideal of vitality-attractiveness and relationship passion. These perceptions of the current partner or relationship are displayed in
results corroborate the earlier findings revealing the strong overlap Table 6. Consistent with Study 4, reports of relationship quality
between the partner and relationship ideal dimensions. attained small but significant positive correlations with the ideal
ratings of partner warmth-trustworthiness, relationship intimacy-
Study 5 loyalty, "and relationship passion. Also, in accord with expecta-
In our previous discussion, we argued that two processes should tions, the ideal ratings were positively and significantly correlated
produce a strong press toward consistency between ideals and with perceptions of the actual (current) partner and relationship on
perceptions of partners and relationships. First, people should be the same dimensions. Because the ideal ratings were positively
motivated to maximize the consistency between ideals and per- intercorrelated (rs = .30 to .76), as were the set of partner-
ceptions of their partner-relationship as a function of a general relationship perceptions (rs = .14 to .77), it is not surprising that
motivation to achieve positive views of their relationships and the off-diagonal discriminant correlations shown in Table 6 also
partners. Second, chronically accessible ideals should operate like tended to be positive and significant. However, the discriminant
schemas, filling in gaps and facilitating the development of spe- validity correlations for each variable were smaller than the diag-
cific person and relationship models. onal correlations (with two exceptions).
In Study 5, we tested the prediction that importance ratings of To further test the convergent validity correlations, we summed
ideals would be significantly and positively related to self-reports the ratings (for both ideals and actual partner-relationship percep-
of ideal-relevant behaviors exhibited in the current relationship. tions) to produce the two superordinate factors suggested by the
This study can also be construed as providing a further test of the CFA: Warmth-Loyalty and Vitality-Status-Passion. Next, we re-
convergent validity of the ideal scales. calculated the convergent validity correlations between ideals and
actual perceptions for each of the two dimensions, partialing out
Method the other set of ratings for both ideals and actual perceptions (see
Participants. Participants completed the relationship quality scale, the Aron & Westbay, 1996). The convergent correlations remained
ideal scales, and scales assessing ideal-related behavior in their relation- positive and significant (p < .05) for both superordinate ideal
ships. This sample consisted of 44 women and 39 men, all students at the factors (Warmth-Loyalty, r = .31; Vitality-Status-Passion, r =
University of Canterbury who were involved in heterosexual relationships. .44).6 Overall, these results are consistent with our argument that
Of the sample, 57 participants were dating, 21 were living together, and 5 powerful psychological forces will tend to produce consistency
were married. The mean length of relationship was 18.80 months (SD =
22.30 months), and the mean age of the sample was 21.80 years (SD =
4.96). Participants were each paid $5 for taking part.
5
Procedure. Participants completed the relationship and partner ideal The same analyses were also conducted on men and women separately.
scales and the six-item relationship quality measure (Fletcher & Fitness, Although the significance levels were slightly different (as a result, in part,
1990). As before, ideal scale scores were computed for each partici- of the loss of power), the correlational patterns were similar for both sexes.
6
pant by summing the items that composed each scale: Partner The same analyses were also conducted on men and women separately.
Warmth-Trustworthiness, Partner Vitality-Attractiveness, Partner Status- Although the significance levels were slightly different (as a result, in part,
Resources, Relationship Intimacy-loyalty, and Relationship Passion. of the loss of power), the correlational patterns were similar for both sexes.
Table 6
Zero-Order Correlations Among Relationship Quality, Ideal Ratings, and Perceptions of the Partner and Relationship
Partner ideal Relationship ideal
Warmth- Vitality- Status- Intimacy- Relationship

Relationship-partner perception trustworthiness attractiveness resources loyalty Passion quality
Partner warmth-trustworthiness .36* .37* .10 .23* .45* .72*

Partner vitality-attractiveness .17 .38* .26* .13 .41* .44*
Partner status-resources .13 .28* .47* .10 .23* .07
Relationship intimacy-loyalty .35* .22* .14 .49* .39* .84*
Relationship passion .21 .31* .13 .21 .48* .58*
Relationship quality .29* .13 .13 .37* .25* —
Note. Convergent correlations are shown in boldface on the diagonal. N = 83 (Study 5).
*/><.05.
between ideal standards and perceptions of the current partner and partner has a good job, relationship is committed, relationship is support-
relationship.7 ive, relationship is exciting, and relationship is fun. The scales were
administered in a counterbalanced fashion with the six possible orders of
the three questionnaires represented as equally as possible.
Study 6
For the first scale, The participants were instructed to "rank the following
According to our previous arguments, a major way in which items in terms of HOW IMPORTANT each item is in terms of your ideal
individuals can evaluate the merits of their current partner and relationship/partner in a close relationship (dating, living together, or married).
Rank each item from 1-10. Do not give the same number to two or more
relationship is by comparing their current partner or relationship
factors: 1 = your most important ideal; 10 = your least important ideal."
against their ideal standards. If perceptions of the current partner or
The second scale required participants to "rank the following items in
relationship approximate an individual's ideal standards, evalua-
terms of the extent to which each item ACCURATELY DESCRIBES your
tions of the relationship should be positive. It is conceivable, of current (actual) relationship/partner (dating, living together, or married).
course, that relationship evaluations could be based on compari- Rank each item from 1—10. Do not give the same number to two or more
sons between that individuals think they have compared and how factors: 1 = the most accurate description of my relationship/partner; 10 =
easily they believe they could find alternatives who are as good as the least-accurate description of my relationship/partner."
or better than their current relationship or partner (Johnson & The third scale asked participants to "imagine that you were not in a
Rusbult, 1989; Thibaut & Kelley, 1959). This latter possibility was close relationship (dating, living together, or married), but that your life
also addressed in Study 6. circumstances were otherwise the same as they are now. Rank the follow-
ing items in terms of the extent to which each would be DIFFICULT to
In Study 6, participants were required to rank a range of ideals
realistically find in a new partner and form a new romantic relationship that
(in terms of their importance) that were selected from the five fit each of the factors. Rank each item from 1-10. Do not give the same
partner and relationship ideal categories. These same items were number to two or more factors: 1 = the most difficult factor to find in a
also ranked according to how accurately they described the current new partner/relationship; 10 = the least difficult factor to find in a new
partner or relationship and how difficult it would be to realistically partner/relationship."
find a new partner or relationship that matched each ideal dimen-
sion. These scales were then correlated within participants to
7
produce indexes of consistency between partner and relationship We also used a multiple regression approach to test whether relation-
ideals and perceptions of the current partner or relationship. We ship evaluations would be more positive to the extent that perceptions of
predicted that individuals who reported higher consistency be- the current partner and relationship were more consistent with ideals in
tween their ideals and perceptions of their current relationship each category. In this approach, each variable is initially entered as a main
would also report higher relationship quality, even when judg- effect, after which the interaction term (i.e., Ideal X Perception) is entered
to determine whether it accounts for increased variance over the main
ments of the availability of alternative partners or relationships
effects. Although such analyses are notoriously conservative (Alexander &
were controlled.
DeShon, 1994), the results revealed two significant interactions. Higher
levels of consistency between both partner warmth-trustworthiness and
Method relationship intimacy—loyalty ideals and related perceptions of the current
Participants. Sixty-three women and 26 men participated in Study 6. partner or relationship were attained by those who rated their relationship
All participants attended the University of Canterbury, and all were in- more positively (relative to participants who rated their relationship less
volved in heterosexual relationship. Of the sample, 60 were dating, 22 were positively).
8
living together, and 7 were married. The mean length of relationship The number of items chosen (10) was a compromise between having
was 24.07 months {SD = 30.82 months), and the mean age of the sample enough items to create a reliable measure that sufficiently tapped each ideal
was 22.01 years (SD = 4.99). scale and not having so many that participants would have difficulty giving
Procedure. Participants completed the relationship quality measure meaningful rankings. The particular items were chosen because they had
(Fletcher & Fitness, 1990), along with three other scales in which the high loadings in the factor analysis reported in Study 2 and, thus, were
same 10 ideal items were rated (2 items were selected from each factor).8 good exemplars of each ideal category. In addition, items with similar
chosen items were as follows: partner is sensitive, partner is warm, partner means in Study 2 were chosen to allow for the maximum possible variance
is adventurous, partner is physically attractive, partner is financially secure, in the rankings.
Results and Discussion ner factors were Partner Warmth-Trustworthiness, Partner Vi-
tality-Attractiveness, and Partner Status-Resources. The Partner
Measures of consistency between ideals and perceptions of the Warmth-Trustworthiness dimension included personal character-
current partner or relationship and between perceptions of the istics that seem particularly relevant for developing an intimate
current partner-relationship and the difficulty of obtaining the and loyal relationship (e.g., supportive, sensitive, trustworthy, hon-
same features in a new partner or relationship were calculated by est, communicative, and affectionate). The Partner Vitality-
computing within-subject rank-order correlations for each set of Attractiveness dimension assesses an ideal partner's attractiveness,
ranks. The mean within-subject correlation for the actual-ideal health, and personality characteristics that reflect energy, adven-
consistency measures was higher (M p = .54, SD = .40) than the turousness, and general vigor. The Partner Status-Resources di-
equivalent mean within-subject correlation for the actual- mension assesses age, job, financial resources, and so forth. Re-
alternative consistency measure (M p = .17, SD = .52). Before the lationship ideals were defined by two factors that paralleled the
results were analyzed further, the within-subject correlations were first two partner ideals: Relationship Intimacy-Loyalty (with items
converted to Fisher's z scores to produce more normal distribu- such as caring, respect, honest, trusting, and support) and Re-
tions (the original distributions were somewhat negatively lationship Passion (with items such as exciting, fun, and
skewed).
independent).
As shown in Table 7, as expected, both consistency measures As anticipated, there was considerable empirical overlap among
were correlated significantly with reports of relationship quality. the five ideal scales. Individuals who rated relationship intimacy-
Participants who evaluated their relationships more highly ranked loyalty more highly also rated as more important an ideal mate's
their ideals in a direction that was more consistent with their personal characteristics that would be conducive to attaining
perceptions of their current partner or relationship. In addition, greater intimacy and loyalty in relationships. In addition, those
participants who evaluated their relationships more positively who considered relationship passion more important valued more
tended to rate the same items that were described as accurately highly an attractive and robust ideal mate. In contrast, importance
describing their current relationship as also more difficult to find in ratings of ideal partner status-resources were less strongly corre-
an alternative partner or relationship. Both consistency measures lated with the other ideal dimensions. Using CFA, Study 3 repli-
were also positively correlated (r = .25, p < .01). cated this pattern of overlapping factors. Moreover, this analysis
When relationship quality was regressed on both consistency suggested that the best-fitting model was one in which the five
measures, the actual-ideal consistency measure maintained a mod- relationship and partner ideals loaded on two overarching factors
erately strong and significant relation with relationship quality, but we labeled as Warmth-Loyalty and Vitality-Status-Passion.
the actual-alternative consistency measure produced a nonsignif-
Study 4 provided validation evidence for the scales. In accord
icant regression coefficient (see Table 7). Moreover, when the
with convergent validity predictions, individuals who had a more
same analyses were calculated separately for men (n = 26) and
unrestricted sociosexual orientation rated the partner warmth-
women (n = 63), the results were very similar in both analyses,
trustworthiness and relationship intimacy-loyalty ideals as less
with positive significant regression coefficients (p < .05) for the
important than did more restricted individuals. The pattern of
actual-ideal consistency measure (men: .46; women: .35) and
correlations between the Relationship Beliefs Scale and the ideal
nonsignificant regression coefficients for the actual-alternative
scales also generally confirmed predictions. For example, individ-
consistency measure (men: .33; women: .10). In summary, as
uals who believed that intimacy is an important cause of rela-
predicted, these results show that the more consistent an individ-
tionship success rated the partner warmth-trustworthiness and
ual's ideals are with perceptions of his or her current partner or
relationship intimacy-loyalty ideals more highly. Confirming dis-
relationship, the happier the individual is with the relationship.
criminant validity predictions, of the 29 possible correlations we
predicted to be nonsignificant, only 2 were significant. Study 5
General Discussion produced additional convergent validity support, confirming our
expectations that the ideal ratings would be positively and signif-
For the most part, the results of this research were consistent icantly associated with perceptions of the actual (current) partner
with our predictions. Studies 1 and 2 revealed that items college and relationship on the same dimensions.
students claimed were important attributes of their ideal romantic
partners and relationships formed five factors. The ideal part- As expected, Study 6 revealed that the higher the consistency
between ideals and perceptions of the current partner or relation-
ship, the more positively individuals viewed their current partners
or relationships. This finding remained significant after controlling
Table 7 for perceptions of how difficult it would be to obtain a comparable
Zero-Order Correlations and Regression Coefficients From alternative partner or relationship, and the pattern of findings was
Simultaneous Multiple Regressions With Relationship the same for both men and women.
Quality as the Dependent Variable
Standardized Social Cognition in Relationships

Predictor variable r regression coefficient
Our results have important implications for the study of social
Actual-ideal consistency 40**** 37****
cognition in relationships. Overall, the findings suggest that part-
Actual-alternative consistency .23* .17
ner and relationship ideals are stable cognitive constructs that vary
Note. N = 89 (Study 6). across individuals in terms of their chronic accessibility or cen-
*p<.05. ****/>< .0005. trality. Moreover, the results clarify the cognitive structure of
ideals, suggesting that the ideals load onto two overarching con- reported by Murray et al. was that partners who inaccurately
structs that span the partner and relationship: warmth-loyalty and idealized each other and their relationship appeared to construct
vitality-status-passion. relationships that, in reality, increasingly mirrored their illusions.
In terms of the model shown in Figure 1, our results suggest As Murray et al. (1996) concluded, "Love is not blind, but pre-
that, at the most general level, the two major sets of partner and scient" (p. 1155).
relationship ideals (warmth-loyalty and vitality-status-passion) In general, we believe that the application of social-cognitive
are located in the central, overlapping portion of the self, the theories to the study of ideals can provide valuable extensions to
partner, and the relationship. Most people apparently believe (quite some of the key concepts underlying interdependence theory (see
reasonably) that the amount of warmth and loyalty likely to be Thibaut & Kelley, 1959). However, in our view, one limitation of
attained in a hypothetical relationship will be partially dependent many contemporary approaches to relationship cognition is that
on how warm and loyal the partner is, and the same appears to be they do not sufficiently take into account the interaction between
true with respect to vitality and passion. the general knowledge-structure level and the specific relationship
In terms of content, the ideals dimensions are consistent with level. Hence, such approaches cannot adequately explain how
some major typologies of love. In particular, the warmth-loyalty global standards (such as ideals) interact with perceptions and
and vitality-status-passion dimensions are similar to the compan- judgments of specific individual partners and relationships.
ionate and passionate love dimensions proposed by Hatfield and For example, self-expansion theory (Aron & Aron, 1996), which
Walster (1978). They also capture the intimacy, passion, and focuses primarily on the relationship level, specifies the interaction
commitment categories that anchor Steinberg's (1986) triangular between the concepts of self and other as they exist within the
theory of love (see also Aron & Westbay, 1996). Prior research on context of specific relationships. Attachment theory (Shaver et al.,
love, however, has not been concerned with ideal standards. 1996), in contrast, focuses more on the general knowledge level,
Rather, it has focused on judgments of specific relationships or on proposing that working models and attachment styles (which are
exploring lay concepts of love in terms of dimensions or proto- developed across the life span) influence the development of
types. Ideals also differ from the concept of working models in specific adult relationships in a similar way. Avoidant styles, for
attachment theory, which represent broad sets of behavioral and example, are presumed to operate like personality traits, being
affective expectations (not ideals) about potential partners and expressed across different intimate relationships in the same way.
relationships (Shaver, Collins, & Clark, 1996).9 Finally, attribution-based theories focus mainly on the local level,
In our view, ideals are cognitive constructs that differ in crucial usually tracing the links between attributional patterns for behav-
ways from constructs such as prototypes or working models. Ideals iors in specific relationships and relationship satisfaction (see
represent general standards that can predate specific relationships Fletcher & Fincham, 1991).
and are usually located at the positive periphery of evaluative
dimensions rather than at the average or mode (Barsolou, 1985). Evolutionary Approaches
Therefore, ideals tend to have a different meaning relative to Our results also support and extend previous research and the-
concepts of love, realistic expectations, or judgments of specific orizing from an evolutionary perspective. The emergence of the
relationships or partners. Accordingly, ideals ought to serve spe- three ideal partner dimensions (warmth-trustworthiness, vitality-
cific and pivotal functions in relationship settings, some of which attractiveness, and status-resources) makes eminent sense theoret-
have been documented in the present research. ically when one considers what kinds of attributes should have
The primary function of ideals, in our view, is their use as been most relevant to successful human reproduction in ancestral
evaluative standards or goals against which relationships and part- environments. To reproduce successfully, individuals must accom-
ners are judged. Although the data are correlational, the present plish three major tasks: (a) find a cooperative, committed partner
study found evidence for this proposition. However, at present, it with whom a good parenting relationship can be formed; (b)
is not known whether ideal standards, perceptions of relationships, identify a partner who is healthy and shows evidence of high
and relationship evaluations are equally flexible over time. If viability, attributes that could be passed on to subsequent children;
ideals remain fixed at prerelationship levels, for example, they and (c) find a partner who has or can accrue the status and
might, be the central force responsible for shifting relationship resources necessary to sustain and support a family (see Buss &
evaluations as relationships unfold. Alternately, if ideals are mal- Schmitt, 1993; Gangestad & Simpson, 1996).
leable and can be easily altered to match relationship perceptions It is especially noteworthy that this tripartite ideal structure for
and evaluations, they might exert little influence. Ideals could be the ideal partner was found in items that were inductively derived
the cognitive tail wagged by the relationship head. from free-response protocols. To our knowledge, these results
To investigate these possibilities, researchers must use experi- constitute the first evidence that ideal categories are rooted in
mental designs or study relationships as they develop over time. language and folk psychological theory (at least in Western cul-
Murray et al. (1996) tracked dating couples in stable dating rela- ture). These results are also consistent with contemporary versions
tionships over a year and found that people's conceptions of their of evolutionary theory that propose that evolutionary processes
ideal partners tended to change in response to the perceived
qualities of their current partners rather than vice versa. However,
these authors speculated that ideals should exert their maximum 9
In a preliminary study (not reported previously), we assessed the
impact at or near the beginning of relationships. To disentangle the correlations between attachment style ratings for security, avoidance, and
causal links that exist between ideals and relationship or partner anxious ambivalence (using a scale developed by Simpson, 1990) and
evaluations, it will be necessary to study developing relationships ratings from the five ideal scales (n = 45). Results revealed a pattern of
from their earliest stages. Perhaps the most intriguing finding small and nonsignificant correlations ranging from — .20 to .22.
often influence human behavior indirectly, via the way in which the proximal causes and processes that generate relationship judg-
evolution has molded the human mind (Pinker, 1997). If an evo- ments and behavior; evolutionary approaches, on the other hand,
lutionary perspective is correct, the same tripartite ideal structure address some of the evolved origins of these knowledge structures.
should exist in all human cultures (see Buss & Schmitt, 1993). However, blending both approaches does more than merely offer
Even though most people may ideally desire partners who score a more comprehensive causal account. In the present research,
high on all three ideal partner dimensions, most individuals must evolutionary theory directed our attention to—and helped us to
make trade-offs between these attributes when deciding whom to explain—the three dimensions underlying partner-based ideals.
date or marry (Gangestad, 1993; Gangestad & Simpson, 1996). Understanding the evolutionary origins of relationship belief sys-
Our findings are consistent with previous research indicating that tems can enhance knowledge of their nature and functions.
some people (e.g., those with a restricted sociosexual orientation)
are, in fact, willing to sacrifice attractiveness and social status for
Caveats and Future Directions
greater commitment and loyalty in a romantic partner (Simpson &
Gangestad, 1992). In effect, each ideal dimension reflects a pos- Our results may not necessarily generalize across different kinds
sible "solution" to a specific barrier to successful reproduction. of relationships or relationship contexts. Indeed, the accessibility
Different people favor different solutions and, hence, differentially and influence of relationship and partner ideals.may vary consid-
weight each ideal dimension. erably depending on the social context in which judgments are
We anticipated that attractiveness and indicators of health would made. For example, we found that the consistency between peo-
define one of the ideal partner dimensions. This dimension did ple's ratings of their ideal partner's social status-resources and
emerge, containing items reflecting the degree to which an ideal their current partner's standing on this dimension did not signifi-
partner is adventurous, outgoing, and attractive (resulting in the cantly predict evaluations of their current partner or relationship
label of partner vitality-attractiveness). From an evolutionary per- (see Footnote 7). Most of the participants in this research, how-
spective, physical fitness, energy, and vigor are good indicators of ever, were involved in stable and long-term relationships (although
general health and physical well-being (see Buss, 1989). a sizable minority were involved in newly formed dating relation-
The general absence of sex differences in the importance as- ships). As mentioned earlier, within stable relationships, the qual-
signed to different ideals corroborates previous research showing ity and nature of relationship intimacy are likely to be of greater
that sex differences in ratings of preferred mates are considerably concern.
weaker for long-term than short-term relationships (Kenrick, In contrast, status-resource ideals (e.g., age and financial re-
Sadalla, Groth, & Trost, 1990). Our results thus make sense given sources) may exert their peak influence at the beginning of rela-
that most of our participants were involved in long-term, stable tionships, or they may operate as "screening variables" that help
relationships. The only significant sex difference we found re- individuals decide whether a given relationship is worth initiating.
vealed that women rated partner warmth-trustworthiness and re- As relationships move past the earliest stages, passion ideals might
lationship intimacy-loyalty as slightly less important than did men. become more salient during particular relationship phases, such as
This difference may partly reflect the widespread belief (held by when passion fades and boredom casts its pall (see Sternberg,
both sexes) that women tend to be more focused on intimacy and 1986). In short, the developmental stages through which relation-
commitment in relationships than men; accordingly, associated ships pass (and also simply chronological age) should be system-
ideals may be calibrated differently in men and women. Specifi- atically related to when different partner and relationship ideals
cally, if women believe that most men want or expect less com- become salient and influential in relationships.
mitment and intimacy (than women do), women may lower their Contextual factors and the availability of alternative partners
ideal standards on this dimension (relative to men) to give them- also ought to affect the accessibility of ideals. For example, be-
selves a more realistic chance of eventually finding a mate. coming good friends with another couple or reading a book on how
It is interesting to speculate on how short-term versus long-term to have a good marriage may strengthen the accessibility of inti-
mating strategies might be associated with ideal standards. Indi- macy ideals (leading to either increased or decreased satisfaction
viduals who adopt short-term mating strategies should be less with one's current relationship). Conversely, joining a gym and
concerned than individuals who adopt long-term strategies if their constantly observing hordes of nubile bodies might strengthen the
current partner-relationship falls short of their ideal standards for importance of the vitality-attractiveness ideal or heighten aware-
partner warmth-trustworthiness and relationship intimacy-loyalty. ness of the availability of attractive and healthy alternative part-
However, they should be more concerned than long-term individ- ners. Such changes in relationship knowledge structures may, in
uals if their current partner or relationship does not meet their ideal turn, either enhance or undermine evaluations of the current part-
standards for partner vitality-attractiveness and perhaps relation- ner and relationship.
ship passion (cf. Simpson & Gangestad, 1992). In contrast, be- Another important question involves what constitutes an intol-
cause intimacy is more important to people who have long- erable discrepancy between ideals and perceptions of the current
term orientations, lower consistency on the partner warmth- partner or relationship. Most people probably accept that their
trustworthiness and relationship intimacy-loyalty dimensions ideals are unlikely to ever be fully met. However, what constitutes
should have stronger and more deleterious effects on their rela- an "acceptable" disparity and why some people may tolerate
tionships than on the relationships of those who adopt short-term greater disparities than others are intriguing questions.
mating strategies. Finally, even though the "self is included in our original model
We believe that the present research illustrates the fruitfulness (see Figure 1), we did not devote much attention to the self-
of jointly applying social-cognitive and evolutionary approaches concept in our research. Instead, we focused on the partner and the
to relationship phenomena. Social-cognitive approaches focus on relationship, which should be the natural targets of relationship-
FLETCHER, SIMPSON, THOMAS, AND GILES
relevant ideals. Indeed, we do not think it likely that a knowledge with evolutionary principles can improve understanding of both
construct that one might term "ideal self in close relationships" the origins of relationship knowledge structures and how they are
exists as an entity that is separate from the ideal self. A central causally related to cognition and behavior in relationships.
component of the ideal (social) self already contains many items
that implicitly refer to close relationship contexts such as affec- References
tionate, warm, honest, sexy, romantic, kind, and so forth. How- Alexander, R. A., & DeShon, R. P. (1994). Effect of error variance
ever, the way in which self-perceptions are related to ideal stan- heterogeneity on the power of tests for regression slope differences.
dards is an important question for further research and theory. For Psychological Bulletin, 115, 308-314.
example, it is plausible that individuals' ideal standards may be Aron, A., & Aron, A. (1996). Self and self-expansion in relationships. In
influenced by their self-perceptions on the same dimension (see G. J. O. Fletcher & J. Fitness (Eds.), Knowledge structures in close
Kenrick, Groth, Trost, & Sadalla, 1993). For instance, if the self is relationships: A social psychological approach (pp. 325-344). Mahwah,
rated as relatively weak in terms of vitality and attractiveness, this NJ: Erlbaum.
may downgrade the importance given to vitality and passion in the Aron, A., & Westbay, L. (1996). Dimensions of the prototype of love.
ideal partner and relationship. Journal of Personality and Social Psychology, 70, 535-551.
Baldwin, M. W. (1992). Relational schemas and the processing of social

Another important direction for future research and theorizing
information. Psychological Bulletin, 112, 461—484.

concerns the similarities and differences between self-discrepancy Barsalou, L. W. (1985). Ideals, central tendency, and frequency of instan-
and relationship-discrepancy models in relation to ideals (see tiation as determinants of graded structure in categories. Journal of
Klohnen & Mendelsohn, 1998). Of course, there are likely to be Experimental Psychology: Learning, Memory, and Cognition, 11, 629—
strong similarities in terms of the underlying psychological model; 654.
indeed, we drew some of our thinking and predictions from Hig- Bentler, P. M. (1993). EQS structural equations program manual. Los
gins's (1987, 1989) self-discrepancy model. On the other hand, Angeles: BMDP Statistical Software.
there are probably important differences between the two catego- Bentler, P. M., & Wu, E. J. C. (1995). EQS for Windows user's guide.
ries. For example, if the gap between the partner-relationship ideal Encino, CA: Multivariate Software.
and the perception of reality is too great, this can be resolved by Berscheid, E. (1994). Interpersonal relationships. Annual Review of Psy-
chology, 45, 79-129.
the individual leaving the relationship, an outcome not available to
Bryne, B. M. (1994). Structural equation modeling with EQS and EQS/
the self (unless one counts suicide as leaving the self). Another
Windows. Thousand Oaks, CA: Sage.
example is that the regulatory function of partner-relationship
Buss, D. M. (1989). Sex differences in human mate preferences: Evolu-
ideals—such as changing the partner's behavior to produce a more tionary hypotheses tested in 37 cultures. Behavioral and Brain Sci-
consistent fit between ideal and perception of reality—might gen- ences, 12, 1-49.
erally be more difficult than for the self (i.e., changing self- Buss, D?M., & Schmitt, D. P. (1993). Sexual strategies theory: A contex-
behavior). This may mean that, over time in relationships, ideals tual evolutionary analysis of human mating. Psychological Review, 100,
will bend to fit the perception of the partner or relationship rather 204-232.
than the other way around (see Murray et al, 1996). Chou, C. P., Bentler, P. M., & Satorra, A. (1991). Scaled test statistics and
robust standard errors of nonnormal data in covariance structure analy-
Finally, one possible drawback in terms of other researchers
sis: A Monte Carlo Study. British Journal of Mathematical and Statis-
using these scales is the large number of items used and the tical Psychology, 44, 347-357.
substantial variation in the number of items adopted across the Crowne, D. P., & Marlowe, P. (1964). The approval motive. New York:
subscales. To derive more user-friendly measures, we selected the Wiley.
six best-loading items from Study 1 from each of the original Fehr, B. (1998). Prototype analysis of the concepts of love and commit-
scales for the five ideal categories. We then repeated all of the ment. Journal of Personality and Social Psychology, 55, 557-579.
analyses reported in the first five studies and attained results Fehr, B., & Russell, J. A. (1991). The concept of love viewed from a
comparable to those already reported. In particular, the five short prototype perspective. Journal of Personality and Social Psychol-
scales attained adequate internal reliability coefficients in Study 2 ogy, 60, 425-438.
(ranging from .75 to .91), and the CFAs with the short scales were Fletcher, G. J. O., & Fincham, F. D. (1991). Attribution processes in close
very similar to those reported in Study 3. Hence, we are confident relationships. In G. J. O. Fletcher & F. D. Fincham (Eds.), Cognition in
close relationships (pp. 7-36). Hillsdale, NJ: Erlbaum.
that the short versions of the scales will perform adequately in
Fletcher, G. J. O., & Fitness, J. (1990). Occurrent social cognition in close
further research (see the Appendix for a list of the items).
relationship interaction: The role of proximal and distal variables. Jour-
nal of Personality and Social Psychology, 59, 464-474.
Conclusion Fletcher, G. J. O., & Fitness, J. (Eds.). (1996). Knowledge structures in
close relationships: A social psychological approach. Mahwah, NJ:
This research offers some provisional answers to fundamental Erlbaum.
questions concerning the structure, content, and functions of part- Fletcher, G. J. O., Fitness, J., & Blampied, N. M. (1990). The link between
ner and relationship ideals, and it has produced reliable and valid attributions and happiness in close relationships: The roles of depression
and explanatory style. Journal of Social and Clinical Psychology, 9,
scales that measure these ideals. It is critical to recognize that
243-255.
evolutionary approaches do not discount the influence of culture or
Fletcher, G. J. O., & Kininmonth, L. (1992). Measuring relationship
learning on cognition and behavior in relationships (see Simpson beliefs: An individual differences scale. Journal of Research in Person-
& Kenrick, 1997). Indeed, evolutionary forces are responsible for ality, 26, 371-397.
producing the human mind, which possesses remarkable cognitive Fletcher, G. J. O., Rosanowski, J., & Fitness, J. (1994). Automatic pro-
flexibility and a tremendous capacity for learning (Pinker, 1997; cessing in intimate contexts: The role of relationship beliefs. Journal of
Wright, 1994). Research and theory that integrate social cognition Personality and Social Psychology, 67, 888-897.
IDEALS IN INTIMATE RELATIONSHIPS
Fletcher, G. J. O., & Thomas, G. (1996). Lay theories in close relation- models and their invariance across groups. Psychological Bulletin, 97,
ships: Their structure and function. In G. J. O. Fletcher & J. Fitness 562-582.
(Eds.), Knowledge structures in close relationships: A social psycholog- Murray, S. L., Holmes, J. G., & Griffin, D. W. (1996). The self-fulfilling
ical approach (pp. 3—24). Mahwah, NJ: Erlbaum. nature of positive illusions in romantic relationships: Love is not blind
Gangestad, S. W. (1993). Sexual strategies and physical attraction: Impli- but prescient. Journal of Personality and Social Psychology, 71, 1155—
cations for mating dynamics. Human Nature, 4, 205-235. 1180.
Gangestad, S. W., & Simpson, J. A. (1996). On the evolutionary psychol- Pinker, S. (1997). How the mind works. New York: Norton.
ogy of human mating: Trade-offs and strategic pluralism. Unpublished Rusbult, C. E., Onizuka, R. K., & Lipkus, I. (1993). What do we really
manuscript, University of New Mexico, Albuquerque. * want? Mental models of ideal romantic involvement explored through
Hasserbrauck, M. (1997). Cognitions of relationship quality: A prototype multidimensional scaling. Journal of Experimental Social Psychol-
analysis of their structure and consequences. Personal Relationships, 4, ogy, 29, 493-527.
163-185. Shaver, P. R., Collins, N., & Clark, C. L. (1996). Attachment styles and
Hatfield, E., & Walster, G. W. (1978). A new look at love. Lanham, MD: internal working models of self and relationship partners. In G. J. O.
University Press of America. Fletcher & J. Fitness (Eds), Knowledge structures in close relationships:
Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect. A social psychological approach (pp. 25—61). Mahwah, NJ: Erlbaum.
Psychological Review, 94, 319-340.

Simpson, J. A. (1990). Influence of attachment styles on romantic rela-
Higgins, E. T. (1989). Self-discrepancy theory: What patterns of self-

tionships. Journal of Personality and Social Psychology, 59, 971-980.
beliefs' cause people to suffer? In L. Berkowitz (Ed.), Advances in
Simpson, J. A., & Gangestad, S. W. (1991). Individual differences in
experimental social psychology (Vol. 22, pp. 93-136). New York:
sociosexuality: Evidence for convergent and discriminant validity. Jour-
Academic Press.
nal of Personality and Social Psychology, 60, 870-883.
Johnson, D. J., & Rusbult, C. E. (1989). Resisting temptation: Devaluation
Simpson, J. A., & Gangestad, S. W. (1992). Sociosexuality and romantic
of alternative partners as a means of maintaining commitment in close
relationships. Journal of Personality and Social Psychology, 57, 967- partner choice. Journal of Personality, 60, 31-51.
980. Simpson, J. A., & Kenrick, D. T. (Eds.). (1997). Evolutionary social
Kenrick, D. T., Groth, G. E., Trost, M. R., & Sadalla, E. K. (1993). psychology. Mahwah, NJ: Erlbaum.
Integrating evolutionary and social exchange perspectives on relation- Sternberg, R. J. (1986). A triangular theory of love. Psychological Re-
ships: Effects of sex, self-appraisal, and involvement level on mate view, 93, 119-135.
selection criteria. Journal of Personality and Social Psychology, 64, Sternberg, R. J., & Barnes, M. L. (1985). Real and ideal others in romantic
951-969. relationships: Is four a crowd? Journal of Personality and Social Psy-
Kenrick, D. T., Sadalla, E. K., Groth, G., & Trost, M. R. (1990). Evolution, chology, 49, 1586-1608.
traits, and the stages of human courtship: Qualifying the parental invest- Thibaut, J. W., & Kelley, H. H. (1959). The social psychology of groups.
ment model. Journal of Personality, 58, 97-116. New York: Wiley.
Klohnen, E. C , & Mendelsohn, G. A. (1998). Partner selection for per- Tooby, Jif & Cosmides, L. (1992). The psychological foundations of
sonality characteristics: A couple-centered approach. Personality and culture. In J. H. Barkow, L. Cosmides, & J. Tooby (Eds.), The adapted
Social Psychology Bulletin, 24, 268-278. mind: Evolutionary psychology and the generation of culture (pp. 19—
Marsh, H. W., & Hocevar, D. (1985). Application of confirmatory factor 136). New York: Oxford University Press.
analysis to the study of self-concept: First- and higher order factor Wright, R. (1994). The moral animal. New York: Vintage Books.
Appendix
Items in the Short Version of the Partner and Relationship Ideal Scales
Partner Warmth-Trustworthiness: understanding, supportive, considerate, kind, a good listener, sensitive

Partner Vitality-Attractiveness: adventurous, nice body, outgoing, sexy, attractive, good lover
Partner Status-Resources: good job, financially secure, nice house or apartment, appropriate ethnicity, success-
ful, dresses well
Relationship Intimacy-Loyalty: honest, commitment, caring, trusting, support, respect
Relationship Passion: exciting, challenging, humorous, fun, independence, passionate
Received January 29, 1997

Revision received December 15, 1997
Accepted August 6, 1998

Ideals in Intimate Relationships

Uploaded by

Copyright:

Available Formats

You might also like

Ideals in Intimate Relationships

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ideals in Intimate Relationships

Uploaded by

Copyright:

Available Formats

Journal of Personality and Social Psychology Copyright 1999 by the American Psychological Association, Inc.

1999, Vol. 76, No. 1, 72-89 0022-3514/99/S3.00

Ideals in Intimate Relationships

Garth J. O. Fletcher Jeffry A. Simpson

Geoff Thomas and Louise Giles

partner were represented by 3 factors (partner warmth-trustworthiness, vitality-attractiveness, and

studies, Rusbult et al. (1993) used a combination of free-response

ined other kinds of relationship-level knowledge structures, includ-

Rate of Importance Factor 1 Factor 2 Factor 3

Understanding 42 5.99 .87 -.18 .04

Friendly 18 5.87 .67 .10 -.05

Communicative 48 6.08 .62 .14 -.13

Rate of mention Importance rating Factor 1 Factor 2

Honest 48 6.39 .80 -.06

Monogamous 14 6.07 .50 -.11

In love 40 5.90 .50 -.04

Partner Warmth-Trustworthiness .93 .75 5.97 0.63 5.66 0.76

1. Partner Warmth-Trustworthiness .49* .10 .80*** 49***

Note. N = 320 (Study 2).

including good internal consistency and good test-retest reliability. Results

Moreover, our expectation that partner and relationship ideals

' Relations Relationshi

[at seales, order and second- erl atent factors. Al loadings a

higher order and lower order models is the target coefficient

Partner ideal Relationship ideal

Warmth- Vitality- Status- Intimacy-

Social desirability .13 .00 .21 .07 -.11

tance being placed on passion in producing successful relation-

ships. The occurrence of 2 significant correlations is approxi-

Partner ideal Relationship ideal

Warmth- Vitality- Status- Intimacy- Relationship

Partner warmth-trustworthiness .36* .37* .10 .23* .45* .72*

Standardized Social Cognition in Relationships

Baldwin, M. W. (1992). Relational schemas and the processing of social

information. Psychological Bulletin, 112, 461—484.

Psychological Review, 94, 319-340.

Higgins, E. T. (1989). Self-discrepancy theory: What patterns of self-

Partner Warmth-Trustworthiness: understanding, supportive, considerate, kind, a good listener, sensitive

Received January 29, 1997

You might also like

1. Partner Warmth-Trustworthiness .49* .10 .80* 49*