SELF-KNOWLEDGE IN EVERYDAY LIFE
Does what we think of ourselves matter?
Self-concept and self-esteem in everyday life
William B. Swann, Jr. Christine Chang-Schneider Katie Larsen McClarty
University of Texas at Austin
Running Head: SELF-KNOWLEDGE IN EVERYDAY LIFE
1
SELF-KNOWLEDGE IN EVERYDAY LIFE
2
Abstract
We suggest that recent critics of self-esteem research have dismissed it prematurely.
These critics have evaluated the effect sizes of past research using a criterion that is
unrealistic. They have also conflated high self-esteem with narcissism, leading them to
fail to recognize the link between low self-esteem and anti-social behavior. Furthermore,
they have focused their literature reviews too narrowly, examining only the effects of
global self-views (i.e., self-esteem or feelings of self-worth) and neglecting the effects of
specific self-views (self-concepts or beliefs about the self) that contextualize these global
self-views. So constrained, they have related self-esteem to specific outcomes that are
much narrower in conceptual breadth than self-esteem, resulting in underestimates of
predictive validity. When predictor and criterion variables are conceptually equivalent
(global predictors paired with global outcomes and specific predictors paired with
specific outcomes), substantial relationships between self-views and outcomes emerge.
We conclude that self-views do matter and it is desirable to implement theory based
programs to improve them.
Keywords: Self-esteem change, self-concept, self-efficacy, self-verification.
SELF-KNOWLEDGE IN EVERYDAY LIFE
3
Does what we think of ourselves matter?
Self-concept and self-esteem in everyday life
For most of the last century, theory and research on the self-concept and selfesteem were consigned to the backwaters of psychology. Then, in the late 1970s, articles
by Kuiper and Rogers (1979), Markus (1977), and others demonstrated that selfknowledge had properties similar to schemas and beliefs--constructs that had been
championed by cognitive psychologists. In so doing, these researchers legitimized the
self-concept as a viable scientific construct. The result was a marked increase in research
on the self during the 1980s (Swann & Seyle, 2005). At the same time, an independent
wave of enthusiasm within the lay community launched self-esteem into the national
limelight. On the basis of precious little evidence, California’s Task Force to Promote
Self-esteem and Personal and State Responsibility (State of California, 1996)
characterized self-esteem as a panacea whose cultivation would protect people from a
host of ills, including welfare dependency, teenage pregnancy, juvenile delinquency,
dropping out of high school, and so on. Thousands of laypersons across America were
smitten with the hope that in self-esteem they had found a modern day Holy Grail.
No longer. With ample justification, members of the academic community pointed
out that the extravagant claims of the self-esteem movement were nothing more than that
(e.g., Dawes, 1996; Swann, 1996). Yet, in very recent years the pendulum has swung
even further, both reflecting—and inspiring—deep doubts about the viability of work on
the self in general and self-esteem in particular. Several authors (Baumeister, Smart, &
Boden, 1996; Sedikides, Rudich, Gregg, Kumashiro, & Rusbult, 2004) have questioned
SELF-KNOWLEDGE IN EVERYDAY LIFE
4
one of the subareas’s bedrock assumptions regarding the nature of self-esteem: namely,
that narcissism, a construct that has historically been regarded as distinct from selfesteem, is associated with or even a form of self-esteem. Even more provocatively, on
the basis of a qualitative survey of the literature, the authors of a highly visible review
paper (Baumeister, Campbell, Krueger, & Vohs, 2003) asserted that the effect sizes of
self-esteem on socially important outcome variables were small and inconsequential.
They went on to conclude that because self-esteem is largely inconsequential, programs
designed to improve self-esteem are pointless and unwarranted. Baumeister et al.’s
argument has won some support. For example, some authors have endorsed such gloomy
assessments of the predictive validity of self-esteem and gone even further to conclude
that the evidentiary basis of self-esteem research is so fundamentally flawed that the
entire enterprise should be re-examined (Scheff & Fearon, 2004).
We address each of these contentions in this article. We begin by arguing that
recent critics of the self-esteem literature have assessed the predictive validity of selfesteem research using unrealistically high standards. This is particularly true when one
recognizes the social importance of the outcomes that self-esteem has been used to
predict. We then turn to a discussion of the nature of self-esteem, arguing that there are
strong conceptual and empirical reasons for distinguishing it from narcissism.
Furthermore, we note that making this distinction removes recently introduced ambiguity
regarding the capacity of self-esteem measures to predict aggressive and anti-social
behavior.
We suggest that a key problem with recent criticisms of self-esteem research is that
critics have allowed the California task force to define the terms of the debate. This has
SELF-KNOWLEDGE IN EVERYDAY LIFE
5
led them to restrict the focus of their analyses to the effects of self-esteem while ignoring
the effects of self-concepts. This is akin to examining the impact of global evaluations of
someone’s worth while ignoring the impact of the many specific impressions (e.g.,
intelligent, artistic, athletic, interesting) that contextualize and support these global
evaluations. We accordingly take a more expansive approach to the literature on the
effects of self-knowledge, one that simultaneously considers the effect of a broader class
of self-representations dubbed “self-views” (i e., self-esteem and self-concepts). We
contend that measures of self-views should be related to socially important outcome
variables only insofar as the predictor and outcome variables are conceptually equivalent.
In fact, when the conceptual equivalence of self-views and outcomes is high, we show
that self-concepts and self-esteem predict socially important outcomes at levels that
exceed even the unrealistically high expectations of recent critics. Given that self-views
predict outcomes that are important to our society, we conclude that it is appropriate to
devise and implement systematic, theory guided efforts to improve self-views.
Unrealistic expectancies
Baumeister et al. (2003) never explicitly define their standard for an effect size that
they would find impressive. Nevertheless, several of their remarks make it clear that
their standards are quite high. For example, at one point they refer to correlations
between .40 and .60 as “only moderate” (Baumeister et al., p. 6). Later, they note that
correlations between 0 and .30 are “quite weak,” presumably because a .30 correlation
indicates that the predictor accounts for less than 10% of the variance in the outcome
variable (Baumeister et al., p.24).
SELF-KNOWLEDGE IN EVERYDAY LIFE
6
Although there is surely some precedent for such lofty standards in the
psychological literature (e.g., Mischel, 1968; Wicker, 1969), during the last decade or so
researchers have begun to question the appropriateness of the widespread application of
such standards. One group of researchers has compared the predictive validity of
psychological tests, which have been criticized on the grounds that they have low effect
sizes, with the predictive validity of highly regarded medical tests (Meyer et al., 2001).
They concluded that the psychological tests were just as predictive of outcomes as the
medical tests were. Similarly, other work has suggested that the relatively modest effect
sizes of both medical and psychological tests is expected and reasonable if one
recognizes that medical outcomes, like psychological outcomes, are multiply determined.
That is, Monte Carlo simulations conducted by Ahadi and Diener (1989) have shown that
as the number of variables that cause a behavior increase, the maximum possible
correlation between any single predictor and outcome drops precipitously. From this
perspective, the relationships that Baumeister et al. characterized as “very weak” (.30)
actually approach the upper limit of what might be expected when the outcomes are
multiply determined (see also Strube, 1991).
Finally, several thoughtful theorists and researchers have argued that even small
effect sizes are noteworthy if they have clear implications for a significant theoretical or
practical issue (Abelson, 1995; Rosenthal, 1994). This position is highly relevant to
research on self-esteem, as the social importance of the outcome variables with which it
has been associated (e.g., academic performance, marital satisfaction, criminal
convictions, depression, and so on) is quite high. Consider, for example, Rosenthal’s
(1994) discussion of the physician’s aspirin study, which fomented a dramatic revision in
SELF-KNOWLEDGE IN EVERYDAY LIFE
7
the recommended use of aspirin to forestall cardiac events. Rosenthal notes that the
effect size of aspirin was “tiny”—r = .034 (r2 = .0012)—an effect size slightly smaller
than one that Baumeister et al. (2003) derided as “negligible” (p. 8). Yet this effect size
translates roughly to a 4% decrease in heart attacks, which cardiologists agree is
impressive enough to recommend aspirin regimes to all patients who are considered
cardiac risks.
In short, we believe that Baumeister et al. (2003) set the bar much too high when
evaluating the effect sizes of the research they reviewed. The problems with their review,
however, go much deeper than this, including even the manner in which they
conceptualize self-esteem.
The nature of self-knowledge
In this article we use the term “self-view” as a super-ordinate category that
incorporates self-concepts (beliefs about the self) as well as self-esteem (feelings of selfworth). The most important distinction between self-concepts and self-esteem is the
relative specificity of the two terms: whereas self-concepts refer to particular qualities
that people ascribe to themselves (e.g., sociability, dominance, and creativity), selfesteem is relatively global, referring to people’s overall sense of worth (sometimes
differentiated into self-perceived lovability and competence, see Tafarodi & Swann,
2001). Some have also argued that self-concepts are “cognitive” (i.e., what people think
about themselves) and self-esteem is “affective” (i.e., what people feel about themselves).
Although this distinction has some merit, for practical purposes it is often misleading
because both constructs have cognitive as well as affective elements. Most strikingly,
many of the self-concepts that social-personality psychologists study are strongly
SELF-KNOWLEDGE IN EVERYDAY LIFE
8
affectively charged. People often care a great deal, for example, about their belief that
they are intelligent, athletic, or dominant. Even social self-concepts (self-concepts that
align people to groups, such as Christian, American, or Libertarian) can sometimes be
held quite passionately. It is therefore often inaccurate to posit that self-esteem is more
affectively charged than self-concepts (e.g., Baumeister et al., 2003).
Testimony to the importance that people attach to their self-concepts comes from
evidence of the lengths to which they will go to preserve them. So strong is the desire to
preserve self-views that people will forego opportunities to obtain positive evaluations by
choosing negative evaluations and interaction partners over positive ones (e.g., Swann,
1990). Note, however, that this evidence that people with negative self-views prefer
negative evaluations does not mean that they are consumed by an unalloyed desire for
negative feedback. To the contrary, people with negative self-views are deeply
ambivalent; in effect, they are caught in a crossfire between their desire for positive
evaluations (e.g., Jones, 1973) and their desire to maintain stable self-views (Swann,
1983).
This ambivalence also plagues people who possess a mixture of positive and
negative self-views. Narcissists represent a case in point. Superficially, narcissists may
seem like they possess high self-esteem, for they often claim to have positive qualities.
Among narcissists, however, expressions of positive self-regard represent efforts to
compensate for an underlying fragile sense of self-worth. In contrast to people with true
high self-esteem, the positive self-views of narcissists vary dramatically across situations
(Rhodewalt et al., 1998; Rhodewalt, 2005; see also Baumeister, Bushman, & Campbell,
2000). Moreover, unlike people who enjoy true high self-esteem, narcissists are prone to
SELF-KNOWLEDGE IN EVERYDAY LIFE
9
lash out at people. That is, even when the effects of self-esteem are held constant,
narcissism predicts behaviors such as aggression (Bushman & Baumeister, 1998),
negative evaluations of another after an ego threat (Morf & Rhodewalt, 1993), hostility
and disagreeableness (Rhodewalt & Morf, 1995), and reactions to success and failure
(Rhodewalt & Morf, 1998). As such, equating the positive self-evaluative statements of a
narcissist with those of a person with high self-esteem is tantamount to mistaking the
façade of a building with the building itself (e.g., American Psychiatric Association,
2000; Morf & Rhodewalt, 2001; Paulhus, Robins, Trzesniewski, & Tracy, 2004; Raskin,
Novacek, & Hogan, 1991; Tracy & Robins, 2003; Westen, 1990). It is thus arbitrary and
misleading to equate self-esteem and narcissism by suggesting that “high self-esteem is a
heterogeneous category, encompassing people who frankly accept their good qualities
along with narcissistic, defensive, and conceited individuals” (Baumeister et al., 2003, p.
1).
Distinguishing narcissism from true high self-esteem bears directly on assessing the
predictive validity of measures of self-esteem. Consider Baumeister et al.’s (2003)
assertion that narcissism is a form of high self-esteem and that it is therefore unnecessary
to distinguish the two constructs when exploring the links between self-esteem and
aggressive behavior: “People high in self-esteem or narcissism are prone to bully others,
to retaliate aggressively, and to be prejudiced against out-group members” (p. 37).
Contrary to Baumeister et al.’s position, recent research indicates that it is vital to
distinguish narcissism from high self-esteem. For example, Paulhus et al. (2004)
reported that when narcissism is held constant, low rather than high self-esteem predicts
anti-social behavior (parallel to this, when self-esteem was held constant, narcissism
SELF-KNOWLEDGE IN EVERYDAY LIFE 10
predicted anti-social behavior). Longitudinal field studies have revealed similar results.
Specifically, Donnellan, Trzesniewski, Robins, Moffitt, & Caspi (in press) have reported
that low self-esteem was associated with aggression when narcissism was partialled out.
In addition, Trzesniewski et al. (in press) showed that low self-esteem predicted antisocial behavior more than 10 years later. The results of this research thus suggest that
high self-esteem and narcissism are conceptually and empirically distinct and that low
(rather than high) self-esteem predicts aggression and anti-social behavior.
But our point is not simply that high self-esteem is distinct from narcissism or that
low self-esteem predicts aggression and anti-social behavior. Rather, our argument is a
more general one: we suggest that questions regarding the predictive validity of global
self-views such as self-esteem should be considered simultaneously with consideration of
the effects of the specific self-views (i.e., self-concepts) which can amplify, qualify, and
contextualize people’s global conceptions of themselves. This contention was inspired by
our review of the literature on the predictive utility of two closely related constructs,
attitudes and traits.
The predictive validity of attitudes and traits
Attempts to answer the question “Do self-views matter” can benefit from
knowledge gleaned from research on attitudes and traits. In what follows, we briefly
summarize what is known about the variables that influence the predictive validity of
these two constructs.
Attitudes.
Self-views can be considered a subclass of attitudes—specifically attitudes toward
the self. Early reviews of the relationship between global attitudes and specific behaviors
SELF-KNOWLEDGE IN EVERYDAY LIFE 11
concluded that the links between the two were weak and tenuous (e.g., Wicker, 1969).
More recent accounts (e.g., Ajzen & Fishbein, 2005), however, have suggested that such
early accounts were overly pessimistic.
The starting point for contemporary analyses of the attitude-behavior literature is
that (a) most of the behaviors of interest to attitude researchers are multiply determined
and (b) such multiple causation is a problem for researchers interested in using attitudes
to predict behavior, because the greater the number of rival influences on a target
behavior, the smaller the chances that a particular attitude will influence that behavior.
Fortunately, there are at least two distinct ways of dealing with the multiple causation
problem. One is to consider the degree of conceptual equivalence between the attitude
measure and criterion. If one is interested in predicting a specific behavior, then the
impact of rival influences can be minimized by selecting a relatively specific attitude
measure. In contrast, if one wants to predict general behavioral patterns aggregated
across situations, then one should use an appropriately general measure of attitudes. In
short, specific attitudes predict specific behaviors; general attitudes predict general (i.e.,
aggregated) behaviors. This specificity or conceptual equivalence principle has received
ample support in the literature (for an overview, see Ajzen & Fishbein, 2005).
The other strategy for dealing with the multiple causation problem is to measure the
factors that determine whether or not the attitude of interest will influence the behavior.
Thus, for example, based on the assumption that attitudes that are readily accessible will
be more apt to predict behavior (Fazio & Olsen, 2003), investigators can attend to the
factors that determine the accessibility of attitudes. Alternatively, one can measure other
influences on specific behavior (behavior-specific beliefs, subjective norms, perceptions
SELF-KNOWLEDGE IN EVERYDAY LIFE 12
of control, and intentions) and use them in conjunction with attitudes to predict behavior
(Ajzen & Fishbein, 1977; Ajzen, 1991). Both of these strategies for improving attitudebehavior relations could be readily adapted to bolster the predictive validity of self-views.
That is, investigators might identify variables that influence the accessibility of self-views
or rival determinants of the behavior under scrutiny (e.g., Kaplan, 1986).
Traits and personality.
Self-views can be viewed as one component of personality, specifically people’s
beliefs and feelings about their own traits and other attributes. It is thus worth noting that
the history of research on the predictive validity of traits has moved from a period of
pessimism to relative optimism (parallel to the history of the predictive validity of
attitudes). In particular, Mischel’s (1968) early, highly skeptical assessment of the
predictive validity of traits relative to situations triggered a flurry of rebuttals.
Eventually, these efforts resulted in a considerably more nuanced – and optimistic –
understanding of the relation of traits to behavior (see Swann & Seyle, 2005, for an
overview).
In the tradition of the specificity principle that emerged in the attitude-behavior
literature, theorists proposed that because traits are generalized predispositions to
respond, they should closely correspond to general indexes of behavior (assessed on
multiple occasions and different contexts) but more weakly to specific behavioral
outcomes. Consistent with this idea, when researchers aggregated outcome variables,
impressive correlations between traits and behavior emerged (e.g., Epstein, 1979). This
left open questions regarding the conditions under which traits would predict specific
behavioral outcomes. To address this issue, subsequent researchers developed several
SELF-KNOWLEDGE IN EVERYDAY LIFE 13
complementary conceptual and methodological strategies for identifying situational
factors that influence the extent to which traits predict behavior (e.g., Fleeson, 2004;
Funder & Colvin, 1991; Mischel & Shoda, 1999). In principle, it should be possible to
construct parallel formulations to improve the predictive validity of self-views.
Implications for predictive validity of self-views.
Assessments of the predictive validity of attitudes and traits reveal some striking
commonalities (see also Sherman & Fazio, 1983). In particular, unless steps are taken to
improve the strength of the relations between predictor and criterion variables,
correlations seldom exceed .30. This is the same range of correlations that Baumeister et
al. (2003) dismissed in their review of the self-esteem literature as “quite weak” (p.24).
From this vantage point, recent calls for the critical assessment of the utility of the
construct of self-esteem on the grounds of low predictive validity are reminiscent of
somber assessments of the state of research on attitudes (Wicker, 1969) and traits
(Michel, 1968) in the late 1960s. Note that these critiques led to the development of
increasingly sophisticated conceptual and empirical research paradigms which, in turn,
led to stronger findings that justified more optimistic assessments later on. If nothing
else, then, work on attitudes and traits lends perspective to recent skepticism regarding
the predictive validity of self-esteem.
More important, however, the attitude and trait literatures suggest systematic
strategies for bolstering the strength of the relationship between self-views and various
outcome variables. For example, researchers interested in the effects of self-views can
cope with the multiple causation phenomenon by attending to the factors that determine
whether or not the self-view under scrutiny determines the behavior. Thus, one can focus
SELF-KNOWLEDGE IN EVERYDAY LIFE 14
on the factors that determine the accessibility of self-views, with the assumption that
highly accessible self-views will be more apt to influence behavior than non-accessible
ones (cf. Fazio & Olsen’s, 2003 research on attitude accessibility). Alternatively, the
interactionist approach embraced by many attitude researchers (cf. Ajzen & Fishbein,
1977; Ajzen, 1991) as well as personality researchers (e.g., Fleeson, 2004; Funder &
Colvin, 1991; Higgins, 2000; Mischel & Shoda, 1999) could be adapted to research on
the self by having investigators examine the ways in which self-views interact with
situations to predict behavior. Future research should explore these possibilities.
Of more immediate relevance here is the conceptual equivalence principle, which
suggests that measures of self-views should be conceptually equivalent to measures of
outcomes. Thus, for example, researchers interested in predicting relatively specific
outcomes such as math proficiency might measure self-perceived math ability rather than
a global measure such as self-esteem. Similarly, researchers interested in using global
self-esteem as a predictor should focus on an equally global outcome measure, that is,
several outcomes aggregated together (see also Rosenberg, Schooler, Schoenbach &
Rosenberg, 1995). In the section that follows, we show that research that conforms to the
conceptual equivalence principle yields more optimistic assessment of the predictive
validity of self-views than research that does not conform to this principle.
When self-views matter
We begin by examining two lines of research on the effects of self-concepts that
followed the conceptual equivalence principle. This research was ignored by recent
reviewers (e.g., Baumeister et al., 2003; Scheff & Fearon, 2004) because it did not use
self-esteem as a predictor and was therefore beyond the scope of their reviews. In
SELF-KNOWLEDGE IN EVERYDAY LIFE 15
examining both of these literatures, we rely on meta-analyses conducted by others and
ourselves because of the many advantages of meta-analyses over qualitative reviews
(e.g., Johnson & Eagly, 2000). We then turn to some very recent research on self-esteem
that addresses some of the concerns raised by Baumeister et. al., (2003).
Self-efficacy and academic performance.
One research tradition that meets the conceptual equivalence criterion was inspired
by Bandura’s (1986; 1989) social cognitive theory. Bandura defined perceptions of selfefficacy as "people's judgments of their capabilities to organize and execute courses of
action required to attain designated types of performances" (Bandura, 1986, p. 391).
Theoretically, efficacy self-views influence the choices people make, the effort they
expend, how long they persevere in the face of challenge, and the degree of anxiety or
confidence they bring to the task at hand. Although these perceptions do not alter
people’s capabilities, they help determine what individuals do with the knowledge and
skills they have. Efficacy self-views thus help explain why performances differ among
people who have similar knowledge and skills. Consistent with the conceptual
equivalence notion, Bandura (1986) insisted that self-efficacy judgments should be
specifically rather than globally assessed, must correspond directly to the criterion task,
and must be measured as closely as possible in time to that task.
Several meta-analyses have now been conducted that have evaluated the ability of
measures of self-efficacy to predict academic outcomes (e.g., Multon, Brown, & Lent,
1991; Robbins, Lauver, Le, Davis, Langley, & Carlstrom, 2004; Valentine, DuBois, &
Cooper, 2004). In one particularly well-controlled analysis, Robbins et al. (2004)
examined 109 prospective studies in which various psychosocial and study skill factors
SELF-KNOWLEDGE IN EVERYDAY LIFE 16
were used to predict college outcomes. The predictors were categorized into nine broad
constructs, one of which was academic self-efficacy. Two college outcomes were
targeted: performance, as measured by cumulative grade point average (GPA), and
persistence, as measured by the length of time a student remained enrolled at an
institution toward completion of a degree. Of all the studies analyzed, 18 studies
(N=9,598) met the inclusion criteria of academic self-efficacy predicting GPA. A
substantial correlation emerged, ρ = .496 (where ρ = estimated true correlation between
the predictor construct and the performance criterion, corrected for measurement error in
both the predictor and criterion). An additional 6 studies (N = 6,930) met the inclusion
criteria of academic self-efficacy predicting persistence. Once again, a substantial
correlation emerged, ρ = .359.
Other investigators have shown that as the conceptual equivalence of the predictor
and criterion variables increases, so too does the strength of the relationship between
them. For example, in their meta-analyses of a large body of prospective studies,
Hansford and Hattie (1982) found that relatively specific academic self-concepts offered
better predictions of academic ability (r = .42) than global self-esteem (r = .22).
Similarly, Valentine et al. (2004) reported that the predictor-outcome associations were
stronger when the researchers assessed self-views specific to the academic domain and
when measures of self-beliefs and achievement were matched according to subject area.
This finding provides direct support for the notion that the conceptual equivalence of
predictor and criterion variables systematically determines the strength of the
relationships observed between them.
SELF-KNOWLEDGE IN EVERYDAY LIFE 17
Self-verification among married couples.
Another line of research that meets the conceptual equivalence criterion was
designed to examine people’s reactions to evaluations from their marriage partners. These
studies were inspired by Swann’s (1983) self-verification formulation. The theory
assumes that because self-views play a vital role in organizing reality, positioning people
in the world, and guiding behavior, people become invested in preserving them. To this
end, they choose evaluations that confirm their self-views over evaluations that challenge
their self-views. Most provocatively, the theory posits that people should prefer selfverifying evaluations even if the self-views happen to be negative. For example, those
who see themselves as disorganized or unintelligent should prefer evidence that others
perceive them as disorganized or unintelligent. This prediction challenges the widely
held notion that there is a universal preference for positive evaluations (Baumeister et al.,
2003; Jones, 1973).
Among people involved in relationships, self-verification strivings should manifest
themselves by influencing the quality and duration of the relationship. Self-views should
be especially predictive of reactions to evaluations when two conditions are met. First,
there should be an expectation that the relationship with the evaluator will persist (e.g.,
Hixon & Swann, 1993), because evaluations from acquaintances or insignificant
relationship partners have relatively few implications for people’s efforts to position
themselves in the world (e.g., Swann, De La Ronde & Hixon, 1994). In addition, selfverification effects occur when evaluations refer to relatively objective qualities such as
intelligence, social skills, and athleticism. They do not occur when the evaluations consist
of subjective assessments of the target’s likability or suitability as a partner, as self-
SELF-KNOWLEDGE IN EVERYDAY LIFE 18
verification theory assumes (and Murray, Holmes, & Griffin, 1996 have shown) that even
people who have negative self-views still wish to be in relationships and prefer partners
who want to remain with them.
We searched the literature for studies of the relationship of self-verification to
relationship quality that met the foregoing criteria. The conceptual predictor variable was
self-verification among married couples. Operationally, self-verification was defined as
the interaction term in regressions in which the predictors were (a) participants’ selfviews and (b) spouses’ evaluations of participants. The outcome variables were various
indices of relationship quality, such as intimacy, marital happiness, and
separation/divorce. Significant interactions reflected a tendency for spouses to enjoy
better relationship quality insofar as their spouse perceived them as they perceived
themselves. Specifically, self-verification occurred when people with positive self-views
enjoyed better relationship quality insofar as their partners appraised them positively and
people with negative self-views enjoyed better relationship quality insofar as their
partners appraised them negatively.
Several search strategies were used to locate studies that met inclusion criteria.
First, the following computerized reference databases were searched: Academic Search
Premier, Medline, PsycINFO, PsycARTICLES and Sociological Collection. We
searched for any records that contained self-verification and married. Additional search
strategies included reviewing the reference sections of retrieved studies and previous
review articles (Swann, Rentfrow & Guinn, 2003) as well as contacting several
prominent researchers requesting access to any relevant data not available in published
papers. These search strategies yielded over 50 unique references to studies. The first
SELF-KNOWLEDGE IN EVERYDAY LIFE 19
author read abstracts of all of the articles. If the abstract was promising, he read the
article itself to determine if it met the inclusion criterion.
The most common reason for excluding a study was that the authors mentioned, but
did not measure, self-verification. In addition, a study by Burke & Stets (1999) was
removed because the measure of self-views involved roles rather than specific, concrete
self-views that were assessed in all of the other studies (e.g., intellectual ability, social
ability, athletic ability, etc.). Studies by Murray et al. (1996) and Sacco & Phares (2001)
were excluded because the measures of spousal evaluation reflected likability of their
partners (e.g., perceptions of the partners’ virtues, faults, and social commodities; liking)
rather than evaluations of objective, relationship-independent qualities. Finally, Katz et
al. (1996) was removed because they failed to report the interactive effects of self-views
and spousal appraisals on the outcome variables. This left the six studies listed in Table 1
in our final pool.
For each of the six studies, the findings were standardized to a Cohen’s (1988) d
effect size and corrected for sampling error, according to the meta-analysis methods of
Schmidt and Hunter (1990) and Hunter, Schmidt, and Jackson (1982), as outlined by
Lyons (2003). The results of this analysis are outlined in Table 2. As can be seen in
Table 2, the average effect size for the self-verification effects was d = .40 which Cohen
has described as a “medium” effect size. In all studies save one, the interaction effect
reflected a pattern that supported the strong form of the self-verification hypothesis. That
is, whereas people with positive self-views enjoyed higher relationship quality insofar as
their spouse appraised them positively, people with negative self-views displayed the
opposite tendency, enjoying higher relationship quality insofar as their spouse appraised
SELF-KNOWLEDGE IN EVERYDAY LIFE 20
them negatively. The single exception to this pattern was Murray et al. (2000) who found
support for the self-verification hypothesis only among participants with positive selfviews. That is, whereas people with positive self-views were more satisfied with positive
spouses, people with negative self-views were equally satisfied with positive and
negative partners.
The self-verification research provides a nice complement to the research on selfefficacy discussed above because the findings make a stronger case for the causal
influence of self-views on the outcome variable. For example, critics of the self-efficacy
research could argue that people’s self-views merely reflected personal qualities (e.g.,
ability) or situational variables (e.g., experience in academic settings) that later caused
their performance on the outcome variables. If so, then self-views played no causal role
in the findings. The self-verification findings are less susceptible to this explanation,
however, because it is less likely that any single variable could explain the interactive
effects of participants’ self-views and spousal evaluations (i.e., the tendency for people
with positive self-views to be especially satisfied with spouses who evaluated them
positively while people with negative self-views were especially satisfied with partners
who evaluated them negatively).
Well controlled, large scale, longitudinal studies using self-esteem as a predictor.
Critics of past research on the predictive validity of self-esteem (e.g., Baumeister et
al., 2003) disparaged the lack of studies using objective measures, longitudinal designs,
large representative samples, and appropriate controls to test the predictive utility of selfesteem indices. Quite recently one team of researchers (Donnellan et al., 2005;
Trzesniewski et al., in press) has begun to address these concerns.
SELF-KNOWLEDGE IN EVERYDAY LIFE 21
The initial paper (Donnellan et al., 2005) reported that low self-esteem predicted
externalizing problems two years later whether they examined self, teacher, parent, and
interviewer-based measures of self-esteem and externalizing problems, and for
participants from different nationalities (United States and New Zealand) and age groups
(adolescents and college students). Moreover, this relation held when the investigators
controlled for potential confounding variables such as supportive parenting, parent and
peer relationships, SES, and IQ. In a follow-up study that built upon the methodological
strengths of the earlier work, Trzesniewski et al. (in press) followed a group of
adolescents for 11 years into adulthood. Even after controlling for numerous rival
predictors of the outcome measures, the investigators found that self-esteem was a
significant predictor of major depression disorder, anxiety disorder, tobacco dependence,
criminal convictions, school drop out, and money and work problems. Moreover, these
relations held whether the outcome measures were reports by the participants or
observers.
Although the effect sizes reported by Donnellan et al. (2005) and Trzesniewski et
al. (in press) would be considered small to negligible using Baumeister et al.’s (2003)
lofty standards, earlier we explained why these standards are overly high. More
generally, we believe that the results of this line of research are remarkable in at least
three respects. First, numerous potential confounding variables (e.g., depression,
neuroticism) were appropriately controlled for and objective outcome measures were
examined. These methodological features not only address two key problems that marred
previous research on self-esteem (e.g., Baumeister et al., 2003), but they also counter
critics who have claimed that measures of self-esteem predict outcome variables only
SELF-KNOWLEDGE IN EVERYDAY LIFE 22
because they happen to be correlated with variables that are causally related to these
outcome variables (e.g., Judge, Erez, Bono, & Thoresen, 2002; Watson, Suls, & Haig,
2002). Second, the range and social significance of the outcomes predicted by self-esteem
(e.g., depression, anxiety disorders, criminal convictions, school drop out, money and
work problems, etc.) are impressive by any standard. Third, the 11 year time lag between
the measure of the predictor and criterion in the second study was substantial. The fact
that self-esteem scores predicted outcomes over such a long period supports the idea that
self-esteem can have enduring effects on people and strengthens the argument that the
effects of self-esteem were causal.
Trzesniewski et al. (in press) also conducted a set of analyses indicating that
predictive validity was particularly impressive when the conceptual equivalence of the
predictor and outcomes was relatively high. That is, when self-esteem was used to
predict aggregated outcome measures instead of the specific outcomes discussed above,
the data indicated that teenagers with low self-esteem ran an elevated risk for developing
difficulties as adults. For example, among adults with five or more problems during
adulthood, 63% had low self-esteem during adolescence and only 15% had high selfesteem during adolescence. Similarly, among problem-free adults, 50% had high selfesteem when they were adolescents and only 16% had low self-esteem during
adolescence.
Werner and Smith (1992) have also reported evidence that when a measure of selfself-esteem was used to predict aggregated outcomes, substantial relationships emerged.
They focused on a sample of extremely impoverished youth in the Kauai Longitudinal
Study. Self-esteem was assessed using interviews at age 18. When participants were 32
SELF-KNOWLEDGE IN EVERYDAY LIFE 23
years old, the investigators collected an aggregated measure of quality of adult
adaptation. The findings indicated that the self-esteem ratings of teenagers predicted
their adaptation 14 years later (r = .24 for males, r = .41 for females). Considered
together with the results of the Trzesniewski et al. (in press), these data suggest that the
relation between self-esteem and important outcomes are particularly strong when
aggregated outcomes are examined.
Conclusions and Implications
Recent assessments of the importance of self-views in predicting important social
outcomes (e.g., Baumeister et al., 2003; Crocker & Park, 2004; Scheff & Fearon, 2004)
have raised several conceptual and empirical issues that merit careful consideration. For
example, we agree with Baumeister et al.’s (2003) contention that the modest correlation
between measures of self-esteem and narcissism raises the possibility that narcissism
might sometimes drive effects erroneously attributed to self-esteem. Nevertheless, we
disagree that the best way to handle this potential problem is to broaden the definition of
high self-esteem to encompass narcissism. Instead, theory and recent research suggest
that the optimal way to handle this confound is to measure each construct separately and
control for the effects of one while assessing the effects of the other. This strategy is
supported by other recent theory and research on the self which have demonstrated the
utility of making more rather than less fine-tuned discriminations between distinct types
of self-representations (e.g., Crocker & Wolfe, 2001; Kernis, 2003; Deci & Ryan, 1995;
Marsh, 1990; Tafarodi & Swann, 2002).
We also agree with recent contentions that measures of global self-esteem are only
modestly related to specific outcome variables. Nevertheless, recent developments in
SELF-KNOWLEDGE IN EVERYDAY LIFE 24
psychometric theory have shown that such modest relationships are expected and
respectable when outcome variables are multiply determined. Moreover, even if the
outcome variable is not multiply determined, if it is socially important modest
relationships to self-esteem are still noteworthy. Furthermore, the history of research on
constructs related to self-esteem (e.g., attitudes and traits) suggests that recent critics of
self-esteem research have posed the question in an overly narrow fashion. This has, in
turn, led to a failure to appreciate the implications of the fact that the specific outcomes
examined in the research reviewed by Baumeister et al. (2003) are not equivalent in their
conceptual breadth (i.e., conceptually equivalent) to global predictors such as self-esteem.
When one broadens the question to encompass the predictive utility of self-concepts as
well as self-esteem (i.e., self-views) in predicting conceptually equivalent outcomes, one
finds quite respectable predictor-outcome relationships. For example, self-perceived
efficacy predicts academic performance, and married people are more satisfied when
their spouses see them as they see themselves. Furthermore, even years in advance,
global measures of self-esteem predict outcomes ranging from dropping out of school to
depression, especially when the outcome variables are aggregated (e.g., Trzesniewski et
al., in press; Werner & Smith, 1992).
If self-views are meaningfully related to socially significant outcomes, does this
mean that it makes sense to take steps to improve those self-views? We believe that it
does. Indeed, there is evidence that programs designed to improve self-esteem are
effective in fostering a host of favorable outcomes, such as standardized test scores,
school disciplinary reports, and use of drugs and alcohol (e.g., DuBois & Flay, 2004;
Haney & Durlak, 1998). Nevertheless, we believe that it is essential that such research be
SELF-KNOWLEDGE IN EVERYDAY LIFE 25
buttressed by careful, theory driven research designed to specify the precise mediators of
these effects.
Why? There are at least two reasons why research specifying the mechanisms
underlying self-esteem change programs is important. First, of the effective programs of
which we are aware, all are multifaceted and relatively complex, including efforts to
improve self-efficacy and interpersonal relationships as well as self-esteem. Because the
effectiveness of individual components of these programs is rarely, if ever, documented,
it seems likely that such programs include a mix of effective and ineffective strategies (or
strategies that are effective for some people but ineffective for others). This raises the
possibility that one could strengthen such programs by bolstering the effective
components and eliminating the ineffective ones (e.g., DuBois & Tevendale, 1999).
Even more important, identification of the components that underlie effective
programs could also help silence critics by distinguishing programs based on nonsense
from those based on sound psychological principles. Consider the (in some instances,
well deserved) caricatures of self-esteem programs supplied by the media. Perhaps the
best known example was offered by humorist Al Franken, who parodied self-esteem
enhancement programs by having his character (Stuart Smalley) pretend to gaze
tentatively into the mirror, smile, and then carefully recite "I'm good enough, I'm smart
enough, and gosh darn it, people like me." The newly esteemed Smalley then beamed
triumphantly. This scenario was amusing because it was so obvious to everyone (except
Stuart) that such affirmation procedures are hopelessly misguided.
Although there are self-esteem enhancement programs based on such simpleminded strategies, these Panglossian programs are a far cry from the demonstrably
SELF-KNOWLEDGE IN EVERYDAY LIFE 26
effective ones reviewed in the recent literature (e.g., DuBois & Flay, 2004; Haney &
Durlak, 1998). The effective programs are far more ambitious and comprehensive than
programs based on affirmations, so much so that calling them “Self-esteem enhancement
programs” is a misnomer. Instead of focusing exclusively on people’s momentary selfesteem, the effective programs emphasize procedures that are designed to alter the raw
materials that provide a basis for healthy, sustainable, self-esteem. Ideally, these
programs cultivate behaviors that produce self-views that are both realistic (i.e., based on
objective evidence) and adaptive (i.e., emphasizing activities that are predictive of long
term adjustment in society). As such, the objective of these programs contrasts sharply
with the objective of affirmation procedures, which (at best) will produce positive selfimages that are fanciful and ephemeral (e.g., Crocker & Park, 2004; Swann, 1996). For
this reason, participants in the effective programs do not magically affirm their way into
possessing high self-esteem. Instead, they enact systematic improvements in their lives
which, in turn, nurture and maintain positive self-views--including positive self-concepts
as well as high self-esteem. Therefore, the principles that underlie such programs make
sound theoretical sense and it is misleading and unfortunate to confuse them with
programs that do not.
Coda
Skeptics could object: “If the active ingredients in self-esteem change programs
have as much to with changing the behaviors, life circumstances, and associated specific
self-views that give rise to self-esteem rather than self-esteem itself, then it seems
arbitrary and misleading to use labels that imply that self-esteem is the central focus of
such programs. Perhaps these programs have nothing to do with self-esteem at all!”
SELF-KNOWLEDGE IN EVERYDAY LIFE 27
We believe that it is legitimate to ask whether self-esteem plays as dominant a role
in self-esteem enhancement programs as the name of such programs implies. In fact, we
urge researchers to examine this issue empirically in the future. We expect that such
research will reveal that although it is not enough to change self-views only, it is also not
enough to change people’s life circumstances and behavior only--self-views must also
change. One reason for this is that self-views determine how people react to changes to
their social environment. Imagine, for example, a teacher who announces that he has
decided to disregard the quality of students’ performances on tests and give the entire
class A’s. When students who have poor self-academic views suddenly begin receiving
A’s without any changes in their behaviors, it is unlikely that they will suddenly develop
a sense of academic efficacy—and even more unlikely that they will begin to perform
better. Rather, theory and research on self-efficacy (Bandura, 1989) suggests that merely
changing the social setting will not necessarily produce desired changes in behavior—
self-views must change as well.
Similarly, changing people’s behavior will also fail to produce lasting positive
changes in people’s life circumstances unless self-views undergo corresponding changes.
For instance, if a therapist were to coach clients with negative self-views to marry
spouses who see them more favorably than they see themselves, the therapist might
succeed in bringing them to marry someone who sees them favorably. Nevertheless,
research on self-verification (Swann et al., 2002) suggests that unless the self-views of
such clients become more positive, they will feel that their partners’ appraisals are overly
positive and they will withdraw from the relationship, either psychologically or through
divorce.
SELF-KNOWLEDGE IN EVERYDAY LIFE 28
From this vantage point, people’s self-esteem, self-concepts, behaviors, and social
conditions are embedded in cycles in which each of the four components influence and
constrain one another in profound ways. For this reason, just as producing lasting
changes in self-views requires altering the behaviors and social conditions that nourish
those self-views, producing lasting changes in people’s behaviors and social conditions
requires changing their self-views. In the final analysis, whether one is simply trying to
understand why people behave as they do or attempting to improve their life conditions,
one must recognize a basic fact of social life: people’s self-views do matter.
SELF-KNOWLEDGE IN EVERYDAY LIFE 29
Author Note
We thank Serena Chen, Brent Donnellan, David DuBois, Russ Fazio, Nancy Hazen,
Lee Jussim, Stephanie Madon, Fred Rhodewalt, Rick Robins, Robert Rosenthal and Kali
Trzesniewski for helpful comments on an earlier version of this paper and Sandra Murray
for providing us with data for use in our meta-analysis. Address correspondence to
William B. Swann, Jr., Department of Psychology, University of Texas at Austin, Austin,
Texas 78712, Email: swann@mail.utexas.edu.
SELF-KNOWLEDGE IN EVERYDAY LIFE 30
References
Abelson, R.P. (1995). Statistics as principled argument. Hillsdale, NJ: Erlbaum.
Ahadi, S. & Diener, E. (1989). Multiple determinants and effect size.
Journal of Personality and Social Psychology, 56, 398-406.
Ajzen, I. (1991). The Theory of Planned Behavior. Organizational Behavior & Human
Decision Processes, 50, 179-211.
Ajzen, I., & Fishbein, M. (1977). Attitude-behavior relations: A theoretical analysis and
review of empirical research. Psychological Bulletin, 84, 888-918.
Ajzen, I., & Fishbein, M. (2005). The Influence of Attitudes on Behavior. In D.
Albarracín, B. T. Johnson, & M. P. Zanna (Eds.), The handbook of attitudes (pp.
173-221). Mahwah, NJ: Erlbaum.
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental
disorders (4th ed., text revision). Washington, DC: Author.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice Hall.
Bandura, A. (1989). Human agency in social cognitive theory. American Psychologist,
44, 1175-1184.
Baumeister, R.F., Bushman, B.J., & Campbell, W.K. (2000). Self-esteem, narcissism, and
aggression: Does violence result from low self-esteem or from threatened
egotism? Current Directions in Psychological Science, 9, 26-29.
Baumeister, R.F., Campbell, J.D., Krueger, J.I., & Vohs, K.D. (2003). Does High SelfEsteem Cause Better Performance, Interpersonal Success, Happiness, or Healthier
Lifestyles? Psychological Science in the Public Interest, 4, 1-44.
SELF-KNOWLEDGE IN EVERYDAY LIFE 31
Baumeister, R.F., Smart, L., & Boden, J.M. (1996). Relation of Threatened Egotism to
Violence and Aggression: The Dark Side of High Self-Esteem. Psychological
Review, 103, 5-33.
Burke, P.J. & Stets, J.E. (1999). Trust and Commitment through Self-Verification.
Social Psychology Quarterly, 62, 347-360.
Bushman, B.J. & Baumeister, R.F. (1998). Threatened Egotism, Narcissism, Self-Esteem,
and Direct and Displaced Aggression: Does Self-Love or Self-Hate Lead to
Violence? Journal of Personality and Social Psychology, 75, 219-229.
Campbell, W. K. (2001). Is narcissism really so bad? Psychological Inquiry, 12, 214216.
Cast, A. D. & Burke, P. J. (2002). A Theory of Self-Esteem. Social Forces, 80, 10411068.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences. (2nd Ed).
Crocker J. & Park (2004). The Costly Pursuit of Self-Esteem. Psychological Bulletin,
130, 392-414.
Crocker, J., & Wolfe, C. T. (2001). Contingencies of self-worth. Psychological Review,
108, 593-623.
Dawes, R.M. (1996). House of cards: Psychology and Psychotherapy Built on Myth.
New York, NY: Free Press.
Deci, E. L., & Ryan, R. M. (1995). Human autonomy: The basis for true self-esteem. In
M. Kernis (Ed.), Efficacy, agency, and self-esteem (pp. 31-49). New York:
Plenum.
SELF-KNOWLEDGE IN EVERYDAY LIFE 32
De La Ronde, C. & Swann, W. B. Jr. (1998). Partner verification: Restoring shattered
images of our intimates. Journal of Personality and Social Psychology, 75, 374382.
Donnellan, B., Trzesniewski, K., Robins, R., Moffitt, T., & Caspi, A. (2005). Low SelfEsteem Is Related to Aggression, Antisocial Behavior, and Delinquency.
Psychological Science, 16, 328-335.
Dubois, D. L. & Flay, B.R. (2004). The Heathly Pursuit of Self-Esteem: Comment on
and Alternative to the Crocker and Park (2004) Formulation. Psychological
Bulletin, 130, 415-420.
Dubois, D.L. & Tevendale, H.D. (1999). Self-esteem in childhood and adolescence:
Vaccine or epiphenomenon? Applied & Preventive Psychology, 8, 103-117.
Epstein, S. (1979). Stability of behavior: On predicting most of the people much of the
time. Journal of Personality and Social Psychology, 37(7), 1097-1126.
Fazio, R. H., & Olson, M. A. (2003). Attitudes: Foundations, functions, and
consequences. In M. A. Hogg & J. Cooper (Eds.), The Handbook of Social
Psychology (pp. 139-160). London: Sage
Fleeson, W. (2004). Moving personality beyond the person-situation debate: The
challenge and opportunity of within-person variability. Current Directions in
Psychological Science, 13, 83-87.
Funder, D. C. & Colvin, C. R. (1991) Explorations in behavioral consistency: Properties
of persons, situations, and behaviors. Journal of Personality and Social
Psychology, 60, 773-94.
SELF-KNOWLEDGE IN EVERYDAY LIFE 33
Haney, P. & Durlak, J.A. (1998). Changing Self-Esteem in Children and Adolescents: A
Meta-Analytic Review. Journal of Clinical Child Psychology, 27, 423-433.
Hansford, B. C., & Hattie, J. A. (1982). The relationship between self and
achievement/performance measures. Review of Educational Research, 52, 123–
142.
Higgins, E. T. (2000). Does personality provide unique explanations for behavior?:
Personality as cross-person variability in general principles. European Journal of
Personality, 14, 391-406.
Hixon, J. G., & Swann, W. B., Jr. (1993). When does introspection bear fruit? Selfreflection, self-insight, and interpersonal choices. Journal of Personality and
Social Psychology, 64, 35-43.
Hunter, J. and Schmidt, F. (1990). Methods of meta-analysis: Correcting error and bias in
research findings. Beverly Hills CA: Sage.
Hunter, J., Schmidt, F., and Jackson, G. (1982). Meta-Analysis: Cumulating research
findings across studies. Beverly Hills CA: Sage.
Johnson, B. T., & Eagly, A. H. (2000). Quantitative synthesis of social psychological
research. In H. T. Reis & C. M. Judd (Eds.), Handbook of research methods in
social and personality psychology (pp. 496-528). New York: Cambridge
University Press.
Jones, S. C. (1973). Self and interpersonal evaluations: Esteem theories versus
consistency theories. Psychological Bulletin, 79, 185-199.
Kaplan, H. B. (1986). Social psychology of self-referent behavior. New York: Plenum.
SELF-KNOWLEDGE IN EVERYDAY LIFE 34
Katz, J. Beach, S. R. H, & Anderson, P. (1996). Self-enhancement versus selfverification: Does spousal support always help? Cognitive Therapy and Research,
20, 345-360.
Kernis, M. H. (2003). Toward a conceptualization of optimal self-esteem. Psychological
Inquiry, 14, 1-26.
Kuiper, N. A., & Rogers, T. B. (1979). Encoding of personal information: self-other
differences. Journal of Personality and Social Psychology, 37, 499-514.
Lyons, L.C. (2003). Meta-analysis: Methods of Accumulating Results Across Research
Domains, 9-13. Retrieved on July 27, 2005 from
http://www.lyonsmorris.com/MetaA/studyfx1.htm
Markus, H. (1977). Self-schemas and processing information about the self. Journal of
Personality and Social Psychology, 35, 63-78.
Marsh, H. W. (1990). A multidimensional, hierarchical model of self-concept:
Theoretical and empirical justification. Educational Psychology Review, 2, 77172
Meyer, G. J., Finn, S. E., Eyde, L. D., Kay, G. G., Moreland, K. L., Dies,
R. R., Eismane, E. J., Kubiszyn, T. W., & Reed, G. M. (2001). Psychological
testing and psychological assessment: A review of the evidence and issues.
American Psychologist, 56, 128-165.
Mischel, W. (1968). Personality and assessment. New York, NY: John Wiley & Sons,
Inc.
SELF-KNOWLEDGE IN EVERYDAY LIFE 35
Mischel, W., & Shoda, Y. (1999). Integrating dispositions and processing dynamics
within a unified theory of personality: The Cognitive Affective Personality
System (CAPS). In L. A. Pervin & O. John (Eds.), Handbook of personality:
Theory and research 2. New York: Guilford, pp.197-218.
Morf, C. C., & Rhodewalt, F. (1993). Narcissism and self-evaluation maintenance:
Explorations in object relations. Personality and Social Psychology Bulletin, 19,
668-676.
Morf, C.C. & Rhodewalt, F. (2001). Unraveling the Paradoxes of Narcissism: A Dynamic
Self-Regulatory Processing Model. Psychological Inquiry, 12 , 177-197.
Multon, K. D., Brown, S. D., & Lent, R. W. (1991). Relation of self-efficacy beliefs to
academic outcomes: A meta-analytic investigation. Journal of Counseling
Psychology, 38, 30-38.
Murray, S. L., Holmes, J. G., & Griffin, D. W. (1996). The benefits of positive illusions:
Idealization and the construction of satisfaction in close relationships. Journal of
Personality and Social Psychology, 70, 79-98.
Murray, S. L., Holmes, J. G., & Griffin, D. W. (2000) Self-Esteem and the Quest for
Felt Security: How Perceived Regard Regulates Attachment Processes. Journal of
Personality and Social Psychology, 78, 478-498.
Paulhus, D.L., Robins, R.W., Trzesniewski, K.H., & Tracy, J.L. (2004). Two Replicable
Suppressor Situations in Personality Research. Multivariate Behavioral
Research, 39, 303-329.
Pelham, B. W. (1991). On confidence and consequence: The certainty and importance of
self-knowledge. Journal of Personality and Social Psychology, 60, 518-530.
SELF-KNOWLEDGE IN EVERYDAY LIFE 36
Pelham, B. W., & Swann, W. B., Jr. (1994). The juncture of intrapersonal and
interpersonal knowledge: Self-certainty and interpersonal congruence. Personality
and Social Psychology Bulletin, 20, 349-357.
Raskin, R., Novacek, J., & Hogan, R. (1991). Narcissism, self-esteem, and defensive
self-enhancement. Journal of Personality, 59, 19-38.
Rhodewalt, F. (2005). Social Motivation and Object Relations: Narcissism and
interpersonal self-esteem regulation. In J. P. Forgas and K. D. Williams and S.
Laham (Eds.), Social Motivation: Conscious and Unconscious Processes.
(pp.332-350). New: Psychology Press.
Rhodewalt, F., & Morf, C. C. (1995). Self and interpersonal correlates of the Narcissistic
Personality Inventory: A review and new findings. Journal of Research in
Personality, 29, 1-23.
Rhodewalt, F., & Morf, C. C. (1998). On self-aggrandizement and anger: A temporal
analysis of Narcissism and affective reactions to success and failure. Journal of
Personality and Social Psychology, 74, 672-685.
Rhodewalt, F., Madrian, J. C., & Cheney, S. (1998). Narcissism, self-knowledge
organization, and emotional reactivity: The effect of daily experiences on selfesteem and affect. Personality and Social Psychology Bulletin, 24, 75-87.
Ritts, V., & Stein, J. R. (1995). Verification and commitment in marital relationships: An
exploration of self-verification theory in community college students.
Psychological Reports, 76, 383-386.
SELF-KNOWLEDGE IN EVERYDAY LIFE 37
Robbins, S.B., Lauver, K., Le, H., Langley, R., Carlstrom, A. (2004). Do Psychosocial
and Study Skill Factors Predict College Outcomes? A Meta-Analysis.
Psychological Bulletin, 130, 261-288.
Roberts, J. E., Gotlib, I. H., & Kassel, J. D. (1996). Adult attachment security and
symptoms of depression: The mediating roles of dysfunctional attitudes and low
self-esteem. Journal of Personality and Social Psychology, 70, 310-320.
Rosenberg, M. (1973). Which significant others? American Behavioral Scientist, 16,
829-860.
Rosenberg, M. R., Schooler, C., Schoenbach, C., & Rosenberg, F. (1995). Global selfesteem and specific self-esteem: Different concepts, different outcomes. American
Sociological Review, 60, 141-156.
Rosenthal, R. (1994). Interpersonal expectancy effects: A 30-year perspective. Current
Directions in Psychological Science, 3, 176-179.
Rosenthal, R. (1994). Parametric measures of effect size. In H. Cooper & L. V. Hedges,
The Handbook of Research Synthesis, (Ch. 16, pp. 231-244). New York: Russell
Sage.
Rosenthal, R.& Rubin, D.B. (2003). requivalent: A Simple Effect Size Indicator.
Psychological Methods, 8, 492-496
Sacco, W.P. & Phares, V. (2001) Partner Appraisal and Marital Satisfaction: The Role of
Self-Esteem and Depression. Journal of Marriage and Family, 63, 504-413.
Scheff, T.J. & Fearon, D.S. (2004). Cognition and Emotion? The Dead End in SelfEsteem Research. Journal for the Theory of Social Behaviour, 34, 73-91.
SELF-KNOWLEDGE IN EVERYDAY LIFE 38
Schafer, R.B., Wickrama, K.A.S., & Keith, P.M. (1996). Self-concept disconfirmation,
psychological distress, and marital happiness. Journal of Marriage and the
Family, 58, 167-177.
Sedikides, C, Rudich, E. A., Gregg, A. P., Kumashiro, M, Rusbult, C. (2004). Are
Normal Narcissists Psychologically Healthy?: Self-Esteem Matters. Journal of
Personality and Social Psychology, 87, 400-416.
Sherman, S.J. & Fazio, R. H. (1983). Parallels between attitudes and traits as predictors
of behavior. Journal of Personaity, 51, 308-345.
Snyder, M. & Stukas, A.A., Jr. (2000). Self-fulfilling prophecy. In Kazdin, A E. (Ed.),
Encyclopedia of psychology, 7, pp. 216-218. Washington, DC: American
Psychological Association.
State of California, Assembly Bill No. 3659, Chapter 1065. Approved by the governor,
September 23, 1986.
Strube, M. J. (1991). Multiple determinants and effect size: A more general
method of discourse. Journal of Personality and Social Psychology, 61,
1024-1027.
Swann, W. B., Jr. (1990). To be adored or to be known : The interplay of selfenhancement and self-verification. In R. M. Sorrentino & E. T. Higgins (Eds.)
Foundations of Social Behavior, (Vol. 2, pp. 408-448). New York: Guilford.
Swann, W. B., Jr. (1996). Self-traps: The elusive quest for higher self-esteem. Freeman:
New York
SELF-KNOWLEDGE IN EVERYDAY LIFE 39
Swann, W. B., Jr. (1983). Self-verification: Bringing social reality into harmony with the
self. In J. Suls & A. G. Greenwald (Eds.), Psychological perspectives on the self
(Vol. 2, pp. 33-66), Hillsdale, NJ: Erlbaum.
Swann, W. B., Jr., De La Ronde, C. & Hixon, J. G. (1994). Authenticity and positivity
strivings in marriage and courtship. Journal of Personality and Social
Psychology, 66, 857-869.
Swann, W.B., Jr., Rentfrow, P. J., & Guinn, J. (2002). Self-verification: The search for
coherence. In M. Leary and J. Tagney, Handbook of self and identity (pp. 367383): Guilford, New York.
Swann, W.B., Jr. & Seyle, C. (2005). Personality psychology’s comeback and its
emerging symbiosis with social psychology. Personality and Social Psychology
Bulletin, 31, 155-165.
Tafarodi, R. W., & Swann, W. B., Jr., (2001). Two-dimensional self-esteem: Theory and
measurement. Personality and Individual Differences, 31, 653-673.
Tracy, J.L. & Robins, R.W. (2003). "Death of a (Narcissistic) Salesman:" An Integrative
Model of Fragile Self-Esteem. Psychological Inquiry, 14, 57-63.
Trzesniewski, K., Donnellan, B., Moffitt, T., Robins, R., Poulton, R., & Caspi, A. (in
press). Low self-esteem during adolescence predicts poor health, criminal
behavior, and limited economic prospects during adulthood. Developmental
Psychology.
Valentine, J. C., DuBois, D. L., & Cooper, H. (2004). The relation between self-beliefs
and academic achievement: A meta-analytic review. Educational Psychologist,
39, 111–133.
SELF-KNOWLEDGE IN EVERYDAY LIFE 40
Wallace, H. M., & Baumeister, R. F. (2002). The performance of narcissists rises and
falls with perceived opportunity for glory. Journal of Personality and Social
Psychology, 82, 819-834.
Werner, E.E., & Smith, R.S. (1992). Overcoming the odds: High risk children from birth
to adulthood. Ithaca, NY: Cornell University Press.
Westen, D. (1990). The relations among narcissism, egocentrism, self-concept, and selfesteem: Experimental, clinical, and theoretical considerations. Psychoanalysis &
Contemporary Thought, 13, 183-239.
Wicker, A. W. (1969). Attitudes versus actions: The relation of verbal and overt
behavioral responses to attitude objects. Journal of Social Issues, 25, 41-78.
SELF-KNOWLEDGE IN EVERYDAY LIFE 41
Table 1
Summary of Studies Included in the Analysis
Study
Cast & Burke, 2002
Sample Size
Predictor
Outcome
231
Self-verification
De La Ronde & Swann, 1998
122
Self-verification
Intimacy
Murray, Holmes, & Griffin, 2000
162
Self-verification
Marital Satisfaction
60
Self-verification
Schafer, Wickrama, &
Keith, 1996
155
Self-verification
Swann, De La Ronde, &
Hixon, 1994
165
Self-verification
Ritts & Stein, 1995
Separation or Divorce
Commitment to relationship
Spousal happiness
Intimacy
SELF-KNOWLEDGE IN EVERYDAY LIFE 42
Table 2
Accumulated Effect Sizes, Corrected for Sampling Error
Study
Sample Size
Cohen’s d
Effect of Self-Verification on
husbands and wives:
Cast & Burke, 2002
231
.2518a
Separation or Divorce
De La Ronde & Swann,
1998b
122
.4902
Intimacy with spouse
Murray, Holmes, & Griffin,
2000
162
.2665d,f
Ritts & Stein, 1995
60
.9256
Commitment to relationship
Schafer, Wickrama, &
Keith, 1996
155
.3159e
Marital happiness
Swann, De La Ronde, &
Hixon, 1994c
165
.6589
Intimacy with spouse
Sample weighted mean d:
a
Marital Satisfaction
.4032g
Odds ratio converted to dequivalent effect size according to method outlined by Rosenthal & Rubin (2003).
Non-significant main effects in these studies were estimated to have a value of F(1, 121) = 1 and F(1, 164) =
1, respectively. Estimates were made because actual values for main effects were not reported in original studies. As a
double-check, calculations were performed with F’s = 1.5 in both cases, and the effects on the sample weighted mean d
were negligible.
d,eTo obtain the combined effect for both husbands and wives for each of these studies, we calculated the
Cohen’s d effect size for each reported correlation (husband, wife), then averaged that value within each study. This
was essentially equivalent to averaging the correlations and then obtaining the Cohen’s d from that averaged
correlation. The averaged Cohen’s d was then entered in the analysis in the same manner as the remaining predictors’
effect sizes. These two were the only studies that evaluated the effects on husbands and wives separately.
f These data were not reported in Murray et al. (2000) but Murray kindly provided them to us.
g Biased population variance estimate 2
sd  .0336 for sample weighted mean d.
b, c