10 1111j 1744 6171 1995 tb00519 x

Maintenance of Self-Esteem by Obese Children

Helen E. Manus, MSN, RN, and Maureen R. Killeen, PhD, RN

Forty-five fifih grade children participated in a
study to examine whether obese children differ in
self-esteem from their average-weight peers. The
study also explored whether obese children use the
processes of discounting and cognitive distortion to
maintain self-esteem. The obese children had lower
scores than average-weight children in global selfworth, and in perceived competence in physical
appearance and social acceptance, but not in scholastic competence, athletic competence, or behavioral
conduct. The pattern of results indicate that obese
children do not discount the importance of physical
appearance and zueight, but use distortion. Further,
the use of distortion by obese children was associated
with more positive self-perceptions. Although the
use of defensive cognitions was effective in enhancing obese children’s global self-worth, it was not
powerjkl enough to prevent completely the negative
effectsof obesity on self-esteem.
Key words: Obesity, overweight, self-esteem

JCAPNVolume 8,Number 1,January- March, 1995

Helen E . Manus, M S N , RN, is afilitated with Scottish
Rite Children’s Medical Center, Atlanta, GA. Maureen R.
Killeen, PhD, RN, is Professor of Mental HealthPsychiatric Nursing, Athens, GA.

o b e s i t y is a major health problem affecting increasing
numbers of school-aged children, and often persisting
into adulthood (Obese Children, 1986). Although the

physical health risks associated with obesity have been
well-documented (Koniz- Booher & Koniz, 1986; Weil,
19811, less attention has been given to the mental health
risks. Among the potential psychosocial consequences of
obesity in childhood are social rejection (Staffieri, 1967)
and low self-esteem (McCarthy, 1986).
Self-esteem is also an important component of health,
affecting numerous aspects of mental health and behavior (Roy & Andrews, 1991). Persons with high selfesteem have greater social adjustment, less anxiety and
worry, and a lower incidence of psychosomatic symptoms (Mack, 1983). When persons value the self and
have confidence in their abilities, they engage in activities that are growth-producing and increasingly become
competent (Bandura, 1986).Conversely, low self-esteem
has been linked to decreased motivation and performance (Bandura), as well as emotional disturbances and
psychopathology (Rosenberg, 1986). Among children,
low self-esteem has been associated with depression, and
suicidal ideas (Hammen & Zupan, 1984; Harter, Marold,
& Whxtesell, 1992; Zupan, Hammen, & Jaenicke, 1987).

Obesity and Self-Esteem
Although many factors may influence self-esteem,
physical appearance, particularly one’s weight, is
thought to be a major contributor. Self-perceptions of
physical appearance are more highly correlated with
overall self-esteem than are perceptions of self-competence in scholastics, athletics, social acceptance, or behavior (Harter, 1985; Killeen, 1993). Although this link
between obesity and the self-esteem of individuals, especially adolescents, is often cited (Allon, 1979; Harrison,
17


Maintenance of Self-Esteem by Obese Children

1983; Kizziar & Hagedorn, 1979; Mendelson & White,

1985),several gaps exist in the research literature. For
example, few studies have actually examined the relationship between obesity and children’s self- esteem.
Further, the conclusions drawn from investigations in
t h area have been contradictory.
Some studies have found that obese children report
lower levels of Self-esteem than do their nonobese peers
(Sallade, 1973; Strauss, Smith, Frame, & Forehand, 1985).
Other studies have either failed to find sigxuficant differences (Bell, Kirkpatrick, & Rim, 1986; Wadden, Foster,
Brownell, & Finley, 1984)or have found equivocal results
(Mendelson & Wkute, 1982; Worsley, 1981).A similar pattern of conflicting results has been found in studies of
obesity and other indicators of children’s and adolescents’ self-perceptions, such as body esteem and total
esteem (Hendry & Gillies, 1978; Mendelson & White,
1982; 1985; 19911, and distortion of body- image (Bell et
al.; Leon, Bemis, Melland, & Nussbaum, 1978).
-

have not addressed these issues. For example, the measurement tools used in most studies have been criticized
for being &dimensional, or for measuring global selfesteem by summing the scores on different areas of performance (Harter, 1986; Rosenberg, 1979), and for not
addressing developmental differences among various
age groups (Harter, 1985). Unidimensional measures,
and those that sum over different areas of self-concept,
tend to mask both the complexity of self-perceptionsand
ddferences in patterns of self- perceptions between obese
and average-weight children. Virtually no research has
examined whether obese children who exhibit low selfesteem differ from high self-esteem obese children, or
average-weight cluldren, in their patterns of self-perceptions or in the cognitive defenses they use to maintain
self-esteem.
The author’s purpose was to address these issues by
exploring not only whether differences in self-esteem
exist between average-weight and obese children, but

also whether children use cognitive defenses to maintain
high levels of self-esteem.

~

Virtually no research has examined

Maintaining Self-Esteem

whether obese children who exhibit low

Studies by Harter (1986) and Tesser (1988) provide
possible explanations for the findings that some obese
children maintain high self-esteem while others do not.
Both Harter and Tesser view self-evaluation as a multidimensional construct that is related not only to performance in different areas, but also to the importance the
person places on performing well in a given domain.
Harter has suggested that overall self-esteem or global
self-worth (GSW) is a function of the discrepancy
between one’s perceptions of competence in specific areas
and the importance of competence in each area. Her studies suggest that if a chdd is deficient in one domain, he or
she may use a cognitive process, such as discounting or
distortion, to maintain a high level of self-esteem.
Discounting is the process by which children maintain
GSW by diminishing the importance of domains in
which they are less competent. By discounting, obese
children may maintain high self- esteem by viewing

self-esteem differ from high self-esteem
obese children, or average-weight children,
in their patterns of self-perceptions or in

the cognitive defenses they use to maintain
self-esteem.

Such conflicting findings may be the result of indwidual differences in patterns of self-perceptions among
obese children, or they may be the result of differences in
the methods used to assess self-esteem. Prior studies
18

-

JCAPNVolume 8, Number 1, January March, 1995


physical appearance, specificallyweight, as unimportant,
and placing more importance on other domains in which
they excel. Distortion is the process by which children
enhance their perception of competence. Obese children
who view physical appearance as very important, may
maintain high self- esteem by underestimating - distorting - their actual body size or weight. Both discounting and distortion may allow chddren to maintain
high GSW by keeping the discrepancy between perceived competence and importance small. By distorting,
children inflate their competence scores, while by discounting they deflate their importancescores.
Several hypotheses for this study were suggested by
the processes just described. First, it was predicted that
obese, overweight, and average-weight chddren would
differ in their self-perceptions only in areas that were
related to or affected by weight, such as physical appearance and global self-worth, but not scholastic competence or behavioral conduct. No specific predictions
were made concerning the effects of obesity on athletic
competence or social acceptance.
Second, predictions were made concerning the processes by which children maintain self-esteem. In general,
it was expected that obese children would attempt to

enhance their GSW either by discounting the importance
of physical appearance and weight, or by distorfing their
perceptions of weight, and that these processes would be
effective. The discounting hypothesis predicted that GSW
scores would be negatively related to the importance of
physical appearance and weight, for obese and overweight children in the sample. This hypothesis also predicted that perceived competence and importance of
physical appearance and weight would be positively correlated for all children in the sample. It was further
expected that obese children with high self-esteem would
rate physical appearance and weight as less important
than would obese children with low self-esteem.
The distortion hypothesis predicted that weight, the
level of GSW, and the importance of physical appearance
would interact to influence the level of distortion of
weight. Specifically,it was hypothesized that obese children who had high global self- worth and for whom
JCAPNVolume 8, Number 1, January- March, 1995

physical appearance was important would distort their
perception of weight more than those with low GSW,
those for whom physical appearance was not important,
and average- weight children.
Finally, it was expected that distortion would be an
effective process for improving self-esteem. It was predicted that, for obese children, GSW would be positively
related to the degree of distortion of weight perception.

It was expected that distortion would be an

effective process for improving self-esteem.

Methods
Research Participants

A convenience sample was drawn from approximately 120 students in a fifth-grade class in a rural
Southeastern elementary school. Fifth-graders were chosen because of the importance of physical appearancefor
children age 10-12, their ability to compare themselves
with others, and their emphasis on classmates as sources
of information and support (Harter, 1990).In order to
participate, children had to be present on the day of data
collection, had to have given written assent, and had to
have written, informed, parental consent. No monetary
or other incentive was provided. A final sample of 45
students (21 boys, 24 girls), aged 10 to 13 (M = 11.1 yrs.,
S D = 0.7) met these criteria.

Instruments
Self-perceptions.GSW and perceptions of competence in specific domains were measured using the
Self-Perception Profile for Children (SPPC) (Harter, 1985).
The measure was designed to be multidimensional
and age-appropriate and to allow for GSW to be rated
19


Maintenance of Self-Esteem by Obese Children

independently of ratings in specific domains. The SPPC
consists of 36 items in a structured-alternative format.
Five subscales (scholastic competence, social acceptance,
athletic competence, physical appearance, and behavioral conduct) tap children’s domain-specific judgments
of competence. A sixth subscale measures GSW. Scores
on each item range from one to four. Individual item ratings are averaged to yield separate subscale scores.
Harter (1985) assessed reliability and validity of the
SPPC by administering the scale to 1,500 t h r d through

eighth-grade children in four separate samples in
Colorado. Children in these samples were primarily from
lower middle class to upper middle class neighborhoods;
90% were Caucasian. Internal consistency was assessed
using Cronbachs alpha, and was acceptable. Coefficients
for the SPPC subscales ranged from .71 for behavioral
conduct to 3 2 for scholastic competence. Physical appearance internal consistency ranged from .76 to 3 0 among
lifth graders, whde the alpha for GSW was .78.
Harter (1985)assessed validity using factor analysis to
determine whether the five specific subscales constituted
separate factors and could be replicated across samples.
Each subscale defined its own factor; factor loadings for
each subscale were substantial (average greater than .55),
with negligble cross-loadings (average between .04and
.08, none greater than ,181.
Four items were added to the SPPC in order to obtain
additional data on the children’s physical self-perceptions and to serve as a measure of distortion. These items
were written in the identical format to the other SPPC
items. Two of these items tapped perceived competence
in weight, and were used in this study. Validity of the
weight items was assessed by examining their intercorrelations with the items assessing perceived competence in
the other specific domains and their correlation with the
child’s actual weight (measured as a percentage of the
median weight for age, sex, and height). Perceived competence in weight and actual weight were significantly
correlated ( r = -.&, p < .001), as were perceptions of competence in weight and in physical appearance ( r = .49, p
< .001). These correlations indicate the constructs overlapped substantially, but that they were not identical.
20

Finally, the patterns of correlations between perceived
competence in weight and the other domains were identical to the patterns of relations between perceived competence in physical appearance and the other domains.

Physical appearance and weight competence scores were
sigruficantly correlated with each other and with social
acceptance, but not with scholastic, athletic, or behavioral competence scores.
Importance of specific domains. The perceived
importance of competence in each of the speclfic domains
of the SPPC was assessed using Harter’s Importance
Scale. T ~ 10-item
E
scale also uses a four-point structuredalternative format. Two items were added to gain additional data on the importance cluldren place on the areas
of height and weight. Each of these items was worded
identically to the other importance items, with the exception that the word height or ~ u ~ i gwas
k f substituted for
other domain-specific words. The importance of the
weight item was used in this study.
Weight. Each child‘s weight and height were measured using an upright healthometer scale. The scale was
calibrated by w e i g h g an object of known weight (20 lbs.).
The obtained weight and height measurements were used
to class@ children according to weight norms for age, sex,
and height. Chddren were classified as obese (n = 18)if their
weight was at least 20% greater than the median weight
for their age, height, and sex on growth percentile charts
(Hamill, Drizd, Johnson, Reed, & Moore, 1979).Children
whose weight was at least 10% greater than the median
were classified as ozvmeight (17 = lo), while those whose
weight was less than 110% of the medmn were classified as
azwa~e-zu&g~zt
(iz = 17). Only one participant was less than
90%of the median weight for age, sex, and height.
Distortion. In addition to the scores on specific and
global self-perceptions, an additional score was constructed in order to test the hypotheses about distortion.

Distortion of weight was measured by comparing the
children’s perceived competence in weight to their measured weight. First, an actual weight competence score was
derived from the children’s measured weight in relation
to the median weight for their age, sex, and height.
Obese children’s (at least 120% of the median) acfual
JCAPN Volume 8, Number 1,January- March, 1995


weight competence scores were coded as one; overweight
children’s (between 110 and 119 % of the median) scores
were coded as two Children between 100 and 109 percent
of the median weight were coded as three, and those less
than 100% of the median weight were coded as four.
Distortion scores were then constructed by subtracting the
actual weight competence score from the perceived competence in weight score from the SPPC. Higher distortion
scores indicate inflated perceived competence in weight
scores relative to actual weight status. For example, an
obese child (actual weight competence score = 11, who
views herself or himself very positively in weight (perceived competence in weight = 4), would have a distortion score of 3 (4 - 1 = 3). The same child who rates the
self as less competent in weight (perceived competence =
2), would have a distortion score of 1 (2 - 1 = 1).
Procedure

Two weeks prior to the data collection date, the fifthgrade science teacher distributed an informed consent
form and an assent form to each student in science class.
The teacher read written instructions prepared by the
investigator, while distributing the consent and assent
forms. On the day of data collection, children who were
present in science class, and who had signed assent forms
and returned consent forms signed by their parents, were

given the research instruments by the investigator.
Students not participating in the study were sent to
another room to watch a film. The investigator read the
instructions included in the manual to the participants as a
class. After completing and returning the self-reportinstruments, each participant was brought to a separate mom to
be measured and weighed individually and privately. Each
student was weighed wearing school clothes, but without
shoes or jacket. Age and sex were also recorded. Then,
each student was thanked and returned to the classroom.

Results
A preliminary analysis was performed to determine
whether boys and girls differed in self-perceptions on

any of the subscales of the SPPC. Since no sigruficant differences were found, F (1,43) < 1, boys’ and girls’ scores
were combined in all subsequent analyses. To facilitate
comparisons of subscales of the SPPC and other measures, all scores were standardized prior to analyses. To
test the first hypothesis, a 3 (obese, overweight, averageweight) x 6 (perceived competence subscales) mixed
design analysis of variance was performed. A sigruficant
main effect was found for weight status, F(2,42) = 5.181,
p < .01. Obese children generally rated themselves lower
than average-weight or overweight children. Neither the
main effect for the within-subjects factor (SPPC subscales) nor the interaction of weight status and subscales
was sigruficant(both FS < I).
The effect of weight status was then examined for
each subscale separately. Obese children differed from
both overweight and average-weight peers in global selfworth ( F 12,421 = 3.633, p< .04), perceived social competence ( F 12, 421 = 3.259, p < .05), and perceived competence in physical appearance, (F 12,421 = 5.735, p < .007),
but not in athletics, ( F 12,421 = 2.441, p < .lo), scholastics,
or behavior, both (Fs [2,421< I);see Table 1.
Discounting hypothesis. The discounting hypothesis

was supported only for average- weight children. It was
expected that GSW would be negatively related to the
importance of physical appearance and weight for obese
and overweight children. GSW was negatively related to

Table 1. Self-perception Subscale Means (Z Scores) for
Obese, Overweight, and Average-Weight Children
Subscale

Average Overweight Obese
Weight

Scholastic
Social acceptance
Athletic competence
Physical Appearance
Behavioral conduct
Global self-worth

0.00
0.33
0.43
0.49
0.18
0.35

0.33
0.32
-0.04
0.26

0.15
0.32

-0.18
-0.46
-0.36
-0.59
-0.23
-0.49

p
ns
<.05
<.lo
<.01

ns
<.05

21


Maintenance of Self-Esteem by Obese Children

ratings of the importance of physical appearance for
overweight (r = -.39) and average-weight children (r =
-.41), but not for obese children ( Y = .36).GSW was si@icantly negatively related to importance of weight for
average-weight children ( r = -.55), but not for obese (r =
-.27) or overweight cluldren (r = .21). It was further PIPdicted that children’s ratings of competence in physical
appearance and weight would be positively correlated

with their ratings of the importance of those domains.
No relationship was found between ratings of competence and importance of physical appearance ( Y = -.02).
Correlations between competence and importance of
weight were sigruficant,but in the opposite direction to
p < .05).It was also predicted that
that predicted (r = -3,
obese children with high GSW would rate physical
appearance and weight as less important than would
obese children with low self-esteem. High and low
global self-worth children in each weight category were
compared on physical appearance and weight importance scores, using 3 (weight status) x 2 (high, low GSW)
ANOVAs. No sigruficant dlfferences were noted among
groups in their ratings of the importance of physical
appearance or weight (all Fs < 1).Furthermore, in one
domain, the trends for obese children were in the opposite direction from those predicted. Obese chddren with
high global self-worth tended to rate physical appearance as more important than did those with low GSW,
t(16) = 1.69, p = .lo7 (Table2 and Figure 1).
Distortion hypothesis. The distortion hypothesis was
supported. A 3 (obese, overweight, average-weight) x 2
(high, low GSW) x 2 (high, low importance) ANOVA
was performed to determine whether obesity, global selfworth, and importance of physical appearance interacted
to affect distortion. The analysis revealed a main effect
for weight status, ( F 12,331 = 15.269, p < .OOOl).Post-hoc
comparisons revealed that obese children distorted their
perceptions of competence in weight sigxuficantly more
than overweight ( p < ,051 and average weight ( p < .01)
children (Table 3). T h s main effect was modified by a
significant three-way interaction among GSW, weight
status, and importance of physical appearance on distortion of weight, (F [2,331 = 4.889, p < .02). Subsequent 2
22

Table 2. Physical Appearance Subscale Means ( Z Scores)
by Weight Status and Global Self-worth (GSW)
Group

High GSW
- obese
- overweight
- average
weight
Low GSW
- obese
- overweight
- average

n

Perceived
Competence

Importance

4
5
11

-0.32
0.62
0.81

0.75
-0.42
-0.06

14
5
6

-0.67
-0.09
0.11

-0.18
-0.02
0.47

weight

(importance) x 3 (weight status) ANOVAs were performed separately for children with high and low GSW.
The analyses revealed a significant interaction between
importance of physical appearance and weight status on
distortion, for children with high global self-worth ( F
[2,14] = 8.522, p < .004),but not for the children with low
GSW, ( F 12,191 < 1).Despite achieving statistical significance, this interaction involves four cells with three or
fewer subjects, and is therefore difficult to interpret.
Figure 1. Discounting Physical Appearance ( Z Scores)
0.8
0.6

1

Importance Rating

I

0.4
0.2

High GSW

0

L~~ GSW

-0.2
-0.4
-0.6

Obese

Overweight Average-Weight

JCAPN Volume 8, Number 1,January- March, 1995


Table 3. Weight Subscale Means and Distortion Scores
( Z Scores) by Weight Status and Global SelfWorth(GSW)
Group

n

Perceived Importance Distortion
Competence

High GSW
- obese
4
-overweight 5
11
-average
weight
Low GSW
- obese
14
-overweight 5
6
-average
weight

0.21
0.06
0.63

-0.05
0.14
-0.34

1.20
0.22
-0.76

-0.60
-0.13

0.24
-0.26
0.28

0.50
0.05
-0.79

0.21

Sigruhcant main effects were observed for weight status
for both the high, (F [2,14] = 14.82, p < .0006), and low
GSW groups, (F 12,191 = 3.62, p < .05) (Figure2).
Effect of distortion on self-esteem. The final hypothesis examined whether the use of distortion by obese
children was effective in maintaining self-esteem. The
hypothesis was supported. For obese children, GSW
was positively related to distortion ( r = .51, p < .05),
while no relationship was found between these variFigure 2. Distortion of Perceived Weight ( Z Scores)

ables for the entire sample (Y = .03), or for overweight
(r = -.13) or average-weight ( r = 2 1 ) children.
Previous analyses had shown that obese, overweight
and average-weight children differed in GSW.
Consequently, it was possible that the relations between
weight and distortion, and weight and GSW were
responsible for the sigruhcant correlation between distortion and GSW for obese children. A hierarchical regression analysis was performed to determine whether the

distortion score explained unique variance in GSW in
addition to that explained by weight status. Weight (percentage of the median weight for age, sex, and height,
standardized) was entered first, and the standardized
distortion of weight score was entered second. The first
regression equation was not sigruhcant (semipartial r2 =
.01), while the model with distortion scores added was
significant (R2 = .32, p = .056; semipartial r2 = -31,
p < .02), indicating that the distortion score explained a
significant amount of variance in GSW, beyond that
explained by weight alone.

Obese children differ in self-esteem from
their average-weight peers, and obese
children utilize cognitive processes such as
discounting and distortion to maintain

Distortion Scores

1.5

adequate levels of self-esteem.

1

0.5

Discussion

0

This study sought to answer two questions: Do obese
children differ in self-esteem from their average-weight
peers, and do obese children utilize cognitive processes
such as discounting and distortion to maintain adequate
levels of self-esteem?The answer to each question is a

-0.5
-1

-

JCAl"Volume 8, Number 1, January March, 1995

23


Maintenance of Self-Esteem by Obese Children

qualified "yes." The predictions that obese children
would exhibit lower levels of GSW than would their
average weight peers, and that obese chddrrn would use
defensive cognitions to maintain optimum levels of GSW
were generally supported.
Obesity and self-esteem. Contrary to the findings
of some studies examining obesity and self-esteem,
the authors found that obese children have lower levels of global self-worth than either overweight 3r
average weight peers. The authors' results may differ
because they used different instruments, definitions
of self-esteem, and more standardized measures of
obesity. They employed a self-perception scale

designed for use with children that measured global
self-esteem independently of self-perceptions in specific domains.
Examination of the self-competence scores in spechc
domains (Table 1) provides further insight into the role
of obesity in influencing self-perceptions. Obese children
differed from their peers only in domains that would be
expected to be influenced by weight status: physical
appearance and social acceptance, but not scholastics or
behavioral conduct. hfferences in athletics scores were
only marginally sigmhcant. These findings are slmilar to
those of other studies in which social information has
been shown to affect self-perceptions only in specific,
related domains (Harter, 1985).Studies in which self-ratings in all domains are summed to produce a total selfesteem score would tend to mask variations in self-perceptions in specific domains, a n d would reduce
self-esteem to the average across domains. Such studies
also would mask differences between obese and average-weight chddren.
Discounting the importance of physical appearance
and weight. The second purpose of this study was to
examine the processes, specifically discounting and distortion, by whch self-esteem is maintained by obese children. The discounting hypothesis predicted that GSW
would be negatively related to the importance of physical appearance and weight for obese and overweight
children. Additionally, the hypothesis predicted perceived competence and importance of physical appear24

ance and weight would vary directly for all children.
Little support was found for this hypothesis. Results
were consistent with the hypothesis only for averageweight children.
It was further predicted that obese children with high
self- esteem would rate physical appearance and weight
as less important than would obese children with low
self-esteem. Figure 1 illustrates the findings related to
tlus hypothesis. Again, although not sigruficant, the pattern of results was as expected for average-weight and
Overweight cluldren, but not for obese children. In fact,

obese children with high self-esteem rated the importance of physical appearance higher than any of their
peers, regardless of level of GSW, or weight status. The
trends were in the expected direction for importance of
weight, with high self- esteem, obese and averageweight children rating weight as less important than did
their low self-esteem peers.
These results indicate fifth-grade children have difficulty discounting the importance of physical appearance. One explanation for this finding may lie in the
composition of the sample. Sixty-two percent of the
sample were classified as either obese (iz = 18) or overweight ( 1 2 = 10). Physical appearance may have been
the most salient dimension for this sample, and therefore may contribute strongly to their overall perception
of self. The children also knew they would be weighed,
which may have increased the salience of physical
appearance. Finally, others have found that global selfworth is most highly correlated with self-perceptions in
the domain of physical appearance for children of this
age (Harter, 1985). The authors also found that
although physical appearance was not rated as significantly more important than other domains, global selfworth was most highly correlated with self-perceptions
in the physical appearance domain for all children in
the sample (r = .67).
The discounting hypothesis was based on the argument that chddren who discount effectively should have
lower importance scores in domains in which they do
not perceive themselves as competent, thus lowering the
discrepancy between competence and importance, and
JCAF" Volume 8, Number 1,January- March, 1995


enhancing self-esteem. However, the discounting
hypothesis failed to take into account the possibility that
obese children would view themselves unrealistically.
How did some obese children maintain high levels of
global self-worth in the apparent absence of discounting?
The distortion hypothesis addresses this question. If children are not competent in a domain they consider

important, they can maintain self-esteem by increasing
their perception of competence. Obese children, then,
may perceive themselves as less overweight than they
actually are.
Distortion of weight. It was predicted that global
self-worth, weight status, and importance of physical appearance would interact to influence the
degree of distortion of weight. The results of two
analyses supported this hypothesis most clearly.
Obese children distorted their perception of weight
more than overweight or average-weight peers,
regardless of level of self-worth, or importance of
physical appearance. In fact, average-weight children had negative distortion scores, indicating they
perceived themselves as less competent in weight
(heavier) than they actually were.
The main effect of weight status was moderated by a
significant interaction with importance of physical
appearance only for the high self-worth group. Because
of the small numbers of subjects in each cell, the interaction effect should be interpreted with caution; however,
the pattern is as expected (Figure 2). Obese children distorted more than overweight and average-weight children. High self-esteem enhanced distortion for obese and
overweight groups, with low GSW children viewing
themselves as less competent in weight than did their
high self-esteem peers.
Finally, distortion was a relatively effective defensive
cognition for obese children. To the extent that they utilized distortion, their levels of global self-worth
increased. Nevertheless, few obese children had selfworth scores high enough to place them in the upper
half of this sample. Distortion was used less by nonobese
children, and was not reliably associated with any
increase in self-worth.
JCAPNVolume 8, Number 1, January- March, 1995

Limitations. Several aspects of this study limit the
generalizability of the findings. First, the convenience
sample was drawn from a rural elementary school, and
included a disproportionate number of overweight children. The measure of weight status was based on the
national medians for age, sex, and height. Thus, most of
these children were heavier than their peers nationally. It
is not clear whether this study attracted subjects who
were concerned about their weight, or whether most
children in that rural school are also overweight. Since
the sample was small, and included only fifth-grade volunteers, generalization to other ages, settings, or nonvolunteers is difficult.

Global self-worth, weight status, and
importance of physical appearance interact to
4

influence the degree of distortion of weight.

Potential threats to internal validity include the correlational design, which limits knowledge about the direction of effects among weight, self-esteem, and defensive
cognitions. Instrumentation effects were possible, since
children knew that they were going to be weighed,
although some control was achieved by having children
complete the written instruments first. A more serious
threat was that of reactivity, or hypothesis guessing.
Since children knew the study was about height and
weight, they had an opportunity to alter their responses
to questions involving appearance and weight.
However, the authors would expect such results to be in
the opposite direction from the actual findings. If
responding reactively, overweight children should have
distorted their responses about weight less when they

knew they would be weighed. It is unrealistic to expect
that children would guess the hypotheses about discounting and distortion.
25


Maintenance of Self-Esteem by Obese Children

Implications
The results of this study reveal that obese children differ from their average weight peers in how they perceive
their physical appearance and weight, their social acceptance, and general self-esteem. They do not differ in how
important they consider these areas of performance. The
pattern of results indicate that averageweight children
utihze discounting, while obese children utilize distortion. Each process is associated with more positive selfperceptions. Although use of these processes enhances
global self-worth, obese children tend to have lower
global self-worth than their averageweight peers.

We need to rethink our interventions for
children with low self-esteem.

maintain any semblance of self- esteem. A further complication is that obesity is a health risk, requiring treatment (Obese Children, 1986). In emphasizing weight
reduction, nurses must also recognize that confronting
children with the reality of their weight may diminish
their self-esteem. A balanced approach, which includes
teaching about nutrition and exercise in a weight reduction program, as well as allowing children to utilize their
defenses, is warranted until further research demonstrates other means to assist obese children to maintain
positive self-esteem.
Acknowledgment. This research was partially supported by a
Brannen Award, and a traineeship from the National Institute
of Mental Health for the first author, and grants from the
National Institute of Mental Health (5 TO1 MH18281-03 and 1

KO7 MH00764-02) to the second author. Portions of this paper
were presented at the annual meetings of Advocates for Child
Psychiatric Nursing, Atlanta, September 1990, and the Society
for Research in Child Development, Seattle, April 1991.

References
These results are consistent with theories suggesting
that adults and children engage in cognitive defenses to
maintain their levels of self-esteem (Harter, 1988; Tesser,
1988).The results are also consistent with the large body
of evidence indicating mentally healthy individuals are
overly positive in their self-evaluations(Taylor & Brown,
1988).Taylor and Brown assert that these positive dlusions may be especially adaptive when an individual
receives negative feedback or is threatened by negative
information.
These findings suggest that we need to rethink our
interventions for children with low self-esteem. It seems
that children naturally use defensive cogmtions to bolster self-perceptions. Nurses may be able to capitalize on
defensive thoughts, by emphasizing the importance of
areas in which children perceive themselves as being
competent, and may assist children to discount domains
in which they are less competent. Unfortunately,children
have great difficulty in discounting physical appearance.
Thus, distortion becomes necessary for obese children to
26

Allon, N. (1979). Self-perceptions of the stigma of overweight in relationship to weight losing patterns. American Journal of Clinical
Nutrition, 32,47040.
Bandura, A. (1986). Social foundations of thought and action: A social copifive theory. E n g l e w d Cliffs, NJ:h n t i c e Hall.
Bell, C., Kirkpatrick, S., & Rinn, R. (1986). Body image of anorexic,

obese, and normal females. journal of Clinical Psychology, 43,
431-439.
Hamill, I?, Drizd, T., Johnson, C., Reed, R., & Moore, W. (1979).Physical
growth: National Center for Health statistics percentiles. American
Journal of Clinical Nutrition, 32,607-629.
Hammen, C., & Zupan, B. (1984). Self-schemas, depression, and the
processing of personal information in children. lournal of
Experimental Child Psychology, 37,598-608.
Hanison, A. (1983).Body image and self-esteem. In J. Mack & S. Ablon
(Eds.), The dalelopmmit and sustaining of self-esteem in childhood (pp.
90-102). New York International Universities Press.
Harter, S. (1985). Manual for the self-perception profile for childreri. Denver,
CO. University of Denver.
Harter, S. (1986).The determinants and multidimensionalrole of global
self-worth in children. In N. Eisenberg (Ed.), Contemporary issues in
dmlopmental psychology (pp. 219-242). New York: Wiley.

JCAPNVolume 8, Number 1, January- March, 1995


W,S. (1988). k d and dynamicchangg in the M ~ N Of
F the selfcontxpt Implicatiom for child pychohrap. In S. Shirk (Ed.),Cognitioe
deve-t
nnd h7d psychothempy ( p ~119-160).
.
N~Y
v ~ r kplenum

Straws, C., Smith, K., Frame, C., & Forehand, R. (1985). Personal and
interpersonal characterisitcs associated with childhood obesity.

Journal of Pediatric Psychology, 20,337-343.

Harter, S., Marold, D., & Whitesell, N. (1992). Model of psychosocial
risk factors leading to suicidal ideation in young adolescents.
Dmelopment and Psychopathology,4,167-188.

Taylor, S., & Brown, J. (1988).Illusion and well-being: A social psycho
logical perspechve on mental health. Journal of Personality and Social
Psy~hol~gy,
103,193-210.

Hendry, L., & Gillies, P. (1978).Body type, body esteem, school and
leisure: A study of overweight, average and underweight adolescents. Journal of Youth and Adolescence, 7,181-195.

Tesser, A. (1988).Toward a self-evaluation maintenance model of social
behavior. In L. Berkowitz (Ed.),Advances in experimental social psychology. Vol.22 (pp. 181-227). New York Academic Press.

Killeen, M. (1993). P m t influenceson childmfs self-esteem in economicall17 disadvantaged familig. issues in Mental Htdth N U ~ g24,323-336.
,

Wadden, T., Foster, C., Brownwell, K., & Finley, E. (1984).Self-cuncept
in obese and normal-weight children. Journal of Consulting and
Clinical Psychology, 52,1104-1105.

Kizziar..J., & Hagedorn, J. (1979).Search for acceptance: The adolescent and
self-esteem. Chicago: Nelson-Hall.

Weil, W. (1981).Obesity in children. Pediatric Review, 3,180-182.

Koniz-Booher, P., & Koniz, M. (1986). Nutrition and life-style. In C.

Edelman & C. Mandle (Eds.), Health promotion throughout the lifeS ~ T (pp.
I
113-141). St. Louis: Mosby.

Womly, A. (1981).In the eye of the beholder: Social and personal characteristics of teenagers and their impressions of themselves and fat
and slim people. British Journalof Medical Psychology, 54,231-242.

Leon, G., Bemis, K., Melland, M., & Nussbaum, D. (1978). Aspects of
body image perception in obese and normal-weight youngsters.
Journal of Abnonnal Child Psychology, 6,361-371.

Zupan, B., Hammen, C., & Jaenicke, C. (1987). The effects of current
mood and prior depressive history on self-schematic processing in
children.JournalofEXpen'menta1Child Psychology, 43,149-158.

Mack, 1. (1983).Self-esteem and its development: An overview. In J.
Mack & S. Ablon (Eds.), The development and sustaining of self-esteem
in childhood (pp. 142).New York International UniversitiesPress.
McCarthy, A. (1986). The school-age period. In C. Edelman & C.
Mandle (Eds.), Health promotion throughout the life-span (pp.
407.-450). St. Louis: Mosby.
Mendelson, B., & White, D. (1982). Relation between body-esteem and
self-esteem of obese and normal children. Perceptual and Motor
Skills, 54,899-905.
Mendelson,B., & White, D. (1985).Development of self-bodyeteem in
overweight youngsters. Developmental Psychology, 21, W-%.
Mendelson, B., & W t e , D. (1991,April). Dewlopment ofself-body+teem:
A longitudinal analysis. Paper presented at the Biennial meeting of
the Society for Research in Child Development,Seattle, WA.
Obese tlhildren: A growing problem. (1986).Science, 2 3 2 , 2 0 4 .

Rosenberg, M. (1979).Conceiving the self. New York Basic Books.

Rosenberg, M. (1986).self-concept from middle childhood through ad@
lescence. In J. Suls & A. Gmnwald (Eds.), Psychological perspectives
on the sey. Vo1.3 (pp. 107-136). Hillsdale, NJ:Lawrence Erlbaum.

Roy, C., & Andrews, H. (1991). The Roy adaptation model: The definitive
stat,ment. Norwalk, CT: Appleton & Lange.
Sallade; J. (1973). A comparison of psychological adjustment of obese
vs. non-obese children.Journalof Psychosomatic Research, 17,89-%.
Staffieri, J. (1967).A study of social stereotype of body image in children.Journal of Personality and Sorial Psychology, 7,101-104.

-

JCAPN Volume 8, Number 1,January March, 1995

Save the Date
The ACAPN '95 Conference
Moving Ahead in Child & Adolescent
Psychiatric Nursing:
Innovations & Retooling
September 21-23
Ramada Resort Hotel
Orlando, FL
Discount air fare available:
US Air, 800/ 334-8644;
ask for Gold File #38930298
For more information:
ACAPN, 800/826-2950

27