Jsep Article p311

EXERCISE PSYCHOLOGY
JOURNAL OF SPORT & EXERCISE PSYCHOLOGY, 2005, 27, 311-334

© 2005 Human Kinetics, Inc.
The Effect of Exercise on Global Self-Esteem:

A Quantitative Review
John C. Spence1, Kerry R. McGannon2, and Pauline Poon1

1University of Alberta; 2University of Iowa
The purpose of this study was to quantitatively review the body of research
on exercise and global self-esteem (GSE). This review focuses specifically
on studies using adults and also incorporates both published and unpublished
works. Computer and manual searches identified 113 studies matching the
selection criteria. Each study was coded according to 20 study features. A total
of 128 effect sizes (d) were derived. As indicated by effect-size magnitude,
participation in exercise brought about a small change in GSE (dd = +0.23).
Change in physical fitness and type of program were significant moderators of
the effect of exercise on GSE. Larger effect sizes were observed for those who
experienced significant changes in physical fitness and those participating in
exercise or lifestyle programs as opposed to skills training.
Key Words: meta-analysis, self-concept, physical activity
The self is one of the most widely utilized psychological constructs in con-
temporary society (Harter, 1999; Hattie, 1992). In Western culture, the academic
literature and popular press presuppose that everyone not only has a sense of self
but that the self we have is a social phenomenon (Cooley, 1902; James, 1890/1950;
Mead, 1934). Stated differently, who we are is thought to be reciprocally influenced
by our relationships with others, providing the key to understanding how and why
we think, feel, and behave as we do, as well as the impetus for enriching our lives
and relationships.
Researchers interested in exercise and physical activity have also recognized
the importance of self-related phenomena. This is because exercise participation is
linked to mental and psychological benefits (Gauvin, Spence, & Anderson, 1999;
Landers & Arent, 2001). Involvement in regular exercise and physical activity
programs that improve skill, knowledge, fitness, and health are linked to enhanced
self-perceptions (Fox, 1997). These changes may then generalize to more favor-
able views about the self, leading to an improved sense of well-being (Berger &
McInman, 1993; Morgan, 1985).
1Facultyof Physical Education and Recreation, & Alberta Centre for Active Living,
E-424 Van Vliet, University of Alberta, Edmonton, Alberta, T6G 2H9 Canada; 2Dept. of
Health and Sport Studies, University of Iowa, Iowa City, IA 52242.
311
Brought to you by ULB - TEAM HUMANWISSENSCHAFTEN | Unauthenticated | Downloaded 12/01/24 02:42 PM UTC
312 / Spence, McGannon, and Poon
If people experience a more favorable view of themselves as a result of par-

ticipating in physical activity, they should be more likely to continue participating
(Sonstroem, 1997b, 1998). Studying and understanding this connection is important
because despite the well-documented physical, psychological, and social benefits
of regular physical activity (Bouchard, Shephard, & Stephens, 1994; Gauvin et
al., 1999), physical inactivity remains pervasive. It is estimated that upward of
two-thirds of the industrialized world does not achieve minimum physical activity
guidelines (Oja, 1995; U.S. Department of Health and Human Services, 1996).
Physical inactivity thus constitutes a major public health concern (U.S. Dept. HHS,
1996), along with related social and economic costs (Colditz, 1999; Katzmarzyk,
Gledhill, & Shephard, 2000).
These notions have made the self and self-related phenomena an important
topic of study in exercise psychology. In particular, self-esteem has been proposed
as the variable with the highest payoff for reflecting psychological benefits obtained
from participation in regular exercise (Folkins & Sime, 1981; Hughes, 1984). In
turn, self-esteem has been implicated as a central aspect of psychological health
and well-being (e.g., Greenberg, Solomon, Pyszczynski, et al., 1992; DiLorenzo,
Bargman, Stucky-Ropp, et al., 1999; Rosenberg, 1965) and, as mentioned, is a pos-
sible key in helping to solve the problem of lack of exercise participation. Thus an
abundance of literature has attempted to establish a link between physical activity
participation and self-esteem.
Literature reviews (e.g., Gruber, 1986; Sonstroem, 1984) have noted that
approximately 60% of studies reviewed report a positive association between
physical activity participation and high self-esteem. This has led to the widespread
assumption that enhanced overall or global self-esteem is an automatic outcome
of participation in physical activity. However, recent studies (Calfas & Taylor,
1994; McAuley, Mihalko, & Bane, 1997; McAuley, Blissmer, Katula, Duncan, &
Mihalko, 2000) have questioned the magnitude of change that may be expected for
global self-esteem as a result of participating in physical activity.
Additionally, so far only one meta-analysis examining the effect of exercise
on self-concept1 has included adults (McDonald & Hodgdon, 1991), and its results
have several limitations. First, this study explored only aerobic forms of exercise
(e.g., jogging, dancing) and their link to self-concept. Given the rising popularity of
other forms of exercise (e.g., weight training, flexibility training, leisure activities)
and the differential value these might have in different segments of the population
(e.g., the elderly might value leisure forms of activity as opposed to jogging; Martin
& Sinden, 2001), exploring self-esteem and its relationship to other forms of exercise
would also be advantageous. In fact, other meta-analyses found that weight training
was more effective than aerobic exercise for enhancing mood (Arent, Landers, &
Etnier, 2000; North, McCullagh, & Tran, 1990).
Second, McDonald and Hodgdon only included published studies on self-
concept and exercise, which is somewhat problematic if we consider the tendency
for journals to publish findings that are statistically significant as opposed to non-
significant (Spence & Blanchard, 2001). Thus, those researchers who do not find
significant statistical evidence where self-esteem and exercise are concerned could
be influenced to not submit their findings, or to be rejected when they do submit
them. Thus it is possible that the published literature overstates the significance of
findings where self-concept or self-esteem and exercise are concerned.
Exercise and Self-Esteem / 313
Finally, McDonald and Hodgdon combined different aspects of “self” such

as global self-esteem, physical self-concept, and body cathexis. They justified this
procedure by claiming that the constructs were highly correlated. However, since
clear theoretical distinctions have been made between these constructs, particularly
in how they relate to physical activity (Fox & Corbin, 1989; Sonstroem & Morgan,
1989), it is important to summarize them independently. Moreover, Harter (1999)
recently noted that aggregating specific self-evaluations into a single global score
masks the meaningful distinctions between one’s sense of adequacy across vary-
ing domains.
The division of self-esteem into subdomains has been the basis of Sonstroem
and Morgan’s (1989) multidimensional Exercise and Self-Esteem Model (EXSEM).
This model has been advocated as an important theoretical guide to employ for
study and further advancement in understanding exercise and self-esteem (see Fox,
1997, 2000; Marsh, 1997; McAuley et al., 2000). Moreover, it has been noted that
employing the EXSEM to study self-esteem in the physical activity domain will
provide researchers with examinations of processes and pathways (i.e., mechanisms)
that relate to changes in self-esteem (Sonstroem & Morgan, 1989).
The EXSEM is based on a multifaceted hierarchical model of self-esteem
(see Shavelson, Hubner, & Stanton, 1976). In brief, the model is hierarchically
arranged according to increasing situational specificity, with global self-esteem
(GSE) residing at the top, physical self-competence and physical acceptance in the
middle, and physical self-efficacy at the lower level. Constructs at the base/lower
level feed into and affect the global elements, but are more closely related to the
situation/context at hand. Sonstroem and Morgan (1989) have only included those
elements in the model that have been shown in previous research to be typically
associated with exercise behavior. This is a competence-based model in which
changes in physical fitness are proposed to lead to enhanced self-efficacy, which
in turn bolsters physical competence that then results in an increased perception
of global self-esteem.
Hypothesized Moderators of Effect of Exercise on Self-Esteem

Because few studies have used the EXSEM in experimental designs (see
Alfermann & Stoll, 2000; Caruso & Gill, 1992; Glassford, 1993; McAuley et al.,
2000), the model cannot be tested in this meta-analysis. However, the EXSEM
can still be used to help generate hypotheses about how exercise affects GSE and
which factors may influence or moderate that relationship. Based on the original
hypotheses proposed for the model (Sonstroem & Morgan, 1989) and other writ-
ings of Sonstroem (1984, 1997a, 1998), we have identified the following variables
as potential moderators of the effect of exercise on GSE.
First, changes in physical fitness2 are thought to be the mechanism through
which exercise could affect self-efficacy and subsequently GSE. Thus, exercise
participants who experience changes in physical fitness should experience larger
changes in GSE as opposed to those who do not experience such changes.
Second, if physical fitness change is a potential moderator of the exercise/GSE
relationship, then initial physical fitness level should also be considered a potential
moderator, since less physically fit individuals should experience more of a change
in GSE as a result of exercising compared to more physically fit individuals.
Third, similar to the case for initial physical fitness level, initial self-esteem
level should also serve as a moderator of exercise on GSE. Thus, participants with
initially lower GSE scores should experience the most change in self-esteem as a
result of exercise and changes in fitness.
Finally, since training volume and dose are related to changes in physical
fitness (McArdle, Katch, & Katch, 1986), aspects of the exercise program such as
frequency, duration, intensity, and length of the program should be related to changes
in GSE. Specifically, programs that take place over a longer period, with exercise
bouts of higher frequency, intensity, and duration, should lead to larger changes
in GSE vs. programs of shorter length with lower doses of exercise.
Despite the vast number of studies exploring the GSE and exercise relation-
ship, it should be clear that while the potential benefit of physical activity on GSE
has long been espoused in the sport and exercise literature (Folkins & Sime, 1981;
Hughes, 1984), the foregoing studies and their limitations have put into question
the exact magnitude of change in GSE that results from participation in physical
activity. Therefore, the primary purpose of this paper was to take the limitations of
previous research into consideration and present the findings of a more comprehen-
sive meta-analysis in order to explore the effect of exercise on GSE. To accomplish
this purpose, the current meta-analysis included published and unpublished studies
(e.g., dissertations, theses) and also explored the effects of other forms of exercise
(e.g., weight training) in addition to aerobic exercise on GSE.
Method
Literature Search and Study Selection
Search Strategy. We conducted computer-based information searches in
Current Contents (July 1993–Sept. 2001), Dissertation Abstracts (1865–Dec.
2003), Medline (1966–Dec. 2003), PsycINFO (1960–Dec. 2003), and Sport Discus
(1975–Dec. 2003). Also, manual searches were done through Current Contents
(1987–June 1993), Physical Education Index (1983–1993), and Completed Research
in Health, Physical Education, and Recreation (1968–1995). Reference lists of all
major reviews and primary studies were checked for unidentified references. To find
relevant studies on exercise and GSE, the keywords used in the searches included,
but were not limited to, terms such as: exercise, physical activity, self-esteem,
self-concept, and self-perception. The search was limited by language in that only
studies published in English were included. In an attempt to obtain unpublished
material, we emailed researchers with recent publications on the topic and asked if
they would share any unpublished manuscripts or data in their possession. The final
cutoff date for articles to be included in this review was December 2003.
Inclusion/Exclusion Criteria. The criteria for including studies in the final
sample were as follows: (a) Information had to be reported in a study format, either
published or unpublished, as opposed to a case study or editorial; (b) the dependent
variable had to be a global measure of self-concept, self-esteem, or self-percep-
tion; (c) the independent variable had to include exercise/physical activity as an
intervention; (d) because of the large number of studies found in our initial search
on the topic, we decided to limit our focus to adults and thus required that study
participants be at least 18 years of age or that the mean age of the sample likewise
be 18 years or older; (e) studies had to include a nonexercising control group or a
control group that was maintaining its usual level of physical activity; (f) the train-
ing program had to last more than 1 week; and (g) outcomes of the intervention
had to be quantified and sample size be provided.
Included Studies. The original search resulted in 426 potential studies. Based
on our inclusion criteria, 113 studies (42 published, 71 unpublished) revealing 128
effect sizes were included in the analysis.3 In total, 7724 participants were included
in these studies.
Coding of Studies
Once studies were identified for inclusion, they were collected, coded for
moderator effects, and an estimate of treatment effect was calculated.
Hypothesized Moderator Variables. Based on the original hypotheses
proposed for the EXSEM (Sonstroem & Morgan, 1989) and other writings of
Sonstroem (1984, 1997a, 1998), we hypothesized that changes in physical fitness,
initial self-esteem level, initial physical fitness level, and dose of exercise (fre-
quency, duration, intensity, and length of program) would explain any variation in
the effect of exercise on GSE.
Additional Moderator Variables. Further, for each study we recorded the
characteristics of the sample (age, sex, population, health status), methodology
(self-term, questionnaire, participant assignment, control condition), and treatment
(type of intervention, mode of exercise, type of physical fitness measure). In order
to address concerns about publication bias, we also coded the studies for publication
status, i.e., published vs. unpublished. Finally, based on a subjective assessment
of internal validity, we rated each study on a scale from 1 (poor) to 3 (good) for
methodological quality.
The factors considered for this rating were assignment of participants to condi-
tions, dropout rate, dose of exercise, and assessment of physical fitness. This rating
was applied in a hierachical fashion. First we assessed the method of assignment of
participants to conditions and the dropout rate. If participants were not randomly
assigned to conditions and/or the dropout rate was greater than 50%, the study was
rated as poor. Second, if these conditions were met and a specific dose of exercise
was required and/or the physical fitness of participants was measured before and
after the intervention, the study was rated as good. Otherwise the study was rated
as being only fair. For example, studies were rated as good if random assignment
was used to allocate participants to groups, the dropout rate was less than 50%, and
assessments of physical fitness were conducted before and after the intervention.
Publication status and study quality were included as methodology moderators.
Coding was undertaken independently by two coders. Their initial agreement rate
was 84%. Disagreements between the two coders were resolved through discussion
and further review of the disputed studies.
Effect Size Computation. Hedges and Olkin’s (1985) procedure to derive
and analyze effect sizes was used. Specifically, the effect size, g, was defined as
the difference between the means of the experimental group and the control group
divided by the pooled within-group standard deviation. The pooled standard devia-
tion is considered to be a representative indicator of within-study variance (Hedges
& Olkin, 1985). If descriptive statistics were not available, then estimates of g were
calculated from other statistics such as t, F, or p (see Dunlap, Cortina, Vaslow, &
Burke, 1996; Glass, McGaw, & Smith, 1981; Ray & Shadish, 1996).
Effect sizes are positive if exercisers exhibit higher levels of GSE. The effect
sizes were corrected, d, for the bias from g’s overestimate of the population effect
size for small samples (Hedges & Olkin, 1985). Statistical significance of d was
determined by the presence or absence of 0 within the 95% CI for d. The absence
of 0 from the CI indicates a significant effect size. A d of 1 indicates a change in
magnitude equivalent to 1 SD. According to Cohen (1992), effect sizes can be
categorized as small (dd = 0.2), medium (d = 0.5), or large (d = 0.8).
Unit of Analysis. The unit of analysis was the study finding. Multiple effect
sizes within outcomes were taken from studies only if different samples were used
to produce these effect sizes (e.g., males vs. females).
Missing Data. An effect size of 0 was imputed if complete descriptive (i.e.,
means, variances, and sample size) and/or inferential statistics were unavailable and
it was clear that there was no treatment effect (i.e., Meanexp – Meancont = 0). Also,
if a result was only described as not significant and n and/or direction of change
could not be determined, then 0 was substituted for the effect size (see Lou, Abrami,
Spence, et al., 1996). However, if a result was described as not significant, and n
and direction of change could be determined, then the effect size was calculated
based on the value of t for p = .05 and divided by a factor of 2 (Sedlmeier & Giger-
enzer, 1989). Overall, 0 was substituted in 6 cases (5%) and tt/2 was substituted in
4 cases (3.4%).
Combining Effect Sizes. Because a great effort was made to identify and
obtain every study conducted on the topic, a fixed-effects model (Hedges & Olkin,
1985) was used when combining effect sizes. That is, we assumed that the stud-
ies in our meta-analysis constituted the universe of studies. When calculating the
overall average effect size for each outcome, d+, we weighted each study’s effect
size by the reciprocal of its variance before averaging it with other effect sizes. This
procedure typically gives additional weight to effect sizes that come from studies
with larger sample sizes (Hedges, 1994).
Data Analysis
To detect whether the studies shared a common effect size, we tested the
homogeneity of the set of effect sizes with a homogeneity statistic, Qt, which has
an approximate chi-square distribution with k – 1 degree of freedom, where k is
the number of effect sizes (Hedges & Olkin, 1985). If the resulting chi-square was
significant, the effect sizes were determined to be heterogeneous. In such a case,
Hedges and Olkin (1985) recommend that the total homogeneity statistic be par-
titioned into a between-classes effect and a test of the homogeneity of the effect
sizes be conducted within each class. The between-classes effect was estimated by
Qb, which has an approximate chi-square distribution with p – 1 degree of freedom,
where p is the number of classes. A large value of Qb indicated that there were
significant differences among the classes of effect sizes. As suggested by Hedges
(1994), for any variable with more than two levels, statistically significant between-
classes effects were followed-up with fixed-effects contrasts (Bonferroni method).
The homogeneity of the effect sizes within each class was estimated by Qwi, which
has an approximate chi-square distribution with m – 1 degree of freedom, where m
is the number of effect sizes in the class.
Results
Based on the average weighted effect size (d+ = 0.23, SE = 0.02), adults in
the studies in this analysis experienced approximately 1/4 of a standard deviation
increase in GSE when they participated in physical activity or exercise. While
this change was deemed small, it was significantly different from zero (95% CI,
.18/.28) and the effect sizes were found to be homogeneous, Qt(127) = 144.39, p =
.14. This weighted effect size was similar in magnitude to the unweighted median
effect size (MM = 0.24) and the unweighted mean effect size (d = 0.27). Because
of concerns about the imputing of zero for those cases in which the findings were
nonsignificant, and since it was not possible to calculate an effect size, we also
calculated the average weighted effect size with those cases removed. Again, there
was very little difference between the effect size with zero removed (dd0– = 0.24, SE
= 0.02, 95% CI, .19/.28) and the average weighted effect size.
Even though the effect sizes for GSE were homogeneous, we proceeded with
an analysis of both the hypothesized and additional moderators as Rosenthal (1995)
recommends.4 Because the additional moderators were tested in an exploratory
fashion, we used a Bonferroni inequality to obtain a more conservative alpha of
.004 (i.e., .05/13 categorical tests) to determine statistical significance. Tables 1
and 2 provide the average weighted effect size for each level of each moderating
variable, along with indicators of between-group differences (Qb) and within-group
homogeneity (Qw).
Hypothesized Moderators of Exercise on Self-Esteem

We tested if any of the following variables served as significant moderators
of exercise on GSE: change in physical fitness, initial fitness level, initial self-
esteem level, exercise intensity, exercise frequency, exercise duration, and length
of program. Change in physical fitness, Qb (2) = 3.78, p < .05, was the only signifi-
cant hypothesized moderator of GSE. Significant changes in physical fitness (dd =
0.32) were associated with greater changes in GSE when compared to those cases
where no changes occurred (dd = 0.15). Furthermore, the effect size for the latter
category did not differ significantly from zero. In terms of initial physical fitness
level and initial self-esteem level, while not significant, the direction of change
was as we expected. Thus persons with lower initial physical fitness (dd = 0.29)
and lower initial self-esteem (dd = 0.28) demonstrated larger changes in GSE as a
result of exercise than did those reporting moderate initial physical fitness levels
(dd = 0.20) and moderate initial self-esteem (d = 0.22). A nonsignificant trend was
also observed for exercise frequency in that the more frequent the participation,
the larger the increase in GSE.
Additional Moderators of Exercise on Self-Esteem

Sample Variables. None of the sample variables served as significant mod-
erators of exercise and GSE. Although population was a significant moderator
when tested at an alpha of .05, it was not significant when the more conservative
alpha of .004 was used. No significant within-class variability was detected at the
variable level; however, the effect sizes in the subcategories of women, Qw (56) =
76.18, p < .05, and the psychologically unhealthy, Qw (13) = 24.21, p < .05, were
heterogeneous.
Table 1 Summary Statistics for Hypothesized Moderators of Exercise and Self-

Esteem
Variable Qb k d+ SE 95% CI Qw
Initial fitness level 3.10 141.04

Low 53 .29 .05 .20/.38 60.71
Moderate 73 .20 .03 .14/.26 80.33
Initial self-esteem level 0.91 136.52
Low 28 .28 .06 .16/.41 25.83
Moderate 98 .22 .03 .17/.27 110.69
Exercise intensity 0.81 125.36
Low (< 45%) 21 .26 .07 .13/.40 29.85
Moderate 70 .21 .03 .15/.28 76.56
High (70%+) 25 .26 .05 .15/.37 18.95
Exercise frequency 3.27 125.41
< 3 times/wk 33 .22 .05 .13/.31 45.43
3 times/wk 63 .23 .04 .15/.30 67.91
> 3 times/wk 14 .36 .07 .22/.50 12.07
Exercise duration 2.18 128.71
0–30 min 20 .32 .08 .15/.48 13.25
31–45 min 24 .19 .06 .07/.31 42.72
46–60 min 47 .24 .04 .16/.32 49.68
> 60 min 7 .13 .13 –.13/.38 7.94
Length of program 0.52 113.59
< 9 wks 39 .24 .05 .14/.33 61.48*
9 to 14 wks 53 .26 .04 .18/.33 52.72
> 14 wks 20 .21 .05 .11/.31 14.51
Change in physical fitness 3.79* 63.20
Yes 59 .32 .04 .25/.40 52.68
No 16 .15 .08 .00/.31 10.53
Note: CI = confidence interval; Qb = measure of between-group effect; k = number of

effect sizes; d+ = average weighted effect size; SE = standard error of d+; 95% CI = 95%
confidence intervals for d+; Qw = measure of within-group homogeneity.
* p < .05
Methodological Variables. None of the methodological variables served as

significant moderators of GSE. While no significant within-class variability was
detected at the variable level, the effect sizes in the subcategories of unpublished
studies, Qw (79) = 103.58, p < .05, and poor quality studies, Qw (54) = 73.60, p < .05,
were heterogeneous. In all, 43% of the effect sizes were from poor quality studies and
63% were from unpublished sources. Of the effect sizes from poor quality studies,
89% were from nonequivalent designs (i.e., no random assignment or matching),
84% were from unpublished sources, and 61% were from studies in which physical
fitness was not measured. As shown in Table 3, there was a relationship between
Table 2 Summary Statistics for Additional Moderators of Exercise and Self-Esteem
Variable Qb k d+ SE 95%CI Qw
Sample Variables
Age 2.01 113.63
Young 58 .24 .03 .17/.30 64.41
Middle-age 32 .30 .06 .19/.41 33.64
Older adults 18 .18 .08 .03/.33 15.58
Sex 1.00 143.31
Women 57 .21 .04 .14/.29 76.18*
Men 21 .28 .05 .17/.38 15.57
Mixed sample 49 .22 .04 .15/.30 51.56
Population 8.02 136.29
Public 48 .32 .04 .24/.41 44.00
Students 55 .19 .03 .12/.25 64.43
Patients 15 .28 .09 .10/.46 20.54
Other 9 .12 .09 –.06/.30 7.31
Health status 0.64 143.67
Healthy 97 .22 .03 .17/.27 111.13
Physically unhealthy 15 .23 .08 .06/.40 8.33
Psychologically unhealthy 14 .29 .08 .13/.46 24.21*
Methodological Variables
Self term 0.05 142.29
Self-esteem 44 .23 .04 .14/.31 49.48
Self-concept 83 .24 .04 .18/.30 92.81
Scale 3.52 140.87
R-SES 27 .18 .05 .08/.28 20.57
TSCS 67 .24 .03 .18/.31 65.97
SCS 8 .39 .10 .18/.59 13.22
Other 26 .20 .05 .10/.31 41.11*
Publication status 0.22 144.17
Published 48 .24 .04 .17/.32 40.59
Unpublished 80 .22 .03 .16/.28 103.58*
Control group 0.16 144.22
No treatment 110 .23 .03 .18/.28 121.33
Other treatment 13 .24 .08 .07.40 17.83
Some exercise 5 .18 .14 –.09/.44 5.06
Group assignment 6.60 137.82
Random 54 .26 .04 .18/.34 54.63
Matching 9 .48 .11 .26/.71 6.05
Nonequivalent 65 .20 .03 .14/.26 77.14
(continued)
Table 2 (Continued)
Variable Qb k d+ SE 95%CI Qw
Study quality 5.29 139.10

Poor 55 .17 .04 .10/.24 73.60*
Fair 52 .27 .04 .20/.35 52.11
Good 21 .31 .07 .18/.45 13.39
Treatment Variables
Type of program 14.80** 129.58
Exercise a 109 .26 .03 .21/.31 111.26
Lifestyle a 8 .36 .11 .13/.58 7.54
Skills training b 11 –.03 .07 –.17/.11 10.79
Exercise mode 4.78 131.85
Aerobic 59 .25 .04 .18/.32 68.56
Aerobic & other 31 .22 .05 .12/.32 25.42
Martial arts 4 .00 .12 –.22/.23 2.85
Flexibility 9 .20 .10 .00/.40 12.46
Strength 15 .26 .06 .14/.37 20.83
Mix 5 .18 .10 –.02/.38 1.72
Type of fitness measure 4.23 48.44
Aerobic 27 .32 .06 .20/.44 33.68
Body composition 4 .38 .19 –.01/.76 6.13
Strength 8 .37 .08 .22/.53 2.11
Other 5 .19 .15 –.11/.48 2.15
Combined fitness 5 .18 .11 –.03/.38 2.16
Combined fitness & body comp. 10 .41 .09 .24/.58 2.22
Note: CI = confidence interval; Qb = measure of between-group effect; k = number of

effect sizes; d+ = average weighted effect size; SE = standard error of d+; 95% CI = 95%
confidence intervals for d+; Qw = measure of within-group homogeneity; R-SES = Rosen-
berg Self-Esteem Scale; TSCS = Tennessee Self-Concept Scale; SCS = Self-Cathexis
Scale. a, b Variables with different superscripts differ significantly at p < .01.
* < .05; **
*p **p < .001
Table 3 Prevalence of Study Quality Among Published and Unpublished Studies
Published Unpublished
Study quality (n = 48) (n = 80) χ2 (1)
Poor 9 46 24.89***
Fair 28 23 0.49
Good 11 11 0.00
*** < .001

***p
publication status and study quality, χ2 (2, N = 128) = 18.54, p < .001, with effect
sizes from poor quality studies being more likely to remain unpublished, χ2 (1, N
= 55) = 24.89, p < .001.
Treatment Variables. Type of program, Qb (2) = 14.80, p < .001, was a
significant moderator of GSE. Those participating in exercise (dd = 0.26) or lifestyle
enrichment programs (dd = 0.36) experienced a larger increase in GSE than those
in skills training activities (d = –0.03).5 Also, the effect sizes for both exercise and
lifestyle programs were larger than zero. The other variables, exercise mode and
type of fitness change, were not significant moderators of GSE.
Sample Size and Power Issues

In this meta-analysis, the unweighted median effect size (M M = 0.24), the
unweighted mean effect size (dd = 0.27), and the weighted mean effect size (d+ =
0.23) indicate a small treatment effect for exercise on GSE across studies. However,
in order for this small treatment effect to be detected as being significant (at α =
.05, two-tailed test, and power = .80) in any one study, the sample sizes of studies
would have to range between 200 and 235 (Overall & Doyle, 1994). The total sample
sizes actually ranged from 12 (dePiano, dePiano, Carter, & Wanlass, 1984) to 343
(Wescott, 1980), with 95% of effect sizes coming from sample sizes less than 200.
Furthermore, since the average number of participants per effect size was 61 and
the median was 41, it appears that some small treatment effects went undetected
and were classified as nonsignificant.6
In answer to the question of whether unpublished studies were more under-
powered than published ones, results revealed that there was no significant differ-
ence between the average sample size of unpublished (n = 58.55) vs. published (n
= 63.81) studies. Furthermore, the median sample sizes were almost identical for
unpublished (M M = 40.5) and published (M = 42) studies.
Discussion
The current findings demonstrate that exercise participation leads to small
yet significant increases in GSE. The effect sizes in this case were homogeneous.
Thus while there is a positive and reliable effect of exercise on self-esteem, this
relationship is small at the global level. These results are in contrast with McDonald
and Hodgdon’s (1991) meta-analysis which found that exercise leads to moderate
changes in global self-esteem. Some light can be shed on why these findings differ
if considered in terms of multidimensional or hierarchical models of self-esteem
discussed earlier (e.g., Sonstroem & Morgan, 1989). In this regard it has been
noted that the smallest change in self-esteem occurs at the global level. Part of the
reason for this is that exercise participation is thought to have the greatest influ-
ence on self-esteem at domain-specific levels (e.g., physical self-worth, physical
competence).
While McDonald and Hodgdon (1991) included domain-specific self-
evaluations in their study, these were subsumed into, and used interchangeably
with, a global construct, i.e., self-concept. Given that the current meta-analysis
employed only measures of GSE, it is possible that McDonald and Hodgdon’s
(1991) findings regarding self-esteem and exercise participation were larger because
of their inclusion of these domain-specific components.
It has been claimed that self-concept is the psychological construct with the
most potential for change because of participation in an exercise program (e.g.,
Folkins & Sime, 1981; Hughes, 1984). In comparison with meta-analyses of other
psychological constructs, it appears that state anxiety (dd = –0.34, Long & van
Stavel, 1995; d = –0.25, Petruzzello, Landers, Hatfield, Kubitz, & Salazar, 1991)
and depression (dd = –0.72, Craft & Landers, 1998; d = –0.59, North et al., 1990)
show just as much if not more change than GSE as a result of exercise. Thus, from
the results of the present meta-analysis it appears that the benefits of exercise for
GSE are overstated in the literature on physical activity and exercise. Much of this
optimism about the potential effectiveness for exercise increasing GSE is probably
the result of the McDonald and Hodgdon (1991) review. Also, earlier reviews (e.g.,
Folkins & Sime, 1981) base their claims about the benefits of exercise on the results
of a few studies, some with unique populations (e.g., mentally retarded students).
Effects of the Exercise Program

The only two significant moderators of exercise on GSE were type of program
and change in fitness. In the former, exercise and lifestyle programs resulted in small
to moderate increases in self-esteem, respectively, while skills training activities
had no effect on self-esteem. The fact that skills training activities had no effect
on GSE is somewhat surprising, considering that participation in such activities
should lead to enhanced physical competence which is one of the antecedents of
GSE in the EXSEM (Sonstroem & Morgan, 1989).
Alternatively, changes in physical fitness are related to significant increases
in GSE. While this latter finding supports one of the original tenets of the EXSEM,
Sonstroem (1997a) has since modified the claim that increased physical fitness is
required for enhanced GSE and placed more of an emphasis on “perceptions of
improvement” in physical fitness. The idea that perception of fitness is a mediator
of the relationship between exercise and mental health has become popular in the
literature (see Heaps, 1978; O’Donoghue, 1990; Plante, 1999; Plante Coscarelli,
Caputo, & Oppezzo, 2000; Plante, LeCaptain, & McLain, 2000).
We cannot comment on the role that perceived changes in physical fitness
played in our analysis because it was not measured in the original studies. Future
studies should test whether perceived physical fitness serves as a mediator or modera-
tor of the effect of exercise on GSE. However, it does appear from our analysis that
increases in actual physical fitness are related to enhanced GSE. Thus, participation
in programs that lead primarily to enhanced physical fitness vs. enhanced physical
competencies may be more effective for improving the self-esteem of adults. One
caveat is that approximately 40% of effect sizes came from studies in which actual
physical fitness was not measured. Thus, along with perceived fitness, future studies
of exercise and self-esteem should test for changes in actual physical fitness.
Given that change in physical fitness is a significant moderator of the effect of
exercise on GSE, the overall lack of support for a dose-response (i.e., intensity, fre-
quency, duration, or length of program) relationship might be surprising. However,
it has been argued that little data supports the existence of a strong dose-response
relationship for exercise and most psychosocial variables (Gauvin et al., 1999;
Rejeski, 1994). Similar to the case that has been made for the role of perceived
changes in fitness, perhaps it is the subjective perception of the dose (Dishman,
1994) that is more important in determining the effect of exercise on GSE.
Methodological Issues
No moderator effects were observed for methodological variables in the lit-
erature on exercise and self-esteem. The fact that there was no difference in effect
size between published and unpublished studies suggests it is not because of a lack
of treatment effect that the unpublished studies remain unpublished. Since 58% of
unpublished effect sizes were derived from studies deemed to be of poor quality,
it is likely the quality of those studies that explains why they remain unpublished.
An interesting trend was observed for study quality in that studies rated as either
fair (dd = 0.27) or good (d = 0.31) had larger effect sizes than those rated as poor (d
= 0.17). While no significant within-class variability was detected for study quality
overall, the effect sizes in the poor category were heterogeneous, suggesting that
some other factor may explain the effect size for that group.
Based on the sample sizes of the studies in this meta-analysis, it can be surmised
that GSE studies are generally undersized and underpowered. That is, researchers
are not using samples of sufficient size to provide adequate power in their studies
in order to detect significant findings for the effect of exercise on GSE. This was
the case regardless of whether studies were published or unpublished. In fact, 95%
of the effect sizes in this analysis came from sample sizes too small to detect the
average weighted effect size to be significant. Thus it is likely that lack of statistical
power has contributed to some of the confusion surrounding exercise and GSE.
Limitations of This Review

A potential limitation of this meta-analysis was the choice to use a fixed-
effects as opposed to a random-effects model when analyzing our effect sizes. In
the fixed-effects model it is assumed that the population effect size is the same for
all studies in the analysis. The alternative is that the population effect sizes vary
randomly from study to study. As a result, the error term in the fixed-effects model
uses only the within-study variability while the error term in the random-effects
model includes both the within-study variability and the variability due to differ-
ences between studies (Hedges & Vevea, 1998). Thus when a fixed-effects model
is used, it is assumed that all the studies conducted on a particular topic have been
included, whereas in a random-effects model it is assumed that only a sample of
the studies that could be done have been found (Hedges, 1994).
Technically, any findings and subsequent conclusions drawn from a fixed-
effects model should be limited to the studies in that analysis. If one wants to make
more general conclusions about the research topic, then a random-effects approach
is more appropriate. Therefore the findings and conclusions of this meta-analysis
should be limited to the studies reviewed. Because every effort was made to find
studies of exercise and GSE, as evidenced by the fact that approximately two-thirds
of the studies included were unpublished, we believe our findings closely approxi-
mate the universe of studies on the topic and thus provide insight on the population
effect size.
Conclusions and Future Directions

In summary, participation in physical activity results in small significant
improvements in global self-esteem. From a statistical point of view, these improve-
ments are small. However, more studies are needed in order to determine the clini-
cal or practical significance of such changes. Further, it appears that increases in

actual physical fitness are required for enhanced self-esteem to occur. Based on
these findings, we suggest, as previous research by McAuley and colleagues (1997;
2000) has done, that the link between GSE and physical activity participation has
been to some degree overstated.
For many years, self-esteem research in the exercise domain has remained
largely descriptive, with studies documenting group differences in self-esteem in
various settings such as work, school, or sport. However, multidimensional models
have allowed researchers who are interested in self-esteem and its role in exercise
participation to learn more about the specific subcomponents of the physical self
and its link to the global self (Fox, 1997, 2000). Despite this, if we are to progress
beyond the documentation of links between specific subcomponents of the self
and related behaviors and map changes in self-perception more precisely, future
research must focus more directly on the mechanisms of change involved in the
self-system.
The current findings reinforce the call for more studies that focus on domain-
specific self-evaluations if the mechanisms through which self-esteem affects exer-
cise participation and vice versa are to be identified. Our findings also underscore
the need for future studies to include specific measures of physical fitness and
competencies or attributes (e.g., physical self-worth, perceived fitness) that might be
expected to be most responsive to intervention and change (Baranowski, Anderson,
& Carmack, 1998; Kraemer, Stice, Kazdin, Offord, & Kupfer, 2001).
To conclude, despite the current meta-analytic finding that physical activity
participation results in small improvements in GSE, the potential payoff from a more
focused exploration of multidimensional self-esteem change in the exercise domain
is high. Not only is this topic understudied but, as the research of McAuley and
colleagues (1997, 2000) indicates, it has great potential from a practical standpoint
for designing exercise interventions in areas such as rehabilitation from illness
and/or injury and weight management.
References
Note: References marked with an asterisk indicate studies included in the meta-analysis.
*Alexander, M.J.L., Nickel, R., Boreskie, S.L., & Searle, M. (2000). Comparison of the
effects of two types of fitness/flexibility programs on gait, mobility and self-esteem
of older females. Journal of Human Movement Studies, 38, 235-268.
*Alfermann, D., & Stoll, O. (2000). Effects of physical exercise on self-concept and well-
being. International Journal of Sport Psychology, 30, 47-65.
Arent, S.M., Landers, D.M., & Etnier, J.L. (2000). The effects of exercise on mood in older
adults: A meta-analytic review. Journal of Aging and Physical Activity, 8, 407-430.
*Asci, F.H. (2003). The effects of physical fitness training on trait anxiety and physical self-
concept of female university students. Psychology of Sport and Exercise, 4, 255-264.
Baranowski, T., Anderson, C., & Carmack, C. (1998). Mediating variable framework in
physical activity interventions: How are we doing? How might we do better? American
Journal of Preventive Medicine, 15, 266-297.
*Barker, W.T. (1985). The effect of weight training on self-concept and locus of control in
male patients in treatment for chemical dependence. Unpublished doctoral disserta-
tion, University of San Francisco.
*Bartlett, S.J. (1996). The psychological effects of exercise on dieting obese women (Doctoral
dissertation, Syracuse University, 1995). Dissertation Abstracts International, 57(4),

B2929. (UMI No. 9625842)
*Bennett, M.F. (1991). The influence of adapted physical education exercise classes on the
self-concept of the post-stroke student (Master’s thesis, San Jose State University,
1990). Master’s Abstracts International, 29(2), 199. (UMI No. 1341655)
Berger, B.G., & McInman, A. (1993). Exercise and the quality of life. In R.N. Singer, M.
Murphey, & L.K. Tennant (Eds.), Handbook of research on sport psychology (pp.
729-760). New York: Macmillan.
*Blumenthal, J.A., Babyak, M.A., Moore, K.A., Craighead, W.E., Herman, S., Khatri, P.,
Waugh, R., Napolitano, M.A., Forman, L.M., Appelbaum, M., Doraiswamy, P.M.,
& Krishnan, K.R. (1999). Effects of exercise training on older patients with major
depression. Archives of Internal Medicine, 159, 2349-2356.
*Blumenthal, J.A., Emery, C.F., Madden, D.J., George, L.K., Coleman, R.E., Riddle, M.W.,
McKee, D.C., Reasoner, J., & Williams, R.S. (1989). Cardiovascular and behavioral
effects of aerobic exercise training in healthy older men and women. Journal of Ger-
ontology: Medical Sciences, 44, M147-M157.
*Bothwell, J.H. (1989). The effects of an aerobic exercise program on self-concept, health
locus of control, and health-related behaviors of Zuni Indians (Doctoral dissertation,
The Fielding Institute, 1988). DAI, 50(6), B2613. (UMI No. 8912149)
Bouchard, C., Shephard, R.J., & Stephens, T. (Eds.) (1994). Physical activity, fitness, and health:
International proceedings and consensus statement. Champaign, IL: Human Kinetics.
*Bouchard, L.M.F. (1979). The effects of a body awareness exercise program on body-
concept, self-concept and body composition of inactive obese and non-obese young
women. Unpublished master’s thesis, University of British Columbia.
*Boyll, J.R. (1986). The effects of active exercise and passive electronic muscle stimulation
on self-concept, anxiety, and depression (Doctoral dissertation, Northern Arizona
University, 1986). DAI, 47, B2219.
*Brown, D.R., Wang, Y.D., Ward, A., Ebbeling, C.B., Fortlage, L., Puleo, E., Benson, H.,
& Rippe, J.M. (1995). Chronic psychological effects of exercise and exercise plus
cognitive strategies. Medicine and Science in Sports and Exercise, 27, 765-775.
*Brown, M.A., Goldstein-Shirley, J., Robinson, J., & Casey, S. (2001). The effects of a
multi-modal intervention trial of light, exercise, and vitamins on women’s mood.
Women and Health, 34, 93-112.
*Brown, R.D. (1986). Effects of a strength training program on strength, body composi-
tion, and self-concept of females (Doctoral dissertation, Brigham Young University,
1985). DAI, 46, A2615.
*Butt, J. (1980). The effects of a self-administering fitness program on anxiety, self-concept
and fitness. Unpublished master’s thesis, Memorial University, St. John’s, Newfound-
land & Labrador.
*Byrom, J.M. (1984). The effects of a cardiovascular fitness program on depression, anxiety,
self-concept, and perceived physical fitness in college women (Doctoral dissertation,
California School of Professional Psychology, Fresno, 1983). DAI, 45, B1907.
Calfas, K.J., & Taylor, W.C. (1994). Effects of physical activity on psychological variables
in adolescents. Pediatric Exercise Science, 6, 406-423.
*Callahan, J., Wandzilak, T., Ansorge, C.J., & Schmidt, R.J. (1987). Effects of aerobic swim
and karate classes on cardiovascular fitness, body fat, self-esteem and mood states of
college students. Paper presented at annual conference of AAHPERD, Las Vegas.
Caruso, C.M., & Gill, D.L. (1992). Strengthening physical self-perceptions through exercise.
Journal of Sports Medicine and Physical Fitness, 32, 416-427.
*Case, S.C. (2003). The effects of group therapy and exercise on depression and self-esteem
in college students. Dissertation Abstracts International, 64(4), B1893. (UMI No.
3086935)
Caspersen, C.J., Powell, K.E., & Christenson, G.M. (1985). Physical activity, exercise, and
physical fitness. Public Health Reports, 100, 125-131.
*Cerrato, M.R. (1977). Changes in body image and self concept among college students
who participate in a skilled program of physical conditioning (Doctoral dissertation,
East Texas State University, 1976). DAI, 37, A4210.
*Clower, M.A. (1979). The relationships of self-concept, movement concept and physical
fitness for college women (Doctoral dissertation, University of Georgia, 1978). DAI,
39, A4120.
*Cocklin, J.C. (1989). The effects of physical fitness and body cathexis on self-concept
change in women after aerobic conditioning (Doctoral dissertation, Oklahoma State
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-159.
Colditz, G.A. (1999). Economic costs of obesity and inactivity. Medicine and Science in
Sports and Exercise, 31(11, Suppl.), S663-S667.
*Collingwood, T.R. (1972). The effects of physical training upon behavior and self attitudes.
Journal of Clinical Psychology, 28, 583-585.
Cooley, C.H. (1902). Human nature and the social order. New York: Scribner’s.
*Coyle, C.P., & Santiago, M.C. (1995). Aerobic exercise training and depressive symp-
tomatology in adults with physical disabilities. Archives of Physical Medicine and
Rehabilitation, 76, 647-652.
Craft, L.L., & Landers, D.M. (1998). The effect of exercise on clinical depression and
depression resulting from mental illness: A meta-analysis. Journal of Sport & Exercise
Psychology, 20, 339-357.
*Cusumano, J.A., & Robinson, S.E. (1992). The short-term psychophysiological effects of
Hatha yoga and progressive relaxation on female Japanese students. Applied Psychol-
ogy: An International Review, 42, 77-90.
*Davis, M.W. (1971). The effects of a cardiovascular conditioning program on selected
psychological responses of college males (Doctoral dissertation, University of Okla-
homa, 1970). DAI, 32, A221.
*de Geus, E.J., van Doornen, L.J.P., & Orlebeke, J.F. (1993). Regular exercise and aerobic
fitness in relation to psychological make-up and physiological stress reactivity. Psy-
chosomatic Medicine, 55, 347-363.
*de Piano, F.A., de Piano, L.C., Carter, W., & Wanlass, R.L. (1984). Physical fitness training:
Adjunctive treatment for the depressed, low self-esteem and muscular tense patient.
Psychotherapy in Private Practice, 2(3), 75-83.
*Dewlen, R.F. (1991). Self-concept changes in physically disabled adult men following
eight weeks of physical fitness training (Doctoral dissertation, University of Kansas,
1990). DAI, 52, A465.
*Dickman, N., & Shea, J. (2002). Does exercise or the exercise setting affect adult self-con-
cept? In Abstract Book of the 12th Commonwealth International Sport Conference
(p. 159). London: Association of Commonwealth Universities.
*DiLorenzo, T.M., Bargman, E.P., Stucky-Ropp, R., Brassington, G.S., Frensch, P.A., &
LaFontaine, T. (1999). Long-term effects of aerobic exercise on psychological out-
comes. Preventive Medicine, 28, 75-85.
Dishman, R.K. (1994). Prescribing exercise intensity for healthy adults using perceived
exertion. Medicine and Science in Sports and Exercise, 26, 1087-1094.
*Donaldson, C.S. (1979). The effects of self-defense training on women’s self-esteem and
locus of control orientation (Doctoral dissertation, California School of Professional
Psychology, 1978). DAI, 39, B3505.
*Dreher, E.R. (1974). The effects of hatha yoga and judo on personality and self-concept
profiles on college men and women (Doctoral dissertation, University of Utah, 1973).
DAI, 34, A4833.
Dunlap, W.P., Cortina, J.M., Vaslow, J.B., & Burke, M.J. (1996). Meta-analysis of experiments with
matched groups or repeated measure designs. Psychological Methods, 1, 170-177.
*Edlund, C.J. (1972). Physiological, anthropometrical, and self-concept changes in over-
weight college women as affected by exercise and voluntary diet. Unpublished master’s
thesis, South Dakota State University.
*Eickhoff, J., Thorland, W., & Ansorge, C. (1983). Selected physiological and psychological
effects of aerobic dancing among young adult women. Journal of Sports Medicine
and Physical Fitness, 23, 273-280.
Field, A. (2001). Meta-analysis of correlation coefficients: A Monte Carlo comparison of
fixed- and random-effects methods. Psychological Methods, 6, 161-180.
Folkins, C.H., & Sime, W.E. (1981). Physical fitness training and mental health. American
Psychologist, 36, 373-389.
*Follansbee, P.A. (1983). Effects of a self-defense program on women’s psychological
health and well-being (Doctoral dissertation, Southern Illinois University, Carbondale,
1982). DAI, 43, B2388.
*Ford, H.T., Puckett, J.R., Blessing, D.L., & Tucker, L.A. (1989). Effects of selected physi-
cal activities on health-related fitness and psychological well-being. Psychological
Reports, 64, 203-208.
*Ford, H.T., Puckett, J.R., Reeve, T.G., & Lefavi, R.G. (1991). Effects of selected physical
activities on global self-concept and body-cathexis scores. Psychological Reports, 68,
1339-1343.
Fox, K.R. (1997). The physical self: From motivation to well-being. Champaign, IL: Human
Kinetics.
Fox, K.R. (2000). Self-esteem, self-perceptions and exercise. International Journal of Sport
Fox, K.R., & Corbin, C.B. (1989). The Physical Self-Perception Profile: Development and
preliminary validation. Journal of Sport & Exercise Psychology, 11, 408-430.
*Fulton, G.B. (1968). Study of the relationship of body image to selected physical abilities:
A phenomenonological approach to analysis and modification (Doctoral dissertation,
University of Maryland, 1967). DAI, 28, A3486.
*Garofoli, E. (1990). The influence of a walking program on the body cathexis and self-esteem
of adult sedentary women. Unpublished master’s thesis, Springfield College.
*Gary, V., & Guthrie, D. (1972). The effect of jogging on physical fitness and self-concept
in hospitalized alcoholics. Quarterly Journal of Studies on Alcohol, 33, 1073-1078.
Gauvin, L., Spence, J.C., & Anderson, S. (1999). Exercise and psychological well-being in
the adult population: Reality or wishful thinking? In J.M. Rippe (Ed.), Textbook of
medicine, exercise, nutrition and health (pp. 957-966). Malden, MA: Blackwell.
*Gibbs, T.S. (1987). The effects of a geriatric exercise program on self-concept. Unpublished
master’s thesis, Brigham Young University.
*Gillis, A.J., & Perry, A. (1991). The relationships between physical activity and health-pro-
moting behaviours in mid-life women. Journal of Advanced Nursing, 16, 299-310.
Glass, G.V., McGaw, B., & Smith, M.L. (1981). Meta-analysis in social research. Beverly
Hills, CA: Sage.
*Glassford, J.M. (1993). The effect of exercise on physical fitness, physical self-efficacy,
physical competence, physical acceptance, self-esteem, locus of control, and hope
(Master’s thesis, University of Alberta, 1992). MAI, 31, 995.
*Goldberg, W.G., & Fitzpatrick, J.J. (1980). Movement therapy with the aged. Nursing
Research, 29, 339-346.
Greenberg, J., Solomon, S., Pyszczynski, T., Rosenblatt, A., Burling, J., Lyon, D., Simon,
L., & Pinel, E. (1992). Why do people need self-esteem? Converging evidence that
self-esteem serves an anxiety-buffering function. Journal of Personality and Social
Gruber, J.J. (1986). Physical activity and self-esteem development in children: A meta-
analysis. In G.A. Stull & H.M. Eckert (Eds.), Effects of physical activity on children
(pp. 30-48). Champaign, IL: Human Kinetics.
*Gruen, C.T. (1991). The effect of a walking program upon self-concept in male patients in
treatment for chemical dependency (Doctoral dissertation, University of San Francisco,
1990). DAI, 52, A115.
*Hannaford, C.P. (1984). The psychophysiological effects of a running program on depres-
sion, self-esteem, and anxiety (Doctoral dissertation, North Texas State University,
1983). DAI, 44, B3527.
Harter, S. (1999). The construction of the self: A developmental perspective. New York:
Guilford Press.
Hattie, J. (1992). Self-concept. Hillsdale, NJ: Erlbaum.
*Hayden, R.M., Allen, G.J., & Camaione, D.N. (1986). Some psychological benefits resulting
from involvement in an aerobic fitness from the perspectives of participants and knowl-
edgeable informants. Journal of Sports Medicine and Physical Fitness, 26, 67-76.
Heaps, R.A. (1978). Relating physical and psychological fitness: A psychological point of
view. Journal of Sports Medicine and Physical Fitness, 18, 399-408.
Hedges, L.V. (1994). Fixed effects models. In H. Cooper & L.V. Hedges (Eds.), The handbook
of research synthesis (pp. 285-299). New York: Russell Sage Foundation.
Hedges, L.V., & Olkin, I. (Eds.) (1985). Statistical methods for meta-analysis. Orlando,
FL: Academic Press.
Hedges, L.V., & Vevea, J.L. (1998). Fixed- and random-effects models in meta-analysis.
Psychological Methods, 3, 486-504.
*Henderson, J.M. (1974). The effect of physical conditioning on self-concept in college females
(Doctoral dissertation, Washington State University, 1974). DAI, 35, B3063.
*Henning, J.M. (1987). The effect of a regular exercise program on the self-concept and self-
actualization of geriatric persons (Doctoral dissertation, United States International
*Hilyer, J.C., & Mitchell, J.W. (1979). Effect of systematic physical fitness training com-
bined with counseling on the self-concept of college students. Journal of Counseling
*Horowitz, S.M. (1985). Job stress: Correlates, differences, and intervention effects on health,
attitude and lifestyle behavior in university faculty and administrators (Doctoral dis-
sertation, The University of Michigan, 1984). DAI, 45, A3577.
*Howse, R. (2002). The effects of a resistance training program on selected psychological vari-
ables in sedentary females (Master’s thesis, University of Alberta, 1986). MAI, 40.
Hughes, J.R. (1984). Psychological effects of habitual aerobic exercise: A critical review.
Preventive Medicine, 13, 66-78.
*Hurley, M.F. (1971). The effects of basic swimming instruction upon self-concept. Unpub-
lished master’s thesis, University of Montana.
James, W. (1950). The principles of psychology. New York: Dover. (Original work publ.
1890)
*Jette, N.M. (1976). The effect of modern dance and music on body image and self concept in
college women (Master’s thesis, Brigham Young University, 1975). DAI, 37, A178.
*Jones, A.T. (1985). The effects of jazz dance and aerobic dance on self-concept. Completed
Research in Health, Physical Education, Recreation, and Dance, 27, 598.
*Jones, M.R. (1988). The effect of participation in a wellness program on cardiovascular
function and self-concept in middle aged men and women. Unpublished master’s
thesis, South Dakota State University.
Katzmarzyk, P.T., Gledhill, N., & Shephard, R.J. (2000). The economic burden of physical
inactivity in Canada. Canadian Medical Association Journal, 163, 1435-1440.
*Kelsey, C.W. (1973). Therapeutic recreation and self-concepts of selected psychiatric
patients. Unpublished master’s thesis, University of Utah.
*Klein, M.H., Greist, J.H., Gurman, A.S., Neimeyer, R.A., Lesser, D.P., Bushnell, N.J., &
Smith, R.E. (1985). A comparative outcome study of group-psychotherapy vs exercise
treatments for depression. International Journal of Mental Health, 13, 148-177.
*Knittle, V.R. (1988). The effect of twelve weeks of exercise on depression, self-esteem, and
locus-of-control in a wellness program sample of women aged 65 and older (Master’s
thesis, University of Arizona). MAI, 26, 380.
*Kofsky, R.D. (1983). Skill mastery and self-esteem: A comparison of physical condition-
ing with other skills (Doctoral dissertation, Ball State University, 1982). DAI, 43(7),
B2317-B2318. (UMI No. 8227594)
Kraemer, H.C., Stice, E., Kazdin, A., Offord, D., & Kupfer, D. (2001). How do risk factors
work together? Mediators, moderators, and independent, overlapping, and proxy risk
factors. American Journal of Psychiatry, 158, 848-856.
*Kulinovich, B. (1983). The effect of exercise on the self-concept of physically disabled
elderly (Master’s thesis, University of Arizona, 1982). MAI, 21, 77.
Landers, D.M., & Arent, S.M. (2001). Physical activity and mental health. In R.N. Singer,
H.A. Hausenblas, & C.M. Janelle (Eds.), Handbook of sport psychology (2nd ed., pp.
740-765). New York: John Wiley & Sons.
*Lauffer, P.L. (1988). The effect of gender, gender role, and weight training on self-concept
and body cathexis (Master’s thesis, University of Arizona, 1987). MAI, 26(1), 24.
(UMI No. 1331411)
*Lee, J.M. (1989). An investigation of the effects of selected physical education activities
upon self-concept of college students (Doctoral dissertation, East Texas State Uni-
versity, 1988). DAI, 49(10), A2963. (UMI No. 8827268)
*Li, F., Harmer, P., Chaumeton, N.R., Duncan, T.E., & Duncan, S.C. (2002). Tai Chi as
a means to enhance self-esteem: A randomized controlled trial. Journal of Applied
Gerontology, 21, 70-89.
Long, B.C., & van Stavel, R. (1995). Effects of exercise training on anxiety: A meta-analysis.
Journal of Applied Sport Psychology, 7, 167-189.
Lou, Y., Abrami, P.C., Spence, J.C., Poulsen, C., Chambers, B., & d’Apollonia, S. (1996).
Within-class grouping: A meta-analysis. Review of Educational Research, 66, 423-458.
*Lu, F.J.H. (2003). Effects of 10 weeks’ weight training on female college students’ physical
self and mood. Bulletin of Sport and Exercise Psychology of Taiwan, 3, 111-125.
Marsh, H.W. (1997). The measurement of physical self-concept: A construct validation
approach. In K.R. Fox (Ed.), The physical self: From motivation to well-being (pp.
27-58). Champaign, IL: Human Kinetics.
Martin, K.A., & Sinden, A.R. (2001). Who will stay and who will go? A review of older
adults’ adherence to randomized controlled trials of exercise. Journal of Aging and
Physical Activity, 9, 91-114.
*May, B.E. (1979). Effects of cardiovascular training on self-concept. Completed Research
in Health, Physical Education, and Recreation, 21, 40.
McArdle, W.D., Katch, F.I., & Katch, V.L. (1986). Exercise physiology: Energy, nutrition,
and human performance (2nd ed.). Philadelphia: Lea & Febiger.
McAuley, E., Blissmer, B., Katula, J., Duncan, T.E., & Mihalko, S.L. (2000). Physical
activity, self-esteem, and self-efficacy relationships in older adults: A randomized
controlled trial. Annals of Behavioral Medicine, 22, 131-139.
McAuley, E., Mihalko, S.L., & Bane, S.M. (1997). Exercise and self-esteem in middle-aged
adults: Multidimensional relationships and physical-fitness and self-efficacy influences.
Journal of Behavioral Medicine, 20, 67-83.
*McAuley, J.W., Long, L., Heise, J., Kirby, T., Buckworth, J., Pitt, C., Lehman, K.J., Moore,
J.L., & Reeves, A.L. (2001). A prospective evaluation of the effects of a 12-week
outpatient exercise program on clinical and behavioral outcomes in patients with
epilepsy. Epilepsy & Behavior, 2, 592-600.
*McConnell, J.A. (1989). Effects of movement training on body awareness, self-concept,
and antisocial behavior in forensic psychiatric patients (Doctoral dissertation, The
Ohio State University, 1988). DAI, 49(9), B4066. (UMI No. 8824568)
McDonald, D.G., & Hodgdon, J.A. (1991). Psychological effects of aerobic fitness training.
New York: Springer.
Mead, G.H. (1934). Mind, self, and society from the standpoint of a social behaviorist.
Chicago: University of Chicago Press.
*Melnick, M.J., & Mookerjee, S. (1991). Effects of advanced weight training on body-cathexis
and self-esteem. Perceptual and Motor Skills, 72, 1335-1345.
*Metcalfe, L.L. (1980). The effect of a ten week parcours exercise program on self-concept,
body-cathexis, life and job satisfaction of state employees. Unpublished master’s
thesis, Florida State University.
*Millett, H.L. (1974). An evaluation of social and physical changes in social dance stu-
dents at Brigham Young University. Unpublished master’s thesis, Brigham Young
University.
*Mock, V., Burke, M.B., Sheehan, P., Creator, E.M., Winningham, M.L., McKenney-Tedder,
S., Schwager, L.P., & Liebman, M. (1994). A nursing rehabilitation program for
women with breast cancer receiving adjuvant chemotherapy. Oncology Nursing
Forum, 21, 899-907.
*Moore, K.A. (1995). The effect of participation in an aerobic exercise program on the
self-esteem of older adults (Doctoral dissertation, Kent State University, 1995). DAI,
56(6), B3429. (UMI No. 9536673)
Morgan, W.P. (1985). Affective beneficence of vigorous physical activity. Medicine and
Science in Sports and Exercise, 17, 94-101.
*Muenchow, J.B. (1990). The effect of exercise and sport activities on the psychological state
and trait tendencies of college-age nonparticipants (Doctoral dissertation, University
of Wyoming, 1990). DAI, 51(6), A1951. (UMI No. 9030617)
*Muller, S.M. (1993). The effects of exercise-induced changes on self-esteem (Doctoral
dissertation, University of Maryland, College Park, 1993). DAI, 54, A1688.
North, T.C., McCullagh, P., & Tran, Z.V. (1990). Effect of exercise on depression. Exercise
and Sport Sciences Reviews, 18, 379-415.
*Norvell, N., Martin, D., & Salmon, A. (1991). Psychological and physiological benefits of
passive and aerobic exercise in sedentary middle-aged women. Journal of Nervous

and Mental Disease, 179, 573-574.
O’Donoghue, R.K. (1990). A causal analysis of the interrelationship among physical fit-
ness, body esteem, perception of fitness and self-concept for female participants of
noncompetitive sport (Doctoral dissertation, University of Southern California, 1989).
DAI, 50(8), A2424-A2425
Oja, P. (1995). Descriptive epidemiology of health-related physical activity and fitness.
Research Quarterly for Exercise and Sport, 214, 303-312.
*O’Neill, M.P. (1990). Physical self-concept and psychological mood states: Their relation-
ship to aerobic exercise (Doctoral dissertation, Washington State University, 1989).
DAI, 50(10), B4780. (UMI No. 9007803)
Overall, J.E., & Doyle, S.R. (1994). Estimating sample sizes for repeated measures designs.
Controlled Clinical Trials, 15, 100-123.
Overton, C.O. (1998). A comparison of fixed-effects and mixed (random-effects) models for
meta-analysis tests of moderator variable effects. Psychological Methods, 3, 354-379.
*Palmer, J.H. (1985). Some psychological and physiological effects of aerobic exercise
on adult inpatient alcoholics (Doctoral dissertation, University of North Carolina at
Greensboro, 1985). DAI, 45, B1125.
*Palmer, L.K. (1995). Effects of a walking program on attributional style, depression, and
self-esteem in women. Perceptual and Motor Skills, 81(3 Pt 1), 891-898.
*Pavett, C.M., Butler, M., Marcinik, E.J., & Hodgdon, J.A. (1987). Exercise as a buffer
against organizational stress. Stress Medicine, 3, 87-92.
*Perri, S., & Templer, D.I. (1984-1985). The effects of an aerobic exercise program on
psychological variables in older adults. International Journal of Aging and Human
Development, 20, 167-172.
Petruzzello, S.J., Landers, D.M., Hatfield, B.D., Kubitz, K.A., & Salazar, W. (1991). A meta-
analysis on the anxiety-reducing effects of acute and chronic exercise: Outcomes and
mechanisms. Sports Medicine, 11, 143-182.
Plante, T.G. (1999). Could the perception of fitness account for many of the mental and physi-
cal health benefits of exercise? Advances in Mind-Body Medicine, 15, 291-301.
Plante, T.G., Coscarelli, L., Caputo, D., & Oppezzo, M. (2000). Perceived fitness predicts
daily coping better than physical activity or aerobic fitness. International Journal of
Stress Management, 7, 181-192.
Plante, T.G., LeCaptain, S.E., & McLain, H.C. (2000). Perceived fitness predicts daily coping
better than physical activity. Journal of Applied Biobehavioral Research, 5, 66-79.
*Plummer, O.K., & Koh, Y.O. (1987). Effect of “aerobics” on self-concept of college women.
Perceptual and Motor Skills, 65, 271-275.
*Pohl, C.M. (1985). The effect of a 12-week aerobic dance class on body image, self-image,
and fitness in female college students. Completed Research in Health, Physical Educa-
tion, Recreation, and Dance, 27, 195.
*Querry, P.A. (1998). Effects of aerobic exercise training and cognitive functioning in aging
adults (Doctoral dissertation, Oklahoma State University, 1998). DAI, 59, A2904.
*Rainey, D.W., & Wigtil, J. (1985). Aerobic running as a counseling technique for under-
graduates with low self-esteem. Journal of College Student Personnel, 26, 53-57.
Ray, J.W., & Shadish, W.R. (1996). How interchangeable are different estimators of effect
size? Journal of Consulting and Clinical Psychology, 64, 1316-1325.
*Ready, A.E., Fitzpatrick, D.W., Boreskie, S.L., & Hrycaiko, D.W. (1991). The response of
obese females to low impact exercise and diet counselling. Journal of Sports Medicine
and Physical Fitness, 31, 587-595.
Rejeski, W.J. (1994). Dose-reponse issues from a psychosocial perspective. In C. Bouchard,

R.J. Shephard, & T. Stephens (Eds.), Physical activity, fitness, and health: Interna-
tional proceedings and consensus statement (pp. 1040-1055). Chapaign, IL: Human
Kinetics.
*Rhodes, D.L. (1980). Mens sana, corpore sano: A study of the effect of jogging on depres-
sion, anxiety, and self-concept (Doctoral dissertation, Duke University, 1980). DAI,
41(4), A1500. (UMI No. 8022219)
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton
University Press.
Rosenthal, R. (1995). Writing meta-analytic reviews. Psychological Bulletin, 118, 183-192.
*Rubin, B.C. (1989). Comprehensive weight loss employing social support and aerobic
exercise (Doctoral dissertation, University of Tennessee, 1988). DAI, 50(3), B1120.
(UMI No. 8911757)
*Rudner, R.A. (1980). The effects of jogging and assertiveness training on self variables
and assertiveness in women (Doctoral dissertation, University of Florida, 1979). DAI,
40, A4416.
*Samuels, K.L. (1987). An examination of childbirth training: Changes in self-efficacy,
anxiety, and body image (Doctoral dissertation, California School of Professional
Psychology, 1987). DAI, 48, B1162.
*Schmidt, R J. (1989). The effects of Shotokan karate training on self-concept and eastern
cultural value orientation of college men and women (Doctoral dissertation, University
of Nebraska, 1988). DAI, 49(10), A2952-A2953. (UMI No. 8824952)
*Scudder, W.W. (1978). A study of a physical training program and self-concept change
in alcoholics. Completed Research in Health, Physical Education, Recreation, and
Dance, 20, 347.
*Secrist, M.C. (1987). The effects of exercise on self-esteem and morale in the frail el-
derly (Master’s thesis, The Ohio State University, 1986). MAI, 25(3), 290. (UMI No.
1329377)
Sedlmeier, P., & Gigerenzer, G. (1989). Do studies of statistical power have an effect on the
power of studies? Psychological Bulletin, 105, 309-316.
*Segar, M.L., Katch, V.L., Roth, R.S., Garcia, A.W., Portner, T.I., Glickman, S.G., Haslanger,
S., & Wilkins, E.G. (1998). The effect of aerobic exercise on self-esteem and depres-
sive and anxiety symptoms among breast cancer survivors. Oncology Nursing Forum,
25, 107-113.
Shavelson, R.J., Hubner, J.J., & Stanton, G.C. (1976). Self-concept: Validation of construct
interpretations. Review of Educational Research, 46, 407-441.
*Shaw, J.M., Ebbeck, V., & Snow, C.M. (2000). Body composition and physical self-concept
in older women. Journal of Women and Aging, 12(3-4), 59-75.
*Shawver, C.A. (1979). The effects of participation in selected risk college physical educa-
tion classes on a student’s self-concept (Doctoral dissertation, Middle Tennessee State
University, 1978). DAI, 39(10), A6025. (UMI No. 7907712)
*Sheppard, S. (1972). Changes in body concept and self-concept among college students who
learn to swim (Doctoral dissertation, New York University, 1971). DAI, 32, A5599.
*Short, M.A., DiCarlo, S., Steffee, W.P., & Pavlov, K. (1984). Effects of physical condition-
ing on self-concept of adult obese males. Physical Therapy, 64, 194-198.
Sonstroem, R.J. (1984). Exercise and self-esteem. Exercise and Sport Sciences Reviews,
12, 123-155.
Sonstroem, R.J. (1997a). Physical activity and self-esteem. In W. Morgan (Ed.), Physical
activity and mental health. Washington, DC: Taylor & Francis.
Sonstroem, R.J. (1997b). The physical self-system: A mediator of exercise and self-esteem.
In K.R. Fox (Ed.), The physical self: From motivation to well-being (pp. 3-26).
Champaign, IL: Human Kinetics.
Sonstroem, R.J. (1998). Physical self-concept: Assessment and external validity. Exercise
and Sport Sciences Reviews, 28, 133-164.
Sonstroem, R.J., & Morgan, W.P. (1989). Exercise and self-esteem: Rationale and model.
Medicine and Science in Sports and Exercise, 21, 329-337.
*Sorensen, M., Anderssen, S., Hjerman, I., Holme, I., & Ursin, H. (1997). Exercise and diet
interventions improve perceptions of self in middle-aged adults. Scandinavian Journal
of Medicine & Science in Sports, 7, 312-320.
Spence, J.C., & Blanchard, C. (2001). Publication bias in sport and exercise psychology
research: The games we play. International Journal of Sport Psychology, 32, 386-
399.
*Stefani, U. (2002). The effects of a physical conditioning program on the physical fitness and
self-concept of elderly women (Master’s thesis, McGill University, 1988). MAI, 40.
*Stein, P.N., & Motta, R.W. (1992). Effects of aerobic and nonaerobic exercise on depres-
sion and self-concept. Perceptual and Motor Skills, 74, 79-89.
*Stewart, A.L., Mills, K.M., Sepsis, P.G., King, A.C., Mclellan, B.Y., Roitz, K., & Ritter,
P.L. (1997). Evaluation of CHAMPS, a physical activity promotion program for older
adults. Annals of Behavioral Medicine, 19, 353-361.
*Teng, L.N. (1982). The relationship of self-esteem to satisfaction with social supports,
self-report of socially valued resources, and interpersonal reciprocity before and after
participation in a systematic aerobic dance exercise program (Doctoral dissertation,
University of Washington, 1982). DAI, 43(6), B2008. (UMI No. 8226608)
*Titzel, M.R. (1986). The effects of a dance-movement program on the self-concept of
institutionalized, intermediate care, female elderly. Unpublished master’s thesis, The
Pennsylvania State University.
*Tseo, C. (1983). The effect of an aerobic dance program on the self-concepts and physi-
cal fitness levels of young adult institutionalized female mental patients. Completed
Research in Health, Physical Education, Recreation, and Dance, 25, 178.
*Tucker, L.A. (1982). Effect of weight-training program on the self-concepts of college
males. Perceptual and Motor Skills, 54, 1055-1061.
*Tucker, L.A. (1983). Weight training: A tool for the improvement of self and body concepts
of males. Journal of Human Movement Studies, 9, 31-37.
U.S. Dept. of Health and Human Services. (1996). Physical activity and health: A report
of the Surgeon General. Atlanta: U.S. Dept. HHS, Centers for Disease Control and
Prevention, National Center for Chronic Disease Prevention and Health Promotion.
*Weinstein, J.B. (1986). The influences of physical fitness on mood and self-esteem in a
young adult population: A learned efficacy model (Doctoral dissertation, State Uni-
versity of New York at Albany, 1986). DAI, 47, B2637-2638.
*Wescott, R.P. (1980). The effect of elective physical education classes on the self-concept of
college students (Doctoral dissertation, Oklahoma State U., 1980). DAI, 41, A3482.
*White, A.J. (1974). The interrelationships between measures of physical fitness and mea-
sures of self-concept of selected Mississippi State University male students (Doctoral
dissertation, Mississippi State University, 1973). DAI, 34, A4849.
*Whiting, J.W. (1981). The effect of a mild exercise program on the psychological treatment
of inpatient alcoholics (Doctoral dissertation, Fuller Theological Seminary, School
of Psychology, 1980). DAI, 42, B394-B395.
*Williams, D.D. (1972). Instruction stressing physical fitness compared with instruction
stressing skill acquisition upon self-concept in Caucasian and Negro students (Doctoral
dissertation, University of Arkansas, 1972). DAI, 33, A2156.
*Yeagle, G.W. (1982). The effect of a physical recreation program on the self-concept and
flexibility of senior citizens (Doctoral dissertation, University of Utah, 1981). DAI,
42, B3184.
Notes
1 When exploring the self in the physical domain, the terms self-esteem and self-con-
cept tend to be used interchangeably; there is no one unified, operational definition per se.
McDonald and Hodgdon (1991) used the two terms in a unified form, and for the purposes
in this paper so do we.
2 According to Caspersen et al. (1985), physical fitness consists of both health related
components (cardiorespiratory endurance, muscular endurance, muscular strength, body

composition, flexibility) and performance related components (muscular speed, speed,
agility, balance, reaction time).
3 An interesting phenomenon in this meta-analysis was that approximately two-thirds
(n = 80) of the effect sizes came from unpublished sources. This is a much greater proportion
of unpublished effect sizes than is typically seen in meta-analyses. In fact, the majority of
meta-analyses in sport and exercise psychology include no unpublished works (see Gauvin
et al., 1999).
4 Rosenthal (1995) does not agree that moderator analyses should only proceed after
the detection of significant heterogeneity of variance among a set of effect sizes. He argues
that scientific progress can result from scientists “continually reducing the magnitude of
sampling error by increasing their understanding of moderator variables” (p. 186). While
Overton (1998) claims such exploratory analyses are justified in meta-analyses as long as
they are preceded by homogeneity testing and conducted on a priori hypothesized modera-
tors, Rosenthal states that “planned contrasts should be computed without reference to the
overall F, and even unplanned contrasts can be computed with appropriate adjustments of
their levels of significance” (p. 188).
5 Lifestyle programs included exercise along with stress reduction and nutrition
counseling. Skills training activities were those in which the goal was primarily to develop
competence in a skill as opposed to enhancing physical fitness. Such activities included
martial arts and physical education courses.
6 One reviewer of this paper inquired about the statistical power of our analysis.
According to Field (2001), we have sufficient power, based both on number of studies (k
= 118) and average sample size (n = 61), at an overall level to detect any significant effect
sizes. True, we may be underpowered in several levels of our moderator variables. However,
none of the effect sizes for these levels is greater than d = .15. Thus we do not think we are
missing any substantial effects in those categories due to low power.
Acknowledgment
We express our appreciation to Trish Dyck who provided assistance with the study.
Manuscript submitted: June 7, 2004

Revision accepted: March 28, 2005

Jsep Article p311

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

Jsep Article p311

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Jsep Article p311

Uploaded by

Copyright:

Available Formats

EXERCISE PSYCHOLOGY

JOURNAL OF SPORT & EXERCISE PSYCHOLOGY, 2005, 27, 311-334

The Effect of Exercise on Global Self-Esteem:

John C. Spence1, Kerry R. McGannon2, and Pauline Poon1

Key Words: meta-analysis, self-concept, physical activity

If people experience a more favorable view of themselves as a result of par-

Finally, McDonald and Hodgdon combined different aspects of “self” such

Hypothesized Moderators of Effect of Exercise on Self-Esteem

Hypothesized Moderators of Exercise on Self-Esteem

Additional Moderators of Exercise on Self-Esteem

Table 1 Summary Statistics for Hypothesized Moderators of Exercise and Self-

Initial fitness level 3.10 141.04

Note: CI = confidence interval; Qb = measure of between-group effect; k = number of

Methodological Variables. None of the methodological variables served as

Table 2 Summary Statistics for Additional Moderators of Exercise and Self-Esteem

Study quality 5.29 139.10

Note: CI = confidence interval; Qb = measure of between-group effect; k = number of

Table 3 Prevalence of Study Quality Among Published and Unpublished Studies

*** < .001

Sample Size and Power Issues

Effects of the Exercise Program

Limitations of This Review

Conclusions and Future Directions

cal or practical significance of such changes. Further, it appears that increases in

dissertation, Syracuse University, 1995). Dissertation Abstracts International, 57(4),

passive and aerobic exercise in sedentary middle-aged women. Journal of Nervous

Rejeski, W.J. (1994). Dose-reponse issues from a psychosocial perspective. In C. Bouchard,

components (cardiorespiratory endurance, muscular endurance, muscular strength, body

Manuscript submitted: June 7, 2004

You might also like