Gender Differences in Domain-Specific Self-Esteem: A Meta-Analysis

Abstract

This meta-analysis examines gender differences in 10 specific domains of self-esteem across 115 studies, including 428 effect sizes and 32,486 individuals. In a mixed-effects analysis, men scored significantly higher than women on physical appearance (d 0.35), athletic (d 0.41), personal self (d 0.28), and self-satisfaction self-esteem (d 0.33). Women scored higher than men on behavioral conduct (d 0.17) and moral-ethical self-esteem (d 0.38). The gender difference in physical appearance self-esteem was significant only after 1980 and was largest among adults. No significant gender differences appeared in academic, social acceptance, family, and affect self-esteem. The results demon- strate the influence of reflected appraisals on self-esteem.

Full text

Gender Differences in Domain-Specific Self-Esteem:
A Meta-Analysis
Brittany Gentile
University of Georgia
Shelly Grabe
University of California, Santa Cruz
Brenda Dolan-Pascoe
and Jean M. Twenge
San Diego State University
Brooke E. Wells
Center for HIV Educational Studies and Training (CHEST)
and National Development and Research Institutes, Inc.
Alissa Maitino
Alliant International University
This meta-analysis examines gender differences in 10 specific domains of self-esteem across 115 studies,
including 428 effect sizes and 32,486 individuals. In a mixed-effects analysis, men scored significantly
higher than women on physical appearance (d ⫽ 0.35), athletic (d ⫽ 0.41), personal self (d ⫽ 0.28), and
self-satisfaction self-esteem (d ⫽ 0.33). Women scored higher than men on behavioral conduct (d ⫽
⫺0.17) and moral– ethical self-esteem (d ⫽⫺0.38). The gender difference in physical appearance
self-esteem was significant only after 1980 and was largest among adults. No significant gender
differences appeared in academic, social acceptance, family, and affect self-esteem. The results demon-
strate the influence of reflected appraisals on self-esteem.
Keywords: gender differences, self-esteem, adolescence, physical appearance, meta-analysis
Since the mid-1990s, reports in the popular media have sug-
gested that girls—particularly teens— have distressingly low self-
esteem (e.g., Pipher, 1994). Recent advertisements like the “Dove
Self-Esteem Fund” continue to suggest that girls and women suffer
from negative self- images. But are girls and women actually
deficient in self-esteem compared with boys and men? Three
previous meta-analyses found that the effect size for the gender
difference in self-esteem was small, d ⫽ 0.15; in adolescence the
difference was d ⫽ 0.33, a small to medium effect size (Kling,
Hyde, Showers, & Buswell, 1999; Major, Barr, Zubek, & Babey,
1999; Twenge & Campbell, 2001).
However, these previous meta-analyses examined gender dif-
ferences in global but not domain-specific self-esteem. These are
two distinct concepts in the literature. Global self-esteem is “the
positivity of the person’s self-evaluation” (Baumeister, 1998, p.
694) or “the level of global regard that one has for the self as a
person” (Harter, 1993, p. 88). Domain-specific self-esteem, on the
other hand, describes self-satisfaction in specific areas (e.g., ap-
pearance, academics, social). Self-esteem may vary considerably
from one domain to another. Thus, domain-specific self-esteem
may show larger gender differences than global self-esteem (e.g.,
Sondhaus, Kurtz, & Strube, 2001; Tiggemann & Rothblum, 1997).
The present research undertakes a comprehensive meta-analysis of
gender differences in 10 domain-specific areas of self-esteem.
We draw primarily on two theoretical approaches to self-esteem.
First, the reflected appraisals model maintains that people base
their self-esteem on others’ opinions and perceptions. This idea has
a long history; Cooley (1902) argued that self-esteem arises from
the appraisals of others. Mead (1934) took this concept a step
further, maintaining that our self-esteem is also influenced by the
“generalized other”—thus the entire sociocultural environment
(which in modern times would include the media). Thus, the
reflected appraisals model would predict gender differences in
areas where societal standards are different for men and women.
The reflected appraisals model has been conceptualized more
recently as a “sociometer” (Leary, Haupt, Strausser, & Chokel,
1998; Leary, Tambor, Terdal, & Downs, 1995) or the “need to
belong” (Baumeister & Leary, 1995). These authors theorize that
self-esteem is primarily rooted in our relationships with others—
what others think of us, whether they accept us, and so on.
Interactions with others are good predictors of fluctuations in
self-esteem, and even at the level of nations, countries with high
levels of interaction between friends have higher mean self-esteem
(Denissen, Penke, & Schmitt, 2008). For gender differences in
self-esteem, the reflected appraisals model predicts that other peo-
ple’s perceptions of us—or of our gender as a whole—is a key
component of self-esteem. If someone’s interactions with others
around a particular domain are positive, then they would have high
Brittany Gentile, Department of Psychology, University of Georgia;
Shelly Grabe, Department of Psychology, University of California, Santa
Cruz; Brenda Dolan-Pascoe and Jean M. Twenge, Department of Psychol-
ogy, San Diego State University; Brooke E. Wells, Center for HIV Edu-
cational Studies and Training (CHEST) and National Development and
Research Institutes, Inc.; Alissa Maitino, California School of Professional
Psychology, Alliant International University.
Correspondence concerning this article should be addressed to Jean M.
Twenge, Department of Psychology, San Diego State University, 5500
Campanile Drive, San Diego, CA 92182-4611. E-mail:
jtwenge@mail.sdsu.edu
Review of General Psychology © 2009 American Psychological Association
2009, Vol. 13, No. 1, 34– 45 1089-2680/09/$12.00 DOI: 10.1037/a0013689
34

self-esteem in that domain. If they are negative and judgmental,
however, self-esteem would suffer. For example, some have the-
orized that negative interactions often occur around women’s
physical appearance, which is often scrutinized and discussed
(Fredrickson & Roberts, 1997); thus, women might have lower
self-esteem in the domain of physical appearance.
The competencies model (James, 1890) argues that people draw
self-esteem from accomplishments in certain areas. The compe-
tencies model predicts gender differences in areas where actual
performance differs. Thus, it is a model of self-efficacy, predicting
that people will feel high self-esteem when they have performed
well and will perform well when they have high self-esteem (e.g.,
Bandura, 1989). Although self-efficacy and self-esteem are distinct
concepts, domain-specific self-esteem has some overlap with self-
efficacy, because it addresses confidence in a certain area of
competence. Some of these areas, like academics and athletics, are
performance domains that may show a reciprocal relationship
between performance and self-esteem, with each influencing the
other. Thus, the competencies model (and self-efficacy theory)
predicts that when females perform better in an area, the gender
difference in that area will favor females. This most likely occurs
through two mechanisms. First, the average individual female is likely
to perform better in an area, and thus females’ self-esteem will be
higher than males’ in that area on average. For example, the average
girl earns better grades than the average boy, so the gender
difference in academic self-esteem should favor females. Second,
females may know that their gender group performs better in the
area, increasing their self-esteem in that area apart from their
individual performance. This mechanism can break down, how-
ever, if people apply shifting standards by only comparing them-
selves within gender groups (e.g., Biernat & Manis, 1994). Over-
all, however, the competencies model expects that gender differ-
ences in self-esteem in specific domains will reflect the gender
differences in performance within those domains.
In some cases, the reflected appraisals and competencies models
can compliment each other and work together. For example, peo-
ple may appraise a woman’s athletic ability differently because
they realize that on average, men’s strength and speed are greater
(applying shifting standards: Biernat & Manis, 1994). Similarly,
people might appraise a good athlete as such because he truly is
competent in that area. Thus, in some cases the two models will
make similar predictions. In others, however, they will predict
different results because the appraisal and the competence of males
versus females in certain areas are discrepant—that is, the view of
others and actual competence do not agree.
This meta-analysis includes studies that administered one of
the four most widely used scales of multifaceted self-esteem:
the Harter Self-Perception Profile (Harter), the Self-Description
Questionnaire (SDQ), the Tennessee Self-Concept Scale
(TSCS), or the Piers-Harris Self-Concept Scale (Piers-Harris).
Among them, these scales tap 10 domains of self-esteem:
physical appearance, athletic, academic, social acceptance,
family, behavioral conduct, affect, personal self, self-
satisfaction, and moral-ethical. We review previous research on
gender differences in each of these domains and outline what
each theory (reflected appraisals or competencies) would pre-
dict for gender differences in that area.
Physical Appearance
There is no objective difference in attractiveness between the
genders, thus the competencies model would predict no gender
differences. Competencies might even predict a female advantage,
as—at least in Western societies— both men and women prefer to
look at female bodies rather than male bodies, and women focus
more on their appearance. Appearance is more central to girls’
self-esteem than boys’, with body image a stronger predictor of
global self-esteem in females compared with males (Allgood-
Merten, Lewinsohn, & Hops, 1990; Polce-Lynch, Myers, Kilmar-
tin, Forssmann-Falck, & Kliewer, 1998).
However, the reflected appraisals model makes the opposite
prediction. It is often more difficult for girls and women to feel
positive about their appearance because of media messages pro-
moting extremely high standards for female appearance (e.g.,
Fredrickson & Roberts, 1997; Harter, 1993). This includes the
pressure to be thin, which comes not only from the media but also
from family and friends (Wertheim, Paxton, Shultz, & Muir,
1997). In other words, reflected appraisals create a self-esteem
deficit in appearance self-esteem among females. Critical reflected
appraisals may be one of the sources for the perception that girls’
self-esteem decreases during adolescence: Girls’ body satisfaction
and perceptions of their attractiveness decrease during the teen
years, while boys’ increase or remain the same (Hargreaves &
Tiggemann, 2002; Harter, 1990, 1993). This dissatisfaction con-
tinues into adulthood. Tiggemann and Rothblum (1997) found that
the majority of women rated themselves as overweight and said
they wanted to weigh approximately nine pounds less. In contrast,
men rated themselves as average weight and wanted to weigh
slightly more. Women also overestimate male preferences for
slender female bodies; men’s ideal figure is heavier than what
women believed men’s ideal was (Forbes, Adams-Curtis, Rade, &
Jaberg, 2001). More than 90% anorexia and bulimia sufferers are
female (American Psychiatric Association, 2000). Overall, the
high societal standards for female appearance mean that many
women will not attain the promoted ideal. Thus, it is likely that
females will have lower appearance self-esteem than males.
Athletics
Although girls are now more likely to participate in sports than
they once were, athletic activity is still more emphasized for boys.
Even among children who play sports, boys have higher athletic
self-esteem (e.g., Eccles, Wigfield, Harold, & Blumenfeld, 1993;
Jacobs, Lanza, Osgood, Eccles, & Wigfield, 2002). Boys also
believe they are more competent in athletics (Bosacki, 2003;
Klomsten, Skaalvik, & Espnes, 2004; Wu & Smith, 1997). An
assessment of the aspects of physical self-concept found that boys
scored higher than girls in eight of nine subdomains, including
appearance, body fat, coordination, and endurance (Klomsten et
al., 2004). The gender gap in athletic self-esteem widens during
adolescence. Although physical self-concept decreases from ele-
mentary to secondary school for both genders, the drop is more
pronounced for girls (Klomsten et al., 2004). Both theoretical
models suggest that males will score higher than females on
measures of athletic self-esteem; the competencies model favors
males for their measurably higher performance in most athletic
35GENDER DIFFRENCES IN DOMAIN-SPECIFIC SELF-ESTEEM

domains, and reflected appraisals value these skills more in males
than in females.
At first, it may seem paradoxical that males are held to higher
standards in athletics and females in physical appearance, yet we
predict higher self-esteem for males in athletics and lower self-
esteem for females in physical appearance. By this assumption,
self-esteem should be low when expectations are high. However,
athletics and physical appearance, though both involving the body,
have a crucial difference: Athletics is about doing, and physical
appearance is about being looked at (Fredrickson & Roberts,
1997). One is active (and thus may led to self-esteem and good
mental health) and the other is passive, and thus may lead to
lowered self-esteem through rumination and eventual depression
(Nolen-Hoeksema, Larson, & Grayson, 1999). Especially for ad-
olescents, athletic standards are also more attainable than the more
unrealistic standards set for appearance (Fredrickson & Roberts,
1997).
Academics
Females perform better academically and receive better grades
than their male peers (Pomerantz, Altermatt, & Saxon, 2002;
Stetsenko, Little, Gordeeva, Grasshof, & Oetlingen, 2000). How-
ever, this does not always result in girls and women having higher
academic self-esteem. When males outperform females in academ-
ics, female self-esteem suffers, but when females perform at a
higher level, their self-esteem does not increase (Eccles et al.,
1993; Hyde, Fennema, & Lamon, 1990; Jacobs et al., 2002;
Stetsenko et al., 2000; Weiss, Kemmler, Deisenhammer, Fleis-
chhacker, & Delazer, 2003). Even among academically gifted
adolescents, females are more self-critical of their abilities (Lus-
combe & Riley, 2001). This may occur because teachers give less
positive and more negative feedback to girls than to boys (Dweck &
Leggett, 1988; Eccles & Blumenfeld, 1985). Thus, the competencies
model predicts a female advantage in academic self-esteem, but the
reflected appraisals model predicts a male advantage or no gender
difference.
Social Acceptance
Friendships, peer relationships, and social approval are impor-
tant for self-esteem (Leary & Downs, 1995). People with higher
self-esteem have closer personal relationships (Lundgren &
Rudawsky, 1998). However, high self-esteem may have more
bearing on females’ relationships. Josephs, Markus, and Tafarodi
(1992) found that men, but not women, with high self-esteem
differentiated themselves from others. As children, girls play in
smaller groups and maintain more emotionally intimate and less
competitive relationships than boys (Lever, 1978). In adolescence,
females use more expressive pathways to increase intimacy with
friends, whereas males use expressive and instrumental pathways
equally (Radmacher & Azmitia, 2006). During adulthood, women
report that their friendships are more focused on sharing informa-
tion and communicating than men’s friendships are (Sheets &
Lugar, 2005).
However, girls’ social relationships can also be problematic.
Werner and Crick (2004) showed that girls reacted to social
rejection by retaliating; boys did not. This creates an element of
fragility in females’ friendships. Benenson and Christakos (2003)
found that across elementary, middle school, and high school
samples, girls had more previous friendships that had ended, and
current friendships that were shorter than those of boys. When
asked to imagine their closest friendships ending, girls believed
they would be more distressed. In adulthood, women are more
likely than men to respond to rejection cues by becoming more
self-critical (Baldwin, Granzberg, Pippus, & Pritchard, 2003).
Given that women focus on maintaining relationships, the compe-
tencies model would predict that they would have higher self-
esteem in this area. However, the reflected appraisals model sug-
gests that this advantage may be eliminated by the critical nature
of some of these relationships. Thus, gender differences in social
acceptance self-esteem may be influenced by opposing forces, with
relationships causing both higher and lower self-esteem in girls
and women. These influences might well cancel each other out,
leading to null or small gender differences in social acceptance
self-esteem.
Family
The family can act as a source of support and help affirm a
child’s beliefs about his or her self-worth. Familial relationships
have a significant impact on female self-esteem and levels of
depression, a result not seen in males (Colarossi & Eccles, 2000).
However, this influence can have both positive and negative ram-
ifications. Girls with strong familial relationships have higher
levels of self-esteem and lower levels of depression; however,
parents gave girls more negative feedback than they give boys
(Lundgren & Rudawsky, 1998) even when their performances
were the same (Lewis, Allesandri, & Sullivan, 1992). When con-
sidered together, this creates a detrimental combination for girls’
self-esteem if they are simultaneously highly influenced and highly
criticized. Reflected appraisals would predict a male advantage on
this scale; because there are no established gender differences in
closeness to family, the competencies model predicts no differ-
ence. Given this mixed picture, we predict null or very small
gender differences in family self-esteem.
Behavioral Conduct
Behavioral conduct self-esteem measures an individual’s per-
ception of how socially acceptable his or her behavior is. People
with high behavioral conduct self-esteem view their behavior as
appropriate (Haynes, 1990). Behavioral conduct is particularly
important in school. Haynes (1990) found that behavioral conduct
self-esteem was a significant predictor of classroom behavior,
group participation, and attitudes toward authority in middle
school children. Girls generally perceive themselves as better
behaved than boys do (e.g., Bosacki, 2003; Cole et al., 2001; Wu
& Smith, 1997). Boys’ problems tend to manifest in the form of
externalizing disorders and girls’ in internalizing disorders
(Wicks-Nelson & Israel, 2003). Perhaps as a result, boys are more
likely to be punished or reprimanded because of their behavior.
Both competencies and reflected appraisals predict a female ad-
vantage, as girls and women have less problematic behavioral
conduct and are recognized for this good behavior by others. Thus,
the gender difference in behavior self-esteem is likely to favor girls
and women.
36 GENTILE ET AL.

Affect
Feeling happy, satisfied, and free from anxiety are all elements
of emotional well-being. In contrast, negative emotions are related
to depression and low self-esteem. In childhood, girls and boys are
equally likely be depressed, but in early adolescence female de-
pression rates begin to rise and by late adolescence females are
twice as likely as males to be depressed (Nolen-Hoeksema, 2001;
Twenge & Nolen-Hoeksema, 2002). This incidence rate remains
high throughout adulthood. This large gender difference has been
attributed to many sources, including chronic strain, rumination,
and unequal status at home and work (Nolen-Hoeksema, 1999).
Because women experience more sexist incidents than men, they
were subsequently more likely to report feeling angry, anxious,
and depressed (Swim, Hyers, Cohen, & Ferguson, 2001). Further-
more, being a member of a devalued social group increases one’s
risk for emotional distress (Katz, Joiner, & Kwon, 2002). Women
are also more likely to be anxious. Two meta-analyses found that
women score higher in anxiety than men do, with an effect size
around d ⫽ 0.24 (Feingold, 1994; Twenge, 2000). However, the
gender difference in positive affect also favors women. When only
positive emotions are measured, women report slightly greater
happiness and life satisfaction than men (Wood, Rhodes, &
Whelan, 1989). Women are also more gregarious, trusting, and
tender-minded than men (Feingold, 1994). Thus, the gender differ-
ence in affect self-esteem might be null, because women’s higher
negative affect may be cancelled out by their higher positive affect.
Both competencies and reflected appraisals recognize the gender
difference in emotionality in both positive and negative affect and are
thus in agreement with the prediction for no or only small gender
differences.
Personal Self
Personal self-concept is a measure of one’s evaluation of per-
sonality apart from the physical body or relationship to others
(Haynes, 1990). Research on personal self-esteem is limited and
conflicting. One study found that females score significantly higher
than males (Gadzella & Williamson, 1984); however, personal
self-esteem overlaps with global self-esteem, which usually favors
males. This suggests that the gender difference will slightly favor
males. Reflected appraisals also suggests a male advantage, be-
cause men are usually seen as more autonomous and assertive; the
competencies model predicts no differences, because gender dif-
ferences in agentic traits have disappeared (Twenge, 1997, 2001).
Self-Satisfaction
Self-satisfaction is also a measure of happiness with oneself as
a person (Stringer, Reynolds, & Simpson, 2003). Women have
been shown to have higher life satisfaction than men (Wood et al.,
1989), but little research has investigated self-satisfaction. One
study found that women score significantly higher on this subscale
(Gadzella & Williamson, 1984). On the other hand, self-
satisfaction, like personal self-esteem, overlaps considerably with
global self-esteem. This suggests that the gender difference will
slightly favor males. Similar to personal self, reflected appraisals
would favor males, whereas competencies would predict no dif-
ferences.
Moral–Ethical Self-Concept
Moral– ethical self-concept is a measure of one’s perceptions of
moral– ethical attributes and satisfaction with one’s religion or lack
of it (Haynes, 1990). Among Christian populations, women are
more religious than men on every measure of religiosity (Walter &
Davie, 1998). However, there is little gender difference in religi-
osity among Hindu, Jewish, or Muslim adherents (Loewenthal,
MacLeod, & Cinnirella, 2002).
Women are also more likely to focus on caring for others when
faced with moral dilemmas (Gilligan & Attanucci, 1988; Wark &
Krebs, 1996). Females also exhibit more moral maturity than males
(Wark & Krebs, 1996); however, no gender differences have been
found for moral judgment stage scores (Lifton, 1985). Both compe-
tencies and reflected appraisals recognize that females value this area
more. Thus, we expect that the gender difference in moral– ethical
self-esteem will favor girls and women.
Overview
This meta-analysis aims to determine the size and significance of
gender differences in specific domains of self-esteem. The small
gender differences found in global self-esteem (Kling et al., 1999;
Major et al., 1999; Twenge & Campbell, 2001) may be masking
larger differences among specific domains. It is also not clear which
specific facets favor men and which favor women, and how large any
differences are. There may also be significant moderators of the
effects. The current meta-analysis examines gender differences in
self-esteem based on appearance, athletics, academics, social accep-
tance, family, behavioral conduct, affect, personal self, self-
satisfaction, and moral-ethical self.
Moderator Variables
The meta-analytic design also allows us to examine moderator
variables of gender differences in specific self-esteem. For example,
does the gender difference in appearance self-esteem grow larger
during adolescence? We can also examine changes over time to see
whether gender differences in specific self-concept have changed
from the 1970s to the present. In addition, because we are using
several measures of specific self-concept, we can also examine how
consistent gender differences are across the four measures.
Method
Measures of Self-Esteem
We focused on studies that used the four most widely used
scales of multifaceted self-esteem: the Harter, the SDQ, the TSCS,
and the Piers-Harris, each of which demonstrates high reliability
and validity (Blascovich & Tomaka, 1991). Of course, reliability
and validity differs by population. These scales appear to be
equally valid for males and females, although they differ in age
group targeted. The Harter, SDQ, and Piers-Harris scales are
intended for children and are more valid for that age group,
whereas the TSCS were intended for adolescents and adults (Blas-
covich & Tomaka, 1991). These scales are the most popular
measures of specific self-esteem and thus have the most samples
available for analysis. Of note, each of these scales assesses
37GENDER DIFFRENCES IN DOMAIN-SPECIFIC SELF-ESTEEM

physical appearance self-esteem, which we hypothesize to be one
of the most important facets for gender differences.
Three of these scales (the SDQ, the TSCS, and the Piers-Harris)
use a similar format, with respondents asked to note if they agree
or disagree with simple first-person statements such as “I am
good-looking,” “I am well-behaved in school,” “I am a member of
a happy family,” “I am a good athlete,” or “I am satisfied with my
moral behavior.” The Harter, intended for young children, instead
uses pictures of children performing various activities— one who
is good at the activity (“This boy isn’t very good at numbers”) and
one who isn’t (“This boy is pretty good at numbers”). Children are
then asked if they are good at that activity (“Are you: Not too good
at numbers OR sort of good, pretty good, OR really good at
numbers”).
Locating Studies
We searched the Web of Science by Thompson Scientific (previ-
ously known as the Science and Social Sciences Citation Index), a
comprehensive database that includes both major and minor journals
from all fields in science and social science. We searched for articles
that cited the original sources for the four scales (Fitts, 1965; Harter,
1985; Marsh & O’Neill, 1984; Marsh, Relich, & Smith, 1983;
Marsh, Smith, & Barnes, 1983; Piers, 1969, 1984; Piers & Harris,
1969; Roid & Fitts, 1988).
We also conducted a search for dissertations using the database
ProQuest Dissertations and Theses. We searched for the keywords
“Harter Self-Perception Profile,” “Self-Description Question-
naire,” “Tennessee Self-Concept Scale,” and “Piers-Harris.” An
initial search yielded a prohibitively high number of dissertations.
We examined a subsample of the available dissertations, locating
those published in 1970, 1975, 1980, 1985, 1990, 1995, 2000, and
2005.
We included in the meta-analysis studies that met the fol-
lowing criteria: (1) the study did not preselect participants on
any relevant variable (e.g., extreme depression scores, at-risk
status, alcoholics, learning-disabled students, etc.); (2) the study
involved at least 10 participants; and (3) the authors reported the
domain-specific self-esteem scores broken down by gender. Fi-
nally, studies had to report data in such a way that an effect size
could be calculated (see following section on the calculation of
effect sizes).
Coding of Studies
Studies were coded for the following variables: (a) age or age
group of respondents, (b) year of data collection, coded as two
years prior to the date of publication unless another year was noted
in the article (following Oliver & Hyde, 1993), (c) ethnicity/racial
breakdown of the sample, and (d) measure of self-esteem used.
Not enough studies reported data on socioeconomic status to code
this variable. Most studies reported region, but there were not
enough studies conducted outside the United States to analyze this
variable. Studies reporting statistics separately for different age
groups, self-esteem measures, or ethnic groups were coded as
separate data points.
Final Sample of Studies
The search and review procedures led to a final sample of 115
articles and dissertations. These studies included 32,486 partici-
pants and yielded 428 effect sizes (most studies included multiple
effect sizes, as the self-esteem measures all include multiple do-
mains of self-esteem.) A table with a list of the studies included in
the meta-analysis and their reference list can be obtained from
either the first or second author.
Calculation of Effect Sizes
Formulas for the effect size, d, and homogeneity tests were
taken from Hedges and Becker (1986). When means and SDs were
available, the effect size was computed as the mean self-esteem
score for males minus the mean self-esteem score for females,
divided by the pooled SD. When means and SDs were not avail-
able, the effect size was calculated from reported t or F tests. When
t or F was reported, d was calculated by using the formula
provided by Hedges and Becker (1986).
Because effect sizes tend to be upwardly biased when based on
small sample sizes, effect sizes were corrected for bias in the
estimation of population effect sizes using the formula provided by
Hedges (1981). All effect size analyses were weighted analyses
(i.e., each effect size was weighted by an inverted variance;
Hedges & Versa, 1998; Lipsey & Wilson, 2001).
To conduct the meta-analyses, we used mixed-effects models,
which assume that effect size variance can be explained by both
systematic and random components (Lipsey & Wilson, 2001).
In mixed-effects models, certain identifiable study characteris-
tics may act as moderator variables that are associated with
systematic differences among effect sizes at the same time that
a random component of residual variance remains after the
systematic portion is accounted for. The mixed-effects model is
preferable in this case because a fixed-effects model assumes
that the only source of variation is from systematic variation,
and the random effects model assumes none of the variation is
from systematic sources. Mixed-effects models assume the ef-
fects of between-study variables are systematic, but that there is
a remaining unmeasured random effect in the effect size distri-
bution in addition to sampling error. As is done in random
effects models, a random effects variance component (derived
from the residual homogeneity value after the moderators are
taken into account) is estimated and added to the standard error
associated with each effect size and inverted variance weights
are calculated.
Results
Mean Effect Sizes
Mean effect sizes were calculated for each of the self-esteem
domains. The results are reported in Table 1. The number of
samples (k), the weighted d, the 95% confidence interval for d, and
the total homogeneity statistic (Q
T
) for each self-esteem domain
are reported. To address outlier data points, we eliminated effect
sizes more than 2 SDs above or below the mean (Lipsey & Wilson,
2001). Out of 428 effect sizes, this procedure identified 20 total
outliers that were eliminated from further analyses (4 for physical
38 GENTILE ET AL.

appearance, 3 for athletic, 3 for academic, 2 for social accep-
tance, 1 for family, 4 for behavioral conduct, 1 for affect, 1 for
personal self, and 1 for self-satisfaction).
Significant mean effect sizes ranged from ⫺0.17 to 0.41 (see
Table 1). As expected, males scored higher than females on
physical appearance, athletic, personal, and self-satisfaction self-
esteem, but females scored significantly higher than males on
behavioral conduct and moral– ethical self-esteem. Academic, af-
fect, social acceptance, and family self-esteem did not show signifi-
cant gender differences, because their 95% confidence intervals in-
cluded zero (or nearly included zero, in the case of affect). Table 2
compares the predictions from the reflected appraisals and com-
petencies models with the results of the analyses. Overall, the
reflected appraisals model receives more support.
Only six out of the 10 domains included enough datapoints from
dissertations to compare unpublished versus published data. Of
these, four showed no significant differences by publication status.
Unpublished dissertations produced a higher effect size (d ⫽ 0.72,
k ⫽ 4) than published journal articles (d ⫽ 0.33, k ⫽ 72),
␹
2
(2) ⫽ 16.02, p ⬍ .001, for physical appearance, and dissertations
showed a higher effect size (d ⫽ 0.70, k ⫽ 3) than published
articles (d ⫽ 0.18, k ⫽ 7), ␹
2
(2) ⫽ 8.91, p ⬍ .05 for self-
satisfaction. Thus, there were few differences by publication sta-
tus, and the higher effect sizes for dissertations argue against the
possibility of a file-drawer problem.
Moderator Analyses
Physical appearance. Homogeneity analyses using procedures
specified by Hedges and Becker (1986) and Lipsey and Wilson (2001)
indicated that the set of 76 effect sizes was significantly heteroge-
neous, Q
T
⫽ 288.07, ␹
2
(75) ⫽ 118.60, p ⬍ .001. The significant
between-groups homogeneity statistics for age, ␹
2
(4) ⫽ 13.28, p ⬍
.01, data collection year ␹
2
(3) ⫽ 16.27, p ⬍ .001, and measure,
␹
2
(3) ⫽ 7.82, p ⬍ .05 suggest that there is a significant difference in
the magnitude of the effect sizes as a function of these moderator
variables. As can be seen from Table 3, gender differences in physical
appearance self-esteem were largest during adulthood. However, that
effect size was based on only two samples. Of the age groups with
more samples, the largest gender difference in appearance self-esteem
occurred during early adolescence, consistent with the idea that the
standards of others for a sexualized adult female appearance is at the
root of the difference (Fredrickson & Roberts, 1997).
Time period also moderated the effect; there was not a significant
gender difference during the 1970s, but beginning in the 1980s
women scored significantly lower than men in physical appearance
self-esteem. There are also slight differences based on measure, with
the smallest differences on the Harter and the largest on the TSCS.
Athletic. Homogeneity analyses showed that the set of 68
effect sizes was significantly heterogeneous. As shown in Table 4,
measure is the only significant moderator variable, with effect
sizes for the Harter and the TSCS moderate and comparable but
the effect size for the SDQ near zero.
Academic. The set of 75 effect sizes was significantly heteroge-
neous, Q
T
⫽ 155.60, ␹
2
(74) ⫽ 117.35, p ⬍ .001. As shown in Table
5, age is a significant moderator. While there are no significant
differences between males and females on academic self-esteem at
most ages, a small difference emerges during adulthood favoring
females.
Social acceptance. Homogeneity analyses indicated that the
set of 81 effect sizes was significantly heterogeneous, Q
T
⫽
109.84, ␹
2
(80) ⫽ 101.88, p ⬍ .05. As shown in Table 6
, year was
a significant moderator, with the effect closer to zero during the
1980s and very small during the other eras.
Family. Homogeneity analyses indicated that the set of 21
effect sizes was not heterogeneous, Q
T
⫽ 20.41, ␹
2
(20) ⫽
Table 1
Summary of Main Effects for Gender Differences in Domain-
Specific Self-Esteem
Self-esteem domain kd 95% CI Q
T
Physical appearance 76 .35 .31, .40 288.07
ⴱⴱⴱ
Athletic 68 .41 .36, .46 322.33
ⴱⴱⴱ
Academic 75 .04 .00, .08 155.60
ⴱⴱⴱ
Social acceptance 81 .04 ⫺.02, .10 109.84
ⴱ
Family 21 ⫺.02 ⫺.07, .04 20.41
Behavioral conduct 56 ⫺.17 ⫺.28, ⫺.06 88.85
ⴱⴱ
Affect 17 .11 .04, .19 19.18
Personal self 9 .28 .11, .45 23.82
ⴱⴱ
Self-satisfaction 10 .33 .18, .49 58.78
ⴱⴱⴱ
Moral–ethical 15 ⫺.38 ⫺.48, ⫺.29 59.47
ⴱⴱⴱ
Note. k ⫽ number of studies; d ⫽ effect size; CI ⫽ confidence interval;
Q
T
⫽ total homogeneity. A positive d indicates that males scored higher;
a negative d indicates that females scored higher.
ⴱ
p ⬍ .05.
ⴱⴱ
p ⬍ .01.
ⴱⴱⴱ
p ⬍ .001.
Table 2
Comparison of the Model Predictions and Actual Outcomes
Self-esteem domain
Reflected appraisals model
predictions
Competencies model
predictions Actual findings
Physical appearance Males higher Females higher/no difference Males higher
Athletic Males higher Males higher Males higher
Academic Males higher/no difference Females higher No difference
Social acceptance No difference Females higher No difference
Family Males higher No difference No difference
Behavioral conduct Females higher Females higher Females higher
Affect No difference No difference No difference
Personal Males higher No difference Males higher
Self-satisfaction Males higher No difference Males higher
Moral–ethical Females higher Females higher Females higher
39GENDER DIFFRENCES IN DOMAIN-SPECIFIC SELF-ESTEEM

31.41, p ⬍ .05. Thus, no moderator variable analyses were
conducted.
Behavioral conduct. The set of 56 effect sizes was significantly
heterogeneous, Q
T
⫽ 88.85, ␹
2
(55) ⫽ 82.29, p ⬍ .01. As shown in
Table 7, age is a significant moderator, with the female advan-
tage in behavioral conduct self-esteem increasing as youth get
older. Year is also a moderator, with the female advantage
increasing until the 1990s and then reversing in the 2000s.
Affect. Homogeneity analyses indicated that the set of 17 effect
sizes was not heterogeneous, Q
T
⫽ 19.18, ␹
2
(16) ⫽ 26.30, p ⬍ .05.
Thus, no moderator variable analyses were conducted.
Personal self. Homogeneity analyses indicated that the set of 9
effect sizes was significantly heterogeneous, Q
T
⫽ 23.82,
␹
2
(8) ⫽ 20.09, p ⬍ .01. However, none of the moderator variable
analyses reached significance.
Self-satisfaction. The set of 10 effect sizes was significantly
heterogeneous, Q
T
⫽ 58.78, ␹
2
(9) ⫽ 27.88, p ⬍ .001. Not all
groups had 2 or more studies, which restricted moderator analyses.
It appears that gender differences in self-satisfaction self-esteem
are largest during high school versus college and are larger in the
most recent decade in comparison to the 1970s (see Table 8).
Moral– ethical. Homogeneity analyses indicated that the set
of 15 effect sizes was significantly heterogeneous, Q
T
⫽ 59.47,
␹
2
(14) ⫽ 36.12, p ⬍ .001. The female advantage in moral-ethical
self-esteem gets increasingly larger over time and is larger in
studies that employ the SDQ rather than the Piers-Harris (see
Table 9).
Overall. If we make the assumption that global self-esteem
equals the sum of all domains of specific self-esteem, we can
estimate the gender difference for global self-esteem here as
d ⫽ 0.10. This is smaller than the effect size for measures of global
self-esteem (d ⫽ 0.15: Kling et al., 1999), perhaps because appear-
ance plays a disproportionate role in determining global self-esteem
(Harter, 1993), especially among younger people. Thus, this estimate
must be considered only a rough approximation, as it does not weight
the domains of self-esteem.
Discussion
This study used meta-analytic techniques to investigate 428
effect sizes of gender differences in specific domains of self-
esteem. Males scored significantly higher than females on physical
appearance (d ⫽ 0.35), athletic (d ⫽ 0.41), personal self
(d ⫽ 0.28), and self-satisfaction (d ⫽ 0.33) self-esteem. Females
scored higher than males on behavioral conduct (d ⫽⫺0.17) and
moral– ethical self-esteem (d ⫽⫺0.38). No significant gender dif-
Table 3
Moderating Variables in Physical Appearance Self-Esteem
Variable and class Between-groups Hk d 95% CI for d Within-group H
Age group 17.13
ⴱⴱ
Elementary school (ages 5–10) 24 0.30 0.22, 0.38 102.92
ⴱⴱⴱ
Junior high (ages 11–13) 25 0.41 0.33, 0.49 57.47
ⴱⴱⴱ
High school (ages 14–17) 14 0.30 0.19, 0.42 59.71
ⴱⴱⴱ
College (ages 18–22) 8 0.25 0.10, 0.40 17.53
ⴱ
Adult (ages 23–58) 5 0.73 0.51, 0.96 33.31
ⴱⴱⴱ
Year 32.14
ⴱⴱⴱ
1970–1979 8 0.09 ⫺0.09, 0.28 14.39
ⴱ
1980–1989 22 0.33 0.25, 0.41 50.24
ⴱⴱⴱ
1990–1999 34 0.38 0.31, 0.46 70.74
ⴱⴱⴱ
2000–2006 11 0.32 0.20, 0.45 120.54
ⴱⴱⴱ
Self-Esteem measure 9.15
ⴱ
Piers-Harris 10 0.39 0.24, 0.54 41.69
ⴱⴱⴱ
Self-Description Questionnaire 13 0.33 0.23, 0.43 61.14
ⴱⴱⴱ
Harter Self-Perception Profile 15 0.21 0.09, 0.32 70.09
ⴱⴱⴱ
Tennessee Self-Concept Scale 38 0.41 0.34, .047 106.00
Note. k ⫽ number of studies; d ⫽ difference in terms of SDs; CI ⫽ confidence interval; H ⫽ homogeneity. A positive d indicates that males scored higher;
a negative d indicates that females scored higher.
ⴱ
p ⬍ .05.
ⴱⴱ
p ⬍ .01.
ⴱⴱⴱ
p ⬍ .001.
Table 4
Moderating Variables in Athletic Self-Esteem
Variable and class Between-groups Hk d 95% CI for d Within-group H
Self-Esteem measure 37.60
ⴱⴱⴱ
Self-Description Questionnaire 12 0.06 ⫺0.05, 0.17 48.01
ⴱⴱⴱ
Harter Self-Perception Profile 12 0.58 0.45, 0.71 70.49
ⴱⴱⴱ
Tennessee Self-Concept Scale 44 0.48 0.41, 0.54 166.23
ⴱⴱⴱ
Note. k ⫽ number of studies; d ⫽ difference in terms of SDs; CI ⫽ confidence interval; H ⫽ homogeneity. A positive d indicates that males scored higher;
a negative d indicates that females scored higher.
***
p ⬍ .001.
40 GENTILE ET AL.

ferences appeared in academic, social acceptance, family, and affect
self-esteem. Most of the significant gender differences are moderate in
size (between .20 and .40 SDs). Nevertheless, the differences in
appearance, athletic, self-satisfaction, and moral– ethical self-
esteem are larger than the established gender differences in verbal
abilities, empathy, and adult aggression (Ashmore, 1990) and the
White-Asian difference in self-esteem (Twenge & Crocker, 2002).
Thus, the results challenge the idea that self-esteem differences
between men and women are small. It is more accurate to say that
the difference varies depending on the specific domain. Within
some domains, such as physical appearance, athletic, and moral-
ethical self-esteem, the gender differences are more than double
those found with general self-esteem measures.
Overall, the results favor the reflected appraisals model. Both
models made correct predictions in 4 domains (athletic, behavioral
conduct, affect, moral-ethical); competencies were correct and
reflected appraisals wrong in 1 (family); and reflected appraisals
were right and competencies wrong in 5 (physical appearance,
academics, social acceptance, personal, and self- satisfaction). If
people evaluated their abilities based on objective competence,
males would score higher on athletic self-esteem, females would
score higher on academic, social acceptance, behavioral conduct,
and moral-ethical self-esteem, and there would be no gender
differences in the other domains. However, one of the largest gender
differences was in physical appearance, a domain where competen-
cies predicted no differences or even a female advantage but re-
flected appraisals correctly predicted a considerable male advan-
tage. Competences and reflected appraisals apparently canceled
each other out in the domains of academics and social acceptance.
The male advantage in physical appearance self-esteem was
significant at all ages but was most pronounced during adulthood.
The gender gap did not increase consistently from childhood to
adulthood, but rather increased from childhood to junior high, then
decreased throughout high school and college before rising again
in adulthood. These results are consistent with research showing
that female body satisfaction decreases during adolescence while
males’ stabilizes or increases (Hargreaves & Tiggemann, 2002;
Harter, 1990, 1993). It is also consistent with studies showing that
female body dissatisfaction persists during adulthood (Forbes et
al., 2001; Tiggemann & Rothblum, 1997). Prior to adulthood, the
largest gender gap occurs during junior high school. This may
reflect the fact that females begin puberty earlier than males and
thus may show more concern over their development compared
with their male peers, whose development is more delayed. This is
consistent with the traditional theory that lower self-esteem in
females is related to the physical changes of puberty (Rosenberg,
1986).
The gender difference in appearance self-esteem was not sig-
nificantly different from zero during the 1970s. After 1980, the
difference rose to about a third of a SD and stayed there. This may
have been caused by the increasing media focus on appearance
during the 1980s and afterward. As Fredrickson and Roberts’
(1997) objectification theory would predict, this may have led to
lowered appearance self-esteem for women through reflected ap-
praisals.
There were no significant gender differences on academic self-
esteem. Given that females show higher academic performance,
this is consistent with research showing that females may discount
their academic abilities even when they excel (Eccles et al., 1993;
Hyde et al., 1990; Jacobs et al., 2002; Pomerantz et al., 2002;
Stetsenko et al., 2000; Weiss et al., 2003). In this case, actual
competencies are washed out by reflected (or perhaps self-) ap-
praisals.
Table 5
Moderating Variables in Academic Self-Esteem
Variable and class Between-groups Hk d 95% CI for d Within-group H
Age group 10.25
ⴱ
Elementary school (ages 5–10) 30 0.10 0.04, 0.16 74.92
ⴱⴱⴱ
Junior high (ages 11–13) 30 0.01 ⫺0.05, 0.07 50.26
ⴱ
High school (ages 14–17) 8 0.00 ⫺0.15, 0.15 1.91
College (ages 18–22) 5 ⫺0.07 ⫺0.24, 0.10 17.01
ⴱⴱ
Adult (ages 23–58) 2 ⫺0.21 ⫺0.45, 0.04 1.25
Note. k ⫽ number of studies; d ⫽ difference in terms of SDs; CI ⫽ confidence interval; H ⫽ homogeneity. A positive d indicates that males scored higher;
a negative d indicates that females scored higher.
ⴱ
p ⬍ .05.
ⴱⴱ
p ⬍ .01.
ⴱⴱⴱ
p ⬍ .001.
Table 6
Moderating Variables in Social Acceptance Self-Esteem
Variable and class Between-groups Hk d 95% CI for d Within-group H
Year 16.51
ⴱⴱⴱ
1970–1979 8 0.12 ⫺0.11, 0.34 7.85
1980–1989 19 ⫺0.01 ⫺0.13, 0.11 20.28
1990–1999 36 0.07 ⫺0.03, 0.16 34.16
2000–2006 17 0.09 ⫺0.04, 0.22 31.04
Note. k ⫽ number of studies; d ⫽ difference in terms of SDs; CI ⫽ confidence interval; H ⫽ homogeneity. A positive d indicates that males scored higher;
a negative d indicates that females scored higher.
***
p ⬍ .001.
41GENDER DIFFRENCES IN DOMAIN-SPECIFIC SELF-ESTEEM

Females scored significantly higher on behavioral conduct self-
esteem. This reflects research showing that females behave better
(e.g., Bosacki, 2003; Cole et al., 2001; Wu & Smith, 1997),
whereas males tend to act out more (Wicks-Nelson & Israel,
2003). This difference grows as children grow older, suggesting
that experience in school reinforces these beliefs. Moral– ethical
self-esteem was significantly higher in females, supporting previ-
ous research finding greater religiosity among women (at least in
Christian populations). The gender difference is largest during
high school; thus females may mature faster in their morality than
males (Wark & Krebs, 1996). The difference has grown over the
decades; there was no significant gender difference in the 1970s,
but the effect size now exceeds half a SD.
Males scored significantly higher than females on the personal
self and self-satisfaction subscales. Personal self and self-
satisfaction are similar to global self-esteem (they all measure
happiness with oneself as a person). The differences here are
higher than those in the previous meta-analyses on global self-
esteem, but are very close to the meta-analytic effect sizes for
adolescents, who comprised the majority of the samples on these
domains in this meta-analysis.
Implications
An influential review (Baumeister, Campbell, Kruger, & Vohs,
2003) found that global self-esteem was linked to happiness, but
had few benefits for academic achievement, work performance, or
healthy behaviors. In contrast, domain-specific measures of self-
esteem are consistently correlated with performance within that
domain, apparently in a reciprocal process in which each causes
the other (Marsh & Craven, 2006). These results are explained by
the specificity matching hypothesis, which contends that the most
meaningful linkages are between attitudes and behavior within the
same domain. Nevertheless, many interventions for girls and
women conflate domain-specific self-esteem with general self-
esteem (e.g., the Dove Self-Esteem Fund, which opines that “Too
many girls develop low self-esteem from hang-ups about looks
and, consequently, fail to reach their full potential in later life.”)
Research on global vs. domain-specific self-esteem suggests in-
stead that most girls with “low self-esteem from hangups about
looks” will do just fine in reaching their potential (e.g., academi-
cally), but may continue to have low appearance self-esteem and
might be at risk for eating disorders.
Table 7
Moderating Variables in Behavioral Conduct Self-Esteem
Variable and class Between-groups Hk d 95% CI for d Within-group H
Age group 10.09
ⴱ
Elementary school (ages 5–10) 19 ⫺0.10 ⫺0.29, 0.09 47.13
ⴱⴱⴱ
Junior high (ages 11–13) 26 ⫺0.23 ⫺0.39, ⫺0.08 25.27
High school (ages 14–17) 7 ⫺0.25 ⫺0.55, 0.06 5.62
College (ages 18–22) 3 ⫺0.27 ⫺0.73, 0.20 0.74
Year 14.18
ⴱⴱ
1970⫺1979 7 ⫺0.09 ⫺0.42, 0.24 3.42
1980⫺1989 9 ⫺0.12 ⫺0.37, 0.13 3.34
1990⫺1999 30 ⫺0.30 ⫺0.45, ⫺0.15 32.21
2000⫺2006 9 0.22 ⫺0.05, 0.49 35.70
Note. k ⫽ number of studies; d ⫽ difference in terms of SDs; CI ⫽ confidence interval; H ⫽ homogeneity. A positive d indicates that males scored higher;
a negative d indicates that females scored higher.
ⴱ
p ⬍ .05.
ⴱⴱ
p ⬍ .01.
ⴱⴱⴱ
p ⬍ .001.
Table 8
Moderating Variables in Self-Satisfaction Self-Esteem
Variable and class Between-groups Hk d 95% CI for d Within-group H
Age group 50.32
ⴱⴱⴱ
High school (ages 14–17) 4 0.36 0.13 3.63
0.59
College (ages 18–22) 4 0.05 ⫺0.21 4.83
0.30
Year 14.20
ⴱⴱⴱ
1970–1979 7 0.11 ⫺0.08 6.15
0.31
2000–2006 2 0.81 0.46 38.43
1.16
Note. k ⫽ number of studies; d ⫽ difference in terms of SDs; CI ⫽ confidence interval; H ⫽ homogeneity. A positive d indicates that males scored higher;
a negative d indicates that females scored higher.
***
p ⬍ .001.
42 GENTILE ET AL.

As appearance is an area in which “performance” is difficult to
measure (and looking good may have come at the expense of an
eating disorder) the consequences of low appearance self-esteem
are more difficult to quantify than the consequences of, for exam-
ple, low academic self-esteem. Outcomes linked to low appearance
self-esteem may be fueled by mediators such as objectification
(Fredrickson & Roberts, 1997) and other risk factors for eating
disorders. Thus any interventions designed to help girls and
women avoid the negative consequences of low appearance self-
esteem should be focused specifically on appearance and not on
general self-esteem. An even more expeditious rotue might be to
bypass appearance self-esteem and directly target the root causes
of negative outcomes such as eating disorders. Using the appraisals
model as a basis, such interventions might help girls and women
develop a more objective view of how others see them rather than
relying on media images as universal standards. This may be more
effective than self-esteem boosting efforts that fail to account for
this important social context. “You’re beautiful just for being you”
falls on deaf ears if girls believe that others will see them as
beautiful only if they meet an impossible standard.
Measurement Effects and Limitations
The effect sizes were fairly consistent across the self-esteem
measures used for most of the subscales, with few exceptions. A
large gender difference favoring males was found on the athletic
self-esteem subscale of the Harter and TSCS, compared with a
difference of near zero on the SDQ. However, the SDQ showed a
larger female advantage on moral– ethical self-esteem than the
Piers-Harris. One limitation is that age and measure were con-
founded, because adult samples completed the TSCS and child
samples the other measures. In general, the biggest limitation of
this analysis is the small number of samples from adult popula-
tions; the majority of the data here come from children and
adolescents. Future research should explore how gender differ-
ences in the domains of self-esteem change during the course of
adult life.
Conclusions
The current meta-analysis found that gender differences vary
widely across the subdomains of self-esteem, some showing no
difference at all, and others with gender differences in the mod-
erate range. The differences obtained in this study were moderated
by variables such as age, year of data collection, and measure. Like
the meta-analyses of global self-esteem, these analyses do not find
extremely large gender differences in self-esteem. However, many
of the gender differences in specific domains of self-esteem are
considerably larger than the d ⫽ 0.15 difference in global self-
esteem. In some areas, like that of athletic ability, these gender
differences reflect actual gender differences in competence and
performance. In other areas, such as physical appearance, women’s
lower self-esteem derives not from actual deficits but from the
more critical reflected appraisals of others—including the larger
“other” of idealized media images.
References
Allgood-Merten, B., Lewinsohn, P. M., & Hops, H. (1990). Sex differences
and adolescent depression. Journal of Abnormal Psychology, 99, 55– 63.
American Psychiatric Association. (2000). Diagnostic and statistical man-
ual of mental disorders (4th ed., text revision). Washington, DC: Author.
Ashmore, R. D. (1990). Sex, gender, and the individual. In L. A. Pervin
(Ed.), Handbook of personality: Theory and research (pp. 486 –526).
New York: Guilford Press.
Baldwin, M. W., Granzberg, A., Pippus, L., & Pritchard, E. T. (2003).
Cued activation of relational schemas: Self-evaluation and gender ef-
fects. Canadian Journal of Behavioural Science, 35, 153–163.
Bandura, A. (1989). Regulation of cognitive processes through perceived
self-efficacy. Developmental Psychology, 25, 729–735.
Baumeister, R. F. (1998). The self. In D. T. Gilbert, S. T. Fiske, & G.
Lindzey (Eds.), Handbook of social psychology (4th ed., pp. 680 –740).
New York: McGraw-Hill.
Baumeister, R. F., Campbell, J. D., Krueger, J. I., & Vohs, K. D. (2003).
Does high self-esteem cause better performance, interpersonal success,
happiness, or healthier lifestyles? Psychological Science in the Public
Interest, 4, 1–44.
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for
interpersonal attachments as a fundamental human motivation. Psycho-
logical Bulletin, 117, 497–529.
Benenson, J. F., & Christakos, A. (2003). The greater fragility of females’
versus males’ closest same-sex friendships. Child Development, 74,
1123–1129.
Biernat, M., & Manis, M. (1994). Shifting standards and stereotype-based
judgments. Journal of Personality and Social Psychology, 66, 5–20.
Blascovich, J., & Tomaka, J. (1991). Measures of self-esteem. In J. P.
Robinson, P. R. Shaver, & L. S. Wrightsman (Eds.), Measures of
personality and social psychological attitudes (pp. 115–160). San Di-
ego, CA: Academic Press.
Table 9
Moderating Variables in Moral–Ethical Self-Esteem
Variable and class Between-groups Hk d 95% CI for d Within-group H
Year 44.81
ⴱⴱⴱ
1970–1979 5 0.01 ⫺0.19, 0.21 6.01
1980–1989 5 ⫺0.40 ⫺0.57, ⫺0.24 8.40
1990–1999 3 ⫺0.68 ⫺0.91, ⫺0.46 0.25
Self-esteem measure 12.47
ⴱⴱ
Piers-Harris 6 ⫺0.32 ⫺0.48, ⫺0.17 34.46
ⴱⴱⴱ
Self-Description Questionnaire 8 ⫺0.50 ⫺0.63, ⫺0.37 12.54
Note. k ⫽ number of studies; d ⫽ difference in terms of SDs; CI ⫽ confidence intervals; H ⫽ homogeneity. A positive d indicates that males scored
higher; a negative d indicates that females scored higher.
**
p ⬍ .01.
***
p ⬍ .001.
43GENDER DIFFRENCES IN DOMAIN-SPECIFIC SELF-ESTEEM

Bosacki, S. L. (2003). Psychological pragmatics in preadolescents: So-
ciomoral understanding, self-worth, and school behavior. Journal of
Youth and Adolescents, 32, 141–155.
Colarossi, L. G., & Eccles, J. S. (2000). A prospective study of adoles-
cents’ peer support: Gender differences and the influence of parental
relationships. Journal of Youth and Adolescence, 29, 661–678.
Cole, D. A., Maxwell, S. E., Martin, J. M., Peeke, L. G., Seroczynski,
A. D., Tram, J. M., et al. (2001). The development of multiple domains
of child and adolescent self-concept: A cohort sequential longitudinal
design. Child Development, 72, 1723–1746.
Cooley, C. H. (1902). Human nature and the social order. New York:
Scribner’s.
Denissen, J. J. A., Penke, L., & Schmitt, D. P. (2008). Self-esteem reac-
tions to social interactions: Evidence for sociometer mechanisms across
days, people, and nations. Journal of Personality and Social Psychol-
ogy, 95, 181–196.
Dweck, C. S., & Leggett, E. L. (1988). A social cognitive approach to
motivation and personality. Psychological Review, 95, 256 –273.
Eccles, J., & Wigfield, A., Harold, R. D., & Blumenfeld, P. (1993). Age
and gender differences in children’s self- and task perceptions during
elementary school. Child Development, 64, 830 – 847.
Eccles, J. S., & Blumenfeld, P. (1985). Classroom experiences and student
gender: Are there differences and do they matter? In L. C. Wilkinson &
C. B. Marrett (Eds.), Gender influences in the classroom (pp. 79–113).
New York: Academic Press.
Feingold, A. (1994). Gender differences in personality: A meta-analysis.
Psychological Bulletin, 116(3), 429 – 456.
Fitts, W. H. (1965). Manual for the Tennessee Self Concept Scale. Nash-
ville, TN: Counselor Recordings and Tests.
Forbes, G. B., Adams-Curtis, L. E., Rade, B., & Jaberg, P. (2001). Body
dissatisfaction in women and men: The role of gender-typing and self-
esteem. Sex Roles, 44, 461–484.
Fredrickson, B. L., & Roberts, T. A. (1997). Objectification theory: To-
ward understanding women’s lived experiences and mental health risks.
Psychology of Women Quarterly, 21, 173–206.
Gadzella, B. M., & Williamson, J. D. (1984). Differences between men and
women on selected Tennessee Self-Concept Scales. Psychological Re-
ports, 55, 939–942.
Gilligan, C., & Attanucci, J. (1988). Two moral orientations: Gender
differences and similarities. Merrill-Palmer Quarterly, 34(3), 223–237.
Hargreaves, D., & Tiggemann, M. (2002). The role of appearance sche-
maticity in the development of adolescent body dissatisfaction. Cogni-
tive Therapy and Research, 26, 691–700.
Harter, S. (1985). Manual for the self-perception profile for children.
Denver, CO: University of Denver.
Harter, S. (1990). Causes, correlates and the functional role of global
self-worth: A life Span perspective. In R. J. Sternberg & J. Kolligian, Jr.
(Eds.), Competence considered (pp. 67–98). New Haven, CT: Yale
University Press.
Haynes, N. M. (1990). Influence of self-concept on school adjustment
among middle School students. Journal of Social Psychology, 130,
199 –208.
Hedges, L. V. (1981). Distribution theory for Glass’s estimator of effect
size and related estimators. Journal of Educational Statistics, 6(2),
107–128.
Hedges, L. V., & Becker, B. J. (1986). Statistical methods in the meta-
analysis of research in gender differences. In J. S. Hyde & M. C. Linn
(Eds.), The psychology of gender: Advances through meta-analysis (pp.
14 –50). Baltimore: Johns Hopkins University Press.
Hedges, L. V., & Vevea, J. L. (1998). Fixed- and random-effects models in
meta-analysis. Psychological Methods, 3, 486–504.
Hyde, J. S., Fennema, E., & Lamon, J. (1990). Gender differences in
mathematics performance: A meta-analysis. Psychological Bulletin,
107, 139 –155.
Jacobs, J. E., Lanza, S., Osgood, D. W., Eccles, J. S., & Wigfield, A.
(2002). Changes in children’s self-competence and values: Gender and
domain differences across grades one through twelve. Child Develop-
ment, 73, 509–527.
James, W. (1890). The principles of psychology (Vol. 1). New York: Holt.
Josephs, R. A., Markus, H. R., & Tafarodi, R. W. (1992). Gender and
self-esteem. Journal of Personality and Social Psychology, 63, 391–
402.
Katz, J., Joiner Jr., T. E., & Kwon, P. (2002). Membership in a devalued
social group and emotional well-being: Developing a model of personal
self-esteem, collective self-esteem, and group socialization. Sex
Roles, 47, 419– 431.
Kling, K. C., Hyde, J. S., Showers, C. J., & Buswell, B. N. (1999). Gender
differences in self-esteem: A meta-analysis. Psychological Bulletin, 125,
470 –500.
Klomsten, A. T., Skaalvik, E. M., & Espnes, G. A. (2004). Physical
self-concept and sports: Do gender differences still exist? Sex Roles, 50,
119 –127.
Leary, M. R., & Downs, D. L. (1995). Interpersonal functions of the
self-esteem motive: The self-esteem system as a sociometer. In M. H.
Kernis (Ed.), Efficacy, agency, and self-esteem (pp. 123–144). New
York: Plenum Press.
Leary, M. R., Haupt, A. L., Strausser, K. S., & Chokel, J. T. (1998).
Calibrating the sociometer: The relationship between interpersonal ap-
praisals and the state self-esteem. Journal of Personality and Social
Psychology, 74, 1290–1299.
Leary, M. R., Tambor, E. S., Terdal, S. K., & Downs, D. L. (1995).
Self-esteem as an interpersonal monitor: The sociometer hypothesis.
Journal of Personality Psychology, 68, 518 –530.
Lever, J. (1978). Sex differences in the complexity of children’s play and
games. American Sociological Review, 43, 471– 483.
Lewis, M., Allesandri, S. M., & Sullivan, M. W. (1992). Differences in
shame and pride as a function of children’s gender and task difficulty.
Child Development, 63, 630 – 638.
Lifton, P. D. (1985). Individual differences in moral development: The
relation of sex, gender, and personality to morality. Personality, 53,
306 –334.
Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousand
Oaks, CA: Sage.
Loewenthal, K. M., MacLeod, A. K., & Cinnirella, M. (2002). Are women
more religious than men? Gender differences in religious activity among
different religious groups in the UK. Personality and Individual Differ-
ences, 32, 133–139.
Lundgren, D. C., & Rudawsky, D. J. (1998). Female and male college
students’ responses to negative feedback from parents and peers. Sex
Roles, 39, 409– 429.
Luscombe, A., & Riley, T. L. (2001). An examination of self-concept in
academically gifted adolescents: Do gender differences occur? Roeper
Review, 24, 20–22.
Major, B., Barr, L., Zubek, J., & Babey, S. H. (1999). Gender and
self-esteem: A meta-analysis. In W. B. Swann Jr., J. H. Langlois, &
L. A. Gilbert (Eds.), Sexism and stereotypes in modern society: The
gender science of Janet Taylor Spence (pp. 223–253). Washington, DC:
American Psychological Association.
Marsh, H. W., & Craven, R. G. (2006). Reciprocal effects of self-concept
and performance from a multidimensional perspective: Beyond seduc-
tive pleasure and unidimensional perspectives. Perspectives on Psycho-
logical Science, 1, 133–163.
Marsh, H. W., & O’Neill, R. (1984). Self-description questionnaire III: The
construct validity of multidimensional self-concept ratings by late ado-
lescents. Journal of Educational Measurement, 24, 153–174.
Marsh, H. W., Smith, I. D., & Barnes, J. (1983). Multitrait-multimethod
analyses of the self-description questionnaire: Student-teacher agree-
44 GENTILE ET AL.

ment on multidimensional ratings of student self-concept. American
Education Research Journal, 20, 333–357.
Mead, G. H. (1934). Mind, self, and society. Chicago: University of
Chicago Press.
Nolen-Hoeksema, S. (1999). Gender differences in depression. In C. Ham-
men & I. Gotlib (Eds.), Handbook of depression. NY: Guilford.
Nolen-Hoeksema, S. (2001). Gender differences in depression. Current
Directions in Psychological Science, 10, 173–176.
Nolen-Hoeksema, S., Larson, J., & Grayson, C. (1999). Explaining the
gender difference in depressive symptoms. Journal of Personality and
Social Psychology, 77, 1061–1072.
Oliver, M. B., & Hyde, J. S. (1993). Gender differences in sexuality: A
meta-analysis. Psychological Bulletin, 114, 29–51.
Piers, E. V. (1969). Manual for the Piers-Harris Children’s Self-Concept
Scale. Nashville, TN: Counselor Recordings and Tests.
Piers, E. V. (1984). Revised manual for the Piers-Harries Children’s
Self-Concept Scale. Los Angeles: Western Psychological Services.
Piers, E. V., & Harris, D. B. (1969). Piers-Harris Children’s Self-Concept
Scale. Los Angeles: Western Psychological Services.
Pipher, M. (1994). Reviving Ophelia: Saving the selves of adolescent girls.
New York: Ballantine Books.
Polce-Lynch, M., Myers, B. J., Kilmartin, C. T., Forssmann-Falck, R., &
Kliewer, W. (1998). Gender and age patterns in emotional expression,
body image, and self-esteem: A qualitative analysis. Sex Roles, 38,
1025–1048.
Pomerantz, E. M., Altermatt, E. R., & Saxon, J. L. (2002). Making the
grade but feeling distressed: Gender differences in academic perfor-
mance and internal distress. Journal of Educational Psychology, 94,
396 – 404.
Radmacher, K., & Azmitia, M. (2006). Are there gendered pathways to
intimacy in early adolescents’ and emerging adults’ friendships? Journal
of Adolescent Research, 21, 415–448.
Roid, G. H., & Fitts, W. H. (1988). Tennessee self-concept scale (TSCS),
revised manual. Los Angeles: Western Psychological Services.
Rosenberg, M. (1986). Self-concept from middle childhood through ado-
lescence. In J. Suls & A. G. Greenwald (Eds.), Psychological perspec-
tives on the self (pp. 107–136). Hillsdale, NJ: Erlbaum.
Sheets, V. L., & Lugar, R. (2005). Friendship and gender in Russia and the
United States. Sex Roles, 52(1/2), 131–140.
Sondhaus, E. L., Kurtz, R. M., & Strube, M. J. (2001). Body attitude,
gender, and self-concept: A 30-year perspective. The Journal of Psy-
chology, 135, 413–429.
Stetsenko, A., Little, T. D., Gordeeva, T., Grasshof, M., & Oettingen, G.
(2000). Gender effects in children’s beliefs about school performance: A
cross-cultural study. Child Development, 71, 517–527.
Stringer, S. J., Reynolds, G. P., & Simpson, F. M. (2003). Collaboration
between Classroom teachers and a school counselor through literature
circles: Building self-esteem. Journal of Instructional Psychology, 30,
69 –76.
Swim, J. K., Hyers, L. L., Cohen, L. L., & Ferguson, M. J. (2001).
Everyday sexism: Evidence for its incidence, nature, and psychological
impact from three daily diary studies. Journal of Social Issues, 57,
31–53.
Tiggemann, M., & Rothblum, E. D. (1997). Gender differences in internal
beliefs about weight and negative attitudes towards self and others.
Psychology of Women Quarterly, 21, 581–593.
Twenge, J. M. (1997). Changes in masculine and feminine traits over time:
A meta-analysis. Sex Roles, 36, 305–325.
Twenge, J. M. (2000). The age of anxiety? Birth cohort change in anxiety
and neuroticism, 1952–1993. Journal of Personality and Social Psychol-
ogy, 79, 1007–1021.
Twenge, J. M., & Campbell, W. K. (2001). Age and birth cohort differ-
ences in self-esteem: A cross-temporal meta-analysis. Personality and
Social Psychology Review, 5, 321–344.
Twenge, J. M., & Crocker, J. (2002). Race and self-esteem: Meta-analysis
comparing Whites, Backs, Hispanics, Asians, and American Indians.
Psychological Bulletin, 128, 371– 408.
Twenge, J. M., & Nolen-Hoeksema, S. (2002). Age, gender, race, socio-
economic status, and birth cohort differences on the Children’s Depres-
sion Inventory: A meta-analysis. Journal of Abnormal Psychology, 111,
578 –588.
Walter, T., & Davie, G. (1998). The religiousity of women in the modern
west. British Journal of Sociology, 49, 640 – 660.
Wark, G. R., & Krebs, D. L. (1996). Gender and dilemma differences in
real-life moral judgment. Developmental Psychology, 32, 220 –230.
Weiss, E. M., Kemmler, G., Deisenhammer, E. A., Fleischhacker, W. W.,
& Delazer, M. (2003). Sex differences in cognitive functions. Person-
ality and Individual Differences, 35, 863– 875.
Werner, N. E., & Crick, N. R. (2004). Maladaptive peer relationships and
the development of relational and physical aggression during middle
childhood. Social Development, 13, 495–514.
Wertheim, E. H., Paxton, S. J., Shultz, H. K., & Muir, S. L. (1997). Why
do adolescent girls watch their weight? An interview study examining
sociocultural pressures to be thin. Journal of Psychosomatic Re-
search, 42, 345–355.
Wicks-Nelson, R., & Israel, A. C. (2003). Behavior disorders of childhood.
Upper Saddle River, NJ: Prentice Hall.
Wood, W., Rhodes, N., & Whelan, M. (1989). Sex differences in positive
well-being: A consideration of emotional style and marital status. Psy-
chological Bulletin, 106, 249 –264.
Wu, Y., & Smith, D. E. (1997). Self-esteem of Taiwanese children. Child
Study Journal, 27, 1–19.
Received December 10, 2007
Revision received July 21, 2008
Accepted July 22, 2008 䡲
45GENDER DIFFRENCES IN DOMAIN-SPECIFIC SELF-ESTEEM