Roles of Sexual Objectification Experiences and Internalization of ...

Text-only Preview

Journal of Counseling Psychology
Copyright 2005 by the American Psychological Association
2005, Vol. 52, No. 3, 420 – 428
DOI: 10.1037/0022-0167.52.3.420
Roles of Sexual Objectification Experiences and Internalization
of Standards of Beauty in Eating Disorder Symptomatology:
A Test and Extension of Objectification Theory
Bonnie Moradi, Danielle Dirks, and Alicia V. Matteson
University of Florida
This study extends the literature on eating disorder symptomatology by testing, based on extant literature
on objectification theory (B. L. Fredrickson & T. Roberts, 1997) and the role of sociocultural standards
of beauty (e.g., L. J. Heinberg, J. K. Thompson, & S. Stormer, 1995), a model that examines (a) links
of reported sexual objectification experiences to eating disorder-related variables and (b) the mediating
roles of body surveillance, body shame, and internalization of sociocultural standards of beauty.
Consistent with hypotheses, with a sample of 221 young women, support was found for a model in which
(a) internalization of sociocultural standards of beauty mediated the links of sexual objectification
experiences to body surveillance, body shame, and eating disorder symptoms, (b) body surveillance was
an additional mediator of the link of reported sexual objectification experiences to body shame, and (c)
body shame mediated the links of internalization and body surveillance to disordered eating.
Keywords: objectification theory, eating disorders, sexism, body image, self-objectification
Research, theory, treatment, and prevention of eating disorder
disorder symptomatology. More specifically, objectification the-
symptomatology have been important foci for counseling psychol-
ory posits that women’s life experiences and gender socialization
ogists (Kashubeck-West & Mintz, 2001). Identifying contextual
routinely include experiences of sexual objectification that reduce
and intrapersonal variables linked to eating disorder symptoms is
women to their bodies, body parts, or body functions. Similarly,
crucial for understanding how to prevent and treat such symptoms.
Bartky (1988, 1990) defined sexual objectification as the reduction
Indeed, a perspective that attends to contextual and intrapersonal
of a woman’s body to its parts or functions, including the misper-
variables in understanding mental health is a defining feature of
ception that those parts or functions are capable of representing the
counseling psychology (American Psychological Association,
woman as a whole. Consistent with this definition, Fredrickson
1999). Using such a perspective to understand eating disorder
and Roberts (1997) argued that sexual objectification occurs
symptomatology among college age women is important given the
“whenever a woman’s body, body parts, or sexual functions are
high prevalence of such symptoms in this population. Studies
separated out from her person, reduced to status of mere instru-
focusing on undergraduate women suggest that as many as 64%
ments, or regarded as if they were capable of representing her” (p.
have engaged in disordered eating behaviors or attitudes (Mintz &
175). A frequently cited and subtle example of sexual objectifica-
Betz, 1988). More recently, Tylka and Subich’s (2002) data sug-
tion is the objectifying gaze that can occur in interpersonal and
gested that more than half of college women skipped meals (59%);
social encounters and media representations (Fredrickson & Rob-
approximately one third restricted calorie intake (37%), eliminated
erts, 1997; Goffman, 1979; Kilbourne & Jhally, 2000). Swim,
fats (30%), and eliminated carbohydrates (27%); and about one
Hyers, Cohen, and Ferguson’s (2001) series of diary studies with
fourth fasted for more than 24 hr (26%).
college women and men supported the routine occurrence of
Objectification theory (Fredrickson & Roberts, 1997) represents
sexual objectification of women. In these studies, sexual objecti-
a major advancement in eating disorder scholarship that integrates
fication of women (e.g., whistles or cat calls, sexual comments
extant theory and research and attends to contextual and intraper-
about body parts, inappropriate sexual comments or advances)
sonal variables that might play a role in the development of eating
emerged as a unique category of daily experiences of sexism that
participants reported having observed or experienced.
Objectification theory (Fredrickson & Roberts, 1997) posits that
Bonnie Moradi, Danielle Dirks, and Alicia V. Matteson, Department of
routine sexual objectification experiences socialize girls and
Psychology, University of Florida.
women to treat themselves as objects to be looked upon and
Alicia V. Matteson is a Major in the Air Force. The views expressed in
evaluated such that their bodies become objects for others (Bartky,
this article, however, are those of the authors and do not reflect the official
1988, 1990; de Beauvoir, 1952; McKinley, 1998; Spitzack, 1990).
policy or position of the United States Air Force, Department of Defense,
This internalization of an observer’s perspective upon one’s own
or the U.S. Government. We thank Jacob van den Berg and Michele
body is called self-objectification and is manifested by persistent
Goldstein for their invaluable assistance with data collection for this
body surveillance. As such, a woman’s relationship to her body
comes to parallel an observer’s relationship to an object; essen-
Correspondence concerning this article should be addressed to Bonnie
Moradi, Department of Psychology, University of Florida, P.O. Box
tially, women’s bodies become objects even to themselves (Mc-
112250, Gainesville, FL 32611-2250. E-mail: [email protected]
Kinley, 1998; Noll & Fredrickson, 1998). Within the objectifica-

tion theory framework, self-objectification in turn is theorized to
predicting body shame such that self-objectification was related
lead to greater levels of body shame and anxiety, reduce awareness
more strongly to body shame for women who tried on swimsuits
of internal bodily states, and prevent or disrupt peak motivational
than for women who tried on sweaters. Body shame in turn was
states or flow experiences (Csikszentmihalyi, 1982, 1990) for
related to restrained eating, measured by the amount of cookies
women. These experiences then contribute to depression, sexual
that participants ate. Similar results emerged in a second sample of
dysfunction, and eating disorder symptomatology, each of which is
college age women. More specifically, self-objectification and
more prevalent among women than among men (see Figure 1).
body shame each predicted greater restrained eating, and these
The variables highlighted in objectification theory are proposed
relationships emerged with body mass index (weight divided by
to originate from women’s gender role socialization and from
height squared) controlled as a covariate.
experiences of sexual objectification and share a role in shaping
In another study, Noll and Fredrickson (1998) found support for
women’s symptomatology. Nevertheless, there are some important
the mediating role of body shame posited by objectification theory.
conceptual distinctions among these variables. More specifically,
Across two samples of college age women, they found that body
self-objectification manifested by body surveillance is the act of
shame partially mediated the positive relationship between self-
consistently measuring oneself against some internalized or cul-
objectification and disordered eating. In other words, self-
tural standard, whereas body shame is the emotion that results
objectification was related to greater body shame, which in turn
from measuring oneself against such a standard and coming up
was related to greater eating disorder symptomatology. Beyond
short. Anxiety includes the anticipation of danger or threats to
this indirect relationship, there was also a direct positive relation-
one’s safety and fear about when and how one’s body will be
ship between self-objectification and disordered eating. These
looked at and evaluated. Peak motivational states are optimal
findings emerged with body mass index controlled as a covariate
experiences or “rare moments during which we feel we are truly
and regardless of whether symptoms of bulimia or anorexia were
living, uncontrolled by others, creative and joyful” (Fredrickson &
Roberts, 1997, p. 183). Finally, awareness of bodily states refers to
Tiggemann and Slater (2001) examined the replicability of this
the ability to detect and accurately interpret physiological sensa-
mediational model with a sample of classical ballet dancers and a
tions such as heartbeat, stomach contractions, and physiological
second sample of nondancers. Their results for both samples were
sexual arousal.
quite similar to those reported by Noll and Fredrickson (1998).
Within the larger framework of objectification theory, Noll and
More specifically, Tiggemann and Slater examined both general
Fredrickson (1998) identified self-objectification and body shame
self-objectification and the more specific manifestation of self-
as the key predictors of eating disorder symptomatology (see
objectification as body surveillance and found that body shame
Figure 1). Consistent with this conceptualization, much of the
mediated partially the positive link of body surveillance to eating
extant empirical research on objectification theory has focused on
disorder symptoms for ballet dancers and mediated fully this link
self-objectification and body shame and their links to eating dis-
for nondancers. These authors also examined but found no signif-
order symptoms. Overall, this research has yielded results that
icant unique links of appearance anxiety, flow experiences, or
support the propositions of objectification theory. For example,
awareness of internal bodily states to eating disorder symptoms
Fredrickson, Roberts, Noll, Quinn, and Twenge (1998) experimen-
beyond the links of body surveillance and body shame. Thus, these
tally manipulated the salience of self-objectification by having
variables did not emerge as predictors of eating disorder symptoms
college age women try on a swimsuit or a sweater in front of a
nor mediators of the self-objectification– eating disorder link, re-
full-length mirror. They found a significant interaction of experi-
inforcing the proposition that self-objectification and body shame
mental condition (swimsuit or sweater) by self-objectification in
are the key contributors to eating disorder symptoms within the
Figure 1.
Predictors of eating disorder symptomatology in the context of objectification theory. This model
represents objectification theory in its entirety and does not reflect the model examined in the present study.
Solid lines indicate key conceptual and empirically supported links for eating disorder symptomatology. Dashed
lines indicate other links proposed in the model.

objectification theory framework. Furthermore, Tiggeman and
Heinberg et al., 1995; McKinley & Hyde, 1996). These conceptual
Slater found that although participants’ scores on measures of
and empirical links are consistent with Baron and Kenny’s (1986)
general self-objectification and body surveillance overlapped sub-
definition that a variable functions as a mediator “to the extent that
stantially (r
.61), the specific manifestation of self-
it accounts for the relation between the predictor and criterion.
objectification as body surveillance and not general self-
Mediators explain how external physical events take on internal
objectification was linked uniquely to all other variables. In other
psychological significance” (p. 1176). Consistent with this defini-
words, general self-objectification did not account for unique
tion, internalization of cultural standards of beauty might be a
variance beyond that accounted for by body surveillance in any of
critical mechanism that translates sexual objectification experi-
the other eating disorder-related constructs. Thus, in addition to
ences (external events) into body surveillance, body shame, and
providing further support for Noll and Fredrickson’s mediational
eating disorder symptoms (internal psychological variables).
model, Tiggemann and Slater’s findings highlighted the impor-
Thus, on the basis of the literature reviewed here and the
tance of assessing specifically body surveillance as the manifes-
prevalence of eating disorder symptomatology among young
tation of self-objectification.
women (Mintz & Betz, 1988; Tylka & Subich, 2002), the present
Across these studies then, support has mounted for links among
study tested a model that examines the following hypotheses with
self-objectification, body shame, and eating disorder symptoms.
a sample of college age women:
The critical role of sexual objectification experiences as the pre-
cursor to these links, however, has received very limited attention
Hypothesis 1: Reported sexual objectification experiences are
in research on objectification theory. In fact, we identified only
related to greater levels of internalization of sociocultural
one published study that began to address this gap. Morry and
standards of beauty, body surveillance, body shame, and
Staska (2001) assessed college women’s exposure to beauty mag-
eating disorder symptoms. (Support for this hypothesis is a
azines as one specific type of sexual objectification experience.
precondition for Hypothesis 2.)
These authors found that women’s self-reported exposure to
beauty (but not fitness) magazines was related to greater levels of
Hypothesis 2: Links of reported sexual objectification expe-
(a) self-objectification and (b) eating disorder symptomatology. In
riences to body surveillance, body shame, and eating disorder
each case, however, the link was mediated fully by internalization
symptomatology are mediated by internalization of sociocul-
of sociocultural standards of beauty. In other words, greater expo-
tural standards of beauty. (Given limited research in this area,
sure to beauty magazines was related to greater internalization of
we explore both partial and full mediation.)
cultural beauty standards, and this internalization in turn was
Hypothesis 3: Consistent with prior findings, body shame
related to self-objectification and eating disorder symptoms. Con-
mediates partially the link of body surveillance to eating
sistent with other findings, these authors also found a positive
disorder symptomatology.
correlation between self-objectification and eating disorder symp-
toms. These findings provide preliminary support for the role of
The model testing these hypotheses is depicted in Figure 2.
sexual objectification experiences proposed in objectification the-
Consistent with prior research and to provide a more stringent test
ory. In addition, Morry and Staska’s findings indicate that inter-
of the hypotheses, we controlled body mass index as a covariate in
nalization of sociocultural standards of beauty, a variable not
tests of these hypotheses.
explicitly included in objectification theory, is important to con-
sider in research on objectification theory. Unfortunately, these
authors did not analyze concomitantly the links among sexual
objectification experiences, internalization of sociocultural stan-
dards of beauty, self-objectification, and disordered eating. Also,
they did not include the role of body shame in the links they
Participants were 221 undergraduate women at a large southeastern U.S.
university and ranged in age from 17 to 45 years (M
20.42, Mdn
2.75). Sixty-four percent of the sample identified as White, 11%
Thus, an important next step in extending the literature on
Latina or Hispanic, 8% African American or Black, 8% Asian American or
objectification theory as applied to understanding eating disorder
Pacific Islander, 1% Native American, and 8% as multiracial or other.
symptoms is to empirically examine a model that includes rela-
Overall, 43% of the participants were in their third year of college, 28% in
their fourth year, 19% in their second year, 9% in their first year, and 1%
objectification (assessed specifically as body surveillance), body
in graduate school. Thirty-seven percent of the women reported that they
shame, and eating disorder symptoms. Furthermore, Morry and
were married or in a committed relationship, and 63% were single. In terms
Staska’s (2001) findings and other theoretical and empirical liter-
of family social class, 43% of the sample identified as upper middle class,
ature on the relation of internalization of sociocultural standards of
41% middle class, 11% working class, 3% upper class, and 1% lower class.
beauty suggest that internalization might mediate links of sexual
Ninety-one percent of the sample identified as exclusively heterosexual,
objectification experiences to body surveillance, body shame, and
5% mostly heterosexual, 3% bisexual, and less than 1% exclusively ho-
mosexual. (Some of the descriptive percentages may not add to 100%
eating disorder symptoms. For example, body shame is posited to
because of rounding and a few missing responses.)
result from the internalization of unachievable idealized standards
of beauty (Bartky, 1988, 1990; McKinley & Hyde, 1996). In
addition, internalization of sociocultural standards of beauty has
been shown to be related to body surveillance, body shame, and
Undergraduate women from a variety of courses were invited to partic-
eating disorder symptomatology (e.g., Cashel, Cunningham, Lan-
ipate in a survey study on women’s life experiences and well-being.
deros, Cokley, & Muhammad, 2003; Griffiths et al., 1999, 2000;
Persons willing to participate attended scheduled sessions of up to 5

Figure 2.
Model for testing stated hypotheses.
persons per session and received extra credit toward their course grade in
and body image preoccupation (Morry & Staska, 2001). Alpha internal
the classes from which they were recruited. Procedures were described to
consistency reliability estimate with the present sample was .88.
participants, and written consent was obtained. Participants then completed
Self-objectification manifested as body surveillance.
The Body Sur-
a survey packet that included the following instruments and a demographic
veillance (McKinley & Hyde, 1996) subscale of the Objectified Body
questionnaire. The order of instruments in the survey packets was coun-
Consciousness Scale is an eight-item instrument that measures how much
terbalanced. A multivariate analysis of variance with order as the indepen-
a woman thinks of her body in terms of how it looks rather than how it feels
dent variable and sexual objectification experience, internalization, body
(i.e., self-objectification). Questions include “I rarely worry about how I
surveillance, body shame, and eating disorder symptomatology scores as
look to other people” and “I think more about how my body feels than how
dependent variables revealed no significant order effects.
my body looks.” Participants respond to items on a 7-point Likert-type
scale (1
strongly disagree to 7
strongly agree) and indicate NA (not
applicable) if the item does not apply to them. Consistent with McKinley
and Hyde’s (1996) recommendation, the few “not applicable” responses for
Reported sexual objectification experiences.
The sexual objectification
the subscale were coded as missing. Appropriate items are reverse coded,
subscale of Swim, Cohen, and Hyers’s (1998) 25-item measure of daily
and nonmissing item ratings are averaged to yield a scale score, with higher
sexist events was used to assess participants’ reported sexual objectifica-
scores indicating greater levels of self-objectification. With regard to
tion experiences. This subscale consists of seven self-report Likert-type
validity, consistent with objectification theory, women scored higher than
never to 5
about two or more times a week during the last semester)
men on Body Surveillance (McKinley, 1998). Furthermore, Body Surveil-
items that assess the frequency of reported sexual objectification experi-
lance scores were correlated as expected but were not redundant with other
ences. Items for the subscale were based on events observed or experienced
relevant constructs. More specifically, McKinley and Hyde’s exploratory
by women and men in a diary study of everyday sexist events. Sample
and confirmatory factor analyses indicated that body surveillance emerged
items include “Had people shout sexist comments, whistle, or make cat-
as a factor that was distinct from body shame and control beliefs (i.e., belief
calls at me” and “Had sexist comments made about parts of my body or
that one can control one’s appearance). Furthermore, McKinley and Hyde
clothing.” Item ratings are averaged to yield a subscale score, with higher
found that a three-factor solution with body surveillance, body shame, and
scores indicating more frequent reported sexual objectification experi-
control beliefs as separate factors fit their data significantly better than a
ences. In terms of validity, Swim et al. (2001) found that women reported
two-factor structure that modeled body surveillance and body shame as a
more sexual objectification experiences than did men, and these and other
single factor. Nevertheless, as expected, Body Surveillance scores were
sexist events were related more strongly to anxiety for women than for men
correlated negatively with body esteem and were correlated positively with
but were independent of levels of neuroticism. The alpha internal consis-
body shame and control beliefs (McKinley, 1998). Alpha internal consis-
tency reliability for the sexual objectification subscale with the present
sample was .87.
tency reliability estimates have ranged from .76 to .89 with undergraduate
Internalization of sociocultural standards of beauty.
The Internaliza-
and middle-aged women (McKinley, 1999; McKinley & Hyde, 1996).
tion scale of Heinberg et al.’s (1995) Sociocultural Attitudes Toward
McKinley and Hyde reported a 2-week test–retest reliability of .79 for
Appearance Questionnaire is an eight-item Likert-type (1
Body Surveillance scores. Alpha with the present sample was .82.
disagree to 5
completely agree) measure that assesses how much an
Body shame.
The Body Shame subscale is an eight-item subscale of
individual accepts and internalizes societal standards of beauty (e.g., “I
McKinley and Hyde’s (1996) 24-item Objectified Body Consciousness
wish I looked like a swimsuit model” and “Music videos that show thin
Scale that measures how much a woman feels like a “bad person” when she
women make me wish that I were thin”). Item ratings are averaged to
believes that her body does not achieve cultural body standards. Questions
obtain subscale scores, with higher scores indicating greater levels of
for the subscale include “When I cannot control my weight, I feel like there
internalization of sociocultural standards of beauty. Adequate reliability for
must be something wrong with me” and “When I’m not the size I think I
Internalization scores has been demonstrated across a variety of samples. In
should be, I feel ashamed.” Participants respond to items on a 7-point
the development of the scale, Heinberg et al. reported an alpha of .88 with
Likert-type scale (1
strongly disagree to 7
strongly agree) and circle
undergraduate women. More recently, Morry and Staska (2001) obtained
NA (not applicable) if the item does not apply to them. Consistent with
an alpha of .85 with women in their study. In terms of validity, Internal-
McKinley and Hyde’s recommendation, the few “not applicable” responses
ization scores have been shown to be largely independent of the awareness
for the subscale were coded as missing. Appropriate items are reverse
of sociocultural standards of beauty (Heinberg et al., 1995) but related
coded, and nonmissing item ratings are averaged to yield a scale score,
positively to body dissatisfaction (Griffiths et al., 2000), abnormal eating
with higher scores indicating greater levels of body shame. As indicated
attitudes (Griffiths et al., 1999), restrained eating (Griffiths et al., 2000),
previously, McKinley and Hyde found that Body Shame scores were

correlated as expected but were not redundant with other relevant con-
2.05, SD
0.73), Griffiths et al. (2000) for internalization (M
structs. More specifically, body shame emerged as a factor that is distinct
3.12, SD
0.83), McKinley and Hyde (1996) for body surveil-
from body surveillance and control beliefs, and as expected, Body Shame
lance (M
4.22, SD
0.91) and body shame (M
3.24, SD
scores were correlated positively with body surveillance and negatively
1.04), and Mazzeo (1999) for eating disorder symptoms (M
with body esteem (McKinley, 1998). Alpha internal consistency reliability
2.49, SD
estimates for Body Shame scores ranged from .70 to .84 across samples of
In the present sample, none of the variables of interest were
undergraduate and middle-aged women (McKinley, 1999; McKinley &
related to age, relationship status, year in school, family social
Hyde, 1996). McKinley and Hyde reported a 2-week test–retest reliability
of .84 for Body Shame scores. Alpha with the present sample was .81.
class, or sexual orientation when body mass index was controlled
Eating disorder symptomatology.
Garner, Olmstead, Bohr, and Gar-
(alpha adjusted to .05/25
.002). However, a multivariate anal-
finkel’s (1982) Eating Attitudes Test (EAT-26) was used to measure the
ysis of covariance, with body mass index entered as a covariate,
broad range of disordered eating behaviors and attitudes among partici-
suggested that White participants (n
142) scored significantly
pants. The EAT-26 has 26 items that are scored on a 6-point Likert-type
differently from non-White participants (n
78) on the set of
scale (1
always to 6
never). Questions assess disordered eating
variables of interest, F(5, 213)
3.40, p
attitudes such as “Feel that food controls my life” and “Feel extremely
(Because of very small sample sizes for some racial or ethnic
guilty after eating” and disturbed eating behaviors such as “Avoid eating
groups, non-White participants were combined for this analysis.)
when I am hungry” and “Vomit after I have eaten.” Following Kashubeck-
Follow-up univariate analyses of variance indicated that non-
West, Mintz, and Saunders’s (2001) recommendation, continuous scores
White participants had lower internalization, F(1, 217)
were used to reflect the continuum of eating problems. Item ratings were
averaged to yield a scale score. For ease of interpretation, we reverse
.06; surveillance, F(1, 217)
7.47, p
scored items so that higher scores indicated more maladaptive eating
.03; and eating disorder symptomatology scores, F(1,
behaviors and attitudes (Mintz & O’Halloran, 2000). The EAT-26 is one of
11.68, p
.05; than did White participants.
the most widely used measures of disordered eating (Garner, 1997). In their
Effect sizes for these significant differences suggested that White
review of eating disorder measures, Kashubeck-West et al. reported alphas
versus non-White group status accounted for 3% to 6% of the
for the EAT scores ranging from .79 to .94 across samples. They also
variability in these scores.
reported that EAT scores were related to other measures of eating disorder
symptomatology as expected and differentiated between clinical and non-
clinical groups. These authors recommended use of the EAT as a contin-
Primary Analyses
uous measure of disordered eating in research. Alpha in the present sample
was .92.
Partial correlations (holding body mass index as a covariate)
Body mass index.
Participants reported their height and weight, and
among the variables of interest are reported in Table 1. Partial
these self-reports were used to compute body mass index, which was
correlations among all of the variables of interest were significant
controlled as a covariate in the analyses.
and in the expected directions. Consistent with Hypothesis 1, with
body mass index controlled, reported sexual objectification expe-
riences were correlated positively with internalization of sociocul-
tural standards of beauty, body surveillance, body shame, and
Descriptive Information for the Present Sample
eating disorder symptoms.
Levels of sexual objectification experiences, internalization,
To test the mediations proposed in Hypotheses 2 and 3, we
body surveillance, body shame, and eating disorder symptoms for
followed Baron and Kenny’s (1986) procedures. These authors
our sample were generally close to the midrange of possible scores
indicated that for a variable to be tested as a mediator, there must
on each instrument (see Table 1), and these scores were compa-
be a significant relationship between the predictor and the medi-
rable to those in studies that used the same instruments with
ator and between the mediator and the criterion variable. Both of
samples of undergraduate women. More specifically, the present
these conditions were satisfied for our proposed mediators (i.e.,
sample’s means and standard deviations for sexual objectification
internalization, body shame). More specifically, with regard to
experiences (M
2.30, SD
0.80), internalization (M
Hypothesis 2, partial correlations presented in Table 1 indicated
0.91), body surveillance (M
4.81, SD
1.03), body
that reported sexual objectification experiences (i.e., the predictor)
shame (M
3.36, SD
1.12), and eating disorder symptomatol-
were correlated significantly with internalization (i.e., the potential
ogy (M
2.45, SD
0.76) were comparable to those reported by
mediator), which in turn was correlated significantly with body
Swim et al. (1998) for sexual objectification experiences (M
surveillance, body shame, and eating disorder symptoms (i.e., the
Table 1
Summary Statistics and Partial Intercorrelations Among Variables of Interest With Body Mass Index Controlled
Possible range
Sample range
1. Sexual objectification experiences

2. Internalization of beauty standards

3. Body surveillance

4. Body shame

5. Eating disorder symptoms

Higher scores reflect higher levels of the construct assessed.
* p

criterion variables). With regard to Hypothesis 3, body surveil-
objectification experiences also had a significant direct link to
lance (i.e., the predictor) was correlated significantly with body
body surveillance but not to body shame or eating disorder symp-
shame (i.e., the mediator), which in turn was correlated signifi-
toms. Thus, consistent with Hypothesis 2, internalization of socio-
cantly with disordered eating (i.e., the criterion variable).
cultural standards of beauty partially mediated the link of reported
When these conditions are satisfied, a variable is a mediator to
sexual objectification experiences to body surveillance and fully
the extent that it accounts for the relationship between the predic-
mediated the link of reported sexual objectification experiences to
tor and the criterion. To test this, we used Amos 4.01 (Arbuckle,
body shame and eating disorder symptoms. Body surveillance had
1999) to conduct a path analysis of a model in which all possible
a significant indirect link of .13 (.36
.37; z
4.22, p
direct and indirect paths were estimated (i.e., the model presented
through body shame to eating disorder symptoms and a significant
in Figure 2). Again, we entered participants’ body mass index as a
direct link of .14 to such symptoms. Thus, consistent with Hy-
covariate in the model. We used maximum likelihood estimation
pothesis 3, body shame partially mediated the link of body sur-
with the covariance matrix of the variables of interest as input.
veillance to eating disorder symptoms.
Values for the goodness-of-fit index (GFI), adjusted goodness-of-
In addition to these tests of our hypotheses, the significant direct
fit index (AGFI), comparative fit index (CFI), normed fit index
link of reported sexual objectification experiences to body surveil-
(NFI), and nonnormed fit index (NNFI, also known as the Tucker–
lance and the significant direct links of body surveillance to body
Lewis index) all were 1.0 (given that the model tested was fully
shame and eating disorder symptoms allowed us to explore body
saturated), and the model accounted for 50% of the variance in
surveillance as an additional mediator of the links of sexual ob-
eating disorder symptomatology, 35% of variance in body shame,
jectification experiences to body shame and eating disorder symp-
34% of variance in body surveillance, and 6% of the variance in
toms. Again, we multiplied standardized path coefficients and used
internalization of cultural standards of beauty. As indicated in
Sobel’s formula, and we found that through body surveillance,
Figure 3, standardized path coefficients all were significant and in
reported sexual objectification experiences had a significant indi-
the expected direction with the exception of the nonsignificant
rect link of .05 (.14
.36; z
2.45, p
.05) to body shame but
direct links of sexual objectification experiences to body shame
no significant indirect link to eating disorder symptoms. Thus, in
and eating disorder symptomatology. More specifically, reported
addition to the mediating role of internalization, body surveillance
sexual objectification experiences were related directly to body
simultaneously mediated the link of reported sexual objectification
surveillance and internalization of sociocultural standards of
experiences to body shame but did not mediate that link to eating
beauty; internalization was related directly to body surveillance,
disorder symptoms.
body shame, and eating disorder symptomatology; body surveil-
Similarly, the significant direct link of internalization to body
lance was related directly to body shame and eating disorder
shame and the significant direct link of body shame to eating
symptomatology; and body shame was related directly to eating
disorder symptomatology allowed us to explore body shame as an
disorder symptomatology.
additional mediator of the internalization– eating disorder symp-
The model also suggested a number of indirect links. Significant
toms link. We found that, in addition to the direct link between
indirect effects through the proposed mediators (i.e., internaliza-
internalization and eating disorders symptoms, there was a signif-
tion and body shame) would suggest significant mediator effects.
icant indirect link of .09 (.24
.37; z
3.14, p
.01), through
We multiplied indirect standardized path coefficients to compute
body shame, between these variables. Thus, body shame partially
indirect effects (Cohen & Cohen, 1983) and used Sobel’s formula
mediated the link of internalization to eating disorder symptoms.
(see Baron & Kenny, 1986) to determine whether indirect effects
Next, we compared the fit of the fully saturated model to that of
were significantly different from zero. Through internalization of
an alternative trimmed model that eliminated the nonsignificant
cultural standards of beauty, reported sexual objectification expe-
direct paths from reported sexual objectification experiences to
riences had significant indirect links of .13 (.25
.50; z
body shame and eating disorder symptoms. The change in the
.001), .06 (.25
.24; z
2.61, p
.01), and .09 (.25
chi-square statistic was not statistically significant, and fit index
3.04, p
.01) to body surveillance, body shame, and eating
values for this model were all above acceptable cutoffs (GFI
disorder symptoms, respectively. As mentioned previously, sexual
1.0, AGFI
.99, CFI
1.0, NFI
1.0, NNFI
1.0) and similar
Figure 3.
Trimmed model for examining links among variables of interest. Values reflect standardized
coefficients. All paths depicted are significant at p

to those for the original model. The amount of variance accounted
proposed direct and indirect roles of sexual objectification expe-
for in each of the criterion variables and the magnitude of the
riences in eating disorder symptomatology and their correlates.
significant paths in the trimmed model were identical to those in
The direct and indirect roles of sexual objectification experiences,
the original model (see Figure 3). Thus, the trimmed model was
in turn, suggest that an important area for prevention of eating
more parsimonious but equally appropriate as the fully saturated
disorder symptoms and their precursors is the continuation of work
targeted toward reducing the prevalence of sexual objectification
of women in the media and in private and public interpersonal
contexts. Furthermore, assessing and attending to women’s expe-
riences of sexual objectification in counseling/therapy is critical
Objectification theory (Fredrickson & Roberts, 1997) and other
given the evidence that such experiences might set the stage for
theoretical and empirical literature on eating disorder symptom-
internalization of sociocultural standards of beauty, body shame,
atology (e.g., Cashel et al., 2003; Griffiths et al., 1999, 2000;
body surveillance, and eating disorder symptomatology.
Heinberg et al., 1995; McKinley & Hyde, 1996) have identified
In addition, the mediating roles of internalization and body
sexual objectification experiences and internalization of sociocul-
surveillance in the links of sexual objectification experiences to
tural standards of beauty as important correlates of eating disorder-
eating disorder-related variables suggest that attending to the
related variables. To date, however, no study has examined con-
meaning that women make of their experiences of sexual objecti-
comitantly the roles of these variables in tests of objectification
fication is important in clinical work. Particularly, consistent with
theory. The present study addressed this important gap and ex-
objectification theory, our data suggest that women might be at
tended the literature on objectification theory as applied to under-
risk for experiencing greater body shame and eating disorder
standing eating disorder symptomatology (Fredrickson & Roberts,
symptomatology to the extent that they internalize sexual objecti-
1997) by (a) providing the first examination of the role of reported
fication experiences and translate these experiences into personal
sexual objectification experiences in a model of eating disorder
endorsement of sociocultural standards of beauty and body sur-
symptomatology based on objectification theory and (b) concom-
veillance. Clearly, a critical area for further research is identifying
itantly examining the mediating roles of body surveillance (man-
contextual and intrapersonal variables that might prevent or reduce
ifestation of self-objectification), body shame, and internalization
the occurrence of this process. Such research can inform counsel-
of sociocultural standards of beauty in that model.
ors’/therapists’ use of interventions designed to disrupt the trans-
Although reported experiences of sexual objectification have
lation of sexual objectification experiences into internalization of
received limited attention in research on eating disorder symptom-
sociocultural standards of beauty and self-objectification.
atology, the present findings highlight the importance of including
The general sexist events literature might serve as one guide for
such experiences when examining contextual and intrapersonal
identifying such buffers. For example, Moradi and Subich (2004)
variables that might be related to disordered eating. Correlations of
found that self-esteem moderated the link of reported experiences
reported sexual objectification experiences with internalization of
of sexist events to psychological distress, such that the link was
sociocultural standards of beauty, body surveillance, body shame,
positive for women with low self-esteem but nonsignificant for
and eating disorder symptoms all were significant and positive.
women with high self-esteem. Similarly, Moradi and Subich
Furthermore, direct and indirect relations of reported sexual ob-
(2003) found that recent reported sexist events were related more
jectification experiences with other variables in the model are
strongly to psychological distress for women with high passive
consistent with and build upon prior literature.
acceptance feminist identity development attitudes (denial of sex-
For example, the direct links of reported sexual objectification
ism and unexamined acceptance of traditional gender roles) than
experiences to internalization and body surveillance in the present
for women with low levels of such attitudes. Thus, self-esteem,
study are consistent with objectification theory’s proposition that
gender-related identity and attitudes, and other variables identified
women’s experiences of sexual objectification are an important
as buffers of stressful events (e.g., social support) might also serve
correlate of self-objectification manifested as body surveillance
as moderators in the links of sexual objectification experiences to
(Fredrickson & Roberts, 1997) and with Morry and Staska’s
internalization of sociocultural standards of beauty and body sur-
(2001) finding of a link between sexual objectification experiences
veillance. Research on potential moderators can inform interven-
and internalization of sociocultural standards of beauty. Further-
tions that prevent women from internalizing sexual objectification
more, the results of our mediational analyses highlight a number of
experiences, and such interventions ultimately might reduce wom-
potential mechanisms for the translation of sexual objectification
en’s risk for developing eating disorder symptoms.
experiences to eating disorder symptoms and their correlates. More
Our results also replicate prior findings that support the role of
specifically, we found that body surveillance and internalization of
body shame as a mediator in the link of body surveillance to eating
sociocultural standards of beauty simultaneously mediated the link
disorder symptoms (e.g., Noll & Fredrickson, 1998; Tiggemann &
of reported sexual objectification experiences to body shame. In
Slater, 2001). Our results extend prior work on objectification
addition, internalization of sociocultural standards of beauty me-
theory as applied to eating disorder symptomatology by demon-
diated the links of sexual objectification experiences to body
strating that body shame also mediated the link of internalization
surveillance and eating disorder symptoms. These findings point to
of cultural standards of beauty to eating disorder symptoms. Thus,
internalization and body shame as potential key mechanisms in
reducing level of body shame might be another important area for
translating sexual objectification experiences into eating disorder
prevention and intervention in counseling/therapy with women
symptoms and their precursors outlined in objectification theory.
experiencing or at risk for developing eating disorder symptoms.
Thus, our findings extend the literature on objectification the-
Relatedly, a promising area for research is the exploration of
ory, as applied to eating disorders, by providing support for the
contextual and intrapersonal variables that might buffer the trans-

lation of internalization and self-objectification into body shame
restricted to such women. There is a need to empirically test the
and disordered eating.
model examined in the present study and the propositions of
Another striking finding of our study is that the variables in-
objectification theory with different populations such as noncolle-
cluded in our model accounted for 50% of the variance in eating
giate women and women who represent various geographic re-
disorder symptomatology, a very large effect in correlational re-
gions, racial or ethnic backgrounds, and sexual orientations.
search (Cohen, 1988; Wampold & Freund, 1987). Thus, our find-
Our results indicated small but significant differences between
ings support a more comprehensive framework for research on
White and non-White women on internalization, body surveil-
objectification theory that extends prior work and includes exam-
lance, and eating disorder symptomatology, with body mass index
ination of reported sexual objectification experiences and internal-
controlled. These findings are consistent with prior findings that
ization of sociocultural standards of beauty. Furthermore, the
some eating disorder-related attitudes and behaviors might have
magnitude of the variance that the variables in the model ac-
lower prevalence rates among women of color than among White
counted for in eating disorder symptomatology highlights the
women (Gilbert, 2003). Nevertheless, the differences found in the
importance of attending to these variables in future research,
present study were very small (effects sizes of .03 to .06), and the
prevention, and treatment of eating disorder symptomatology with
paucity of research on eating disorder-related attitudes and behav-
young women. This point must be tempered, however, by the
iors among women of color suggests that much more scholarship
possibility that present results reflect measurement overlap in
is needed in this area before clear conclusions can be made
addition to meaningful conceptual overlap among variables of
(Striegel-Moore & Cachelin, 2001). Similarly, theoretical and em-
pirical work is needed to advance understanding of eating
Although the findings of the present study contribute to advanc-
disorder-related symptoms among men of various ages, sexual
ing research and practice related to eating disorders, several lim-
orientations, racial or ethnic and other backgrounds. Research with
itations must be considered when evaluating and interpreting the
more diverse populations would provide needed information about
present findings. First, our correlational data are consistent with,
the generalizability of objectification theory and the present find-
but do not directly evaluate, the directions of causality proposed in
ings. Integrating population-specific risk factors (e.g., experiences
theoretical conceptualizations on which the model we tested was
of racism and heterosexism or homophobia) and protective factors
based. More specifically, objectification theory (Fredrickson &
(e.g., positive cultural identity and connection with lesbian or gay,
Roberts, 1997; Noll & Fredrickson, 1998) suggests that experi-
Latina or Latino, African American, and other minority cultural
ences of sexual objectification foster self-objectification, which in
values) in such research is also important for advancing the liter-
turn leads to body shame and then to eating disorder symptoms.
ature on objectification theory and eating disorders.
Furthermore, extant theoretical conceptualizations suggest that
The present study provides one more step in the accumulating
internalization of sociocultural standards of beauty leads to eating
body of research that has tested aspects of objectification theory
disorder-related symptoms (e.g., Bartky, 1988, 1990; McKinley &
related to eating disorder symptomatology and contributed to the
Hyde, 1996; Morry & Staska, 2001). Experimental and longitudi-
broader literature on eating disorders. Our study extends prior
nal studies are needed to extend our findings and test directly the
research by testing a more comprehensive framework that included
causal and directional relations implicit in these conceptualiza-
the roles of reported sexual objectification experiences and inter-
tions. Identifying causal or directional links would facilitate fo-
nalization of sociocultural standards of beauty in objectification
cusing limited time and resources on key mechanisms and so prove
theory. Additional work to replicate and extend our findings to
invaluable for designing therapy/counseling interventions and pre-
broader populations is needed. Furthermore, research that begins
vention programs.
to explore potential buffers in the links identified in our study is
The reliance on self-report measures to assess eating disorder-
needed to advance research and practice with women who are
related attitudes and behaviors is also a limitation. Such reports
experiencing or at risk for developing eating disorder symptoms.
might be influenced by factors such as social desirability and
memory recall. Self-reports of sexual objectification experiences
might be particularly affected by recall and differential perceptions
of events. For example, two of the seven items that assessed sexual
American Psychological Association. (1999). Archival description of
objectification experiences asked about “sexist comments.” Re-
counseling psychology. The Counseling Psychologist, 27, 589 –592.
spondents’ perceptions of comments as sexist are likely to vary on
Arbuckle, J. L. (1999). Amos (Version 4.01) [Computer software]. Chi-
the basis of a number of intrapersonal and contextual variables
cago: SmallWaters.
such as perpetrators’ gender and race, intensity and impact of the
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable
event, and the target person’s knowledge about prejudice (Barrett
distinction in social psychological research: Conceptual, strategic, and
& Swim, 1998). These complexities raise an interesting question
statistical considerations. Journal of Personality and Social Psychology,
about whether sexual objectification experiences that are not per-
51, 1173–1182.
ceived or labeled as such will be related to self-objectification,
Barrett, L. F., & Swim, J. K. (1998). Appraisals of prejudice and discrim-
internalization, and their correlates. Thus, exploring the potential
ination. In J. K. Swim & C. Stangor (Eds.), Prejudice: The target’s
(pp. 11–35). San Diego, CA: Academic Press.
differential consequences of perceiving or not perceiving an event
Bartky, S. L. (1988). Foucault, femininity, and the modernization of
as sexual objectification is an important area for future research.
patriarchal power. In I. Diamond & L. Quinby (Eds.), Feminism and
An additional limitation is that our study, and much of the extant
Foucault: Reflections of resistance (pp. 61– 86). Boston: Northeastern
research on objectification theory, has focused primarily on
University Press.
women who are young, White, heterosexual, and middle to upper
Bartky, S. L. (1990). Femininity and domination: Studies in the phenom-
class college students. Thus, the generalizability of our findings is
enology of oppression. New York: Routledge.

Cashel, M. L., Cunningham, D., Landeros, C., Cokley, K. O., & Muham-
esteem: The mediating effect of objectified body consciousness and
mad, G. (2003). Sociocultural attitudes and symptoms of bulimia: Eval-
actual/ideal weight discrepancy. Sex Roles, 39, 113–123.
uating the SATAQ with diverse college groups. Journal of Counseling
McKinley, N. M. (1999). Women and objectified body consciousness:
Psychology, 50, 287–296.
Mothers’ and daughters’ body experience in cultural, developmental,
Cohen, J. (1988). Statistical power analysis for the behavioral sciences
and familial context. Developmental Psychology, 35, 760 –769.
(2nd ed.). Hillsdale, NJ: Erlbaum.
McKinley, N. M., & Hyde, J. S. (1996). The Objectified Body Conscious-
Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation
ness Scale: Development and validation. Psychology of Women Quar-
analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.
terly, 20, 181–215.
Csikszentmihalyi, M. (1982). Toward a psychology of optimal experience.
Mintz, L. B., & Betz, N. E. (1988). Prevalence and correlates of eating
In L. Wheeler (Ed.), Review of personality and social psychology (pp.
disordered behaviors among undergraduate women. Journal of Coun-
13–36). Beverly Hills, CA: Sage.
seling Psychology, 35, 463– 471.
Csikszentmihalyi, M. (1990). Flow. New York: Harper Perennial.
Mintz, L. B., & O’Halloran, M. S. (2000). The Eating Attitudes Test:
de Beauvoir, S. (1952). The second sex. New York: Vintage Books.
Validation with DSM–IV eating disorder criteria. Journal of Personality
Fredrickson, B. L., & Roberts, T. (1997). Objectification theory: Toward
Assessment, 74, 489 –503.
understanding women’s lived experiences and mental health risks. Psy-
Moradi, B., & Subich, L. M. (2003). A concomitant examination of the
chology of Women Quarterly, 21, 173–206.
relations of perceived racist and sexist events to psychological distress
Fredrickson, B. L., Roberts, T., Noll, S. M., Quinn, D. M., & Twenge, J. M.
for African American women. The Counseling Psychologist, 31, 451–
(1998). That swimsuit becomes you: Sex differences in self-
objectification, restrained eating, and math performance. Journal of
Moradi, B., & Subich, L. M. (2004). Examining the moderating role of
Personality and Social Psychology, 75, 269 –284.
self-esteem in the link between experiences of perceived sexist events
Garner, D. M. (1997). Psychoeducational principles in treatment. In D. M.
and psychological distress. Journal of Counseling Psychology, 51, 50 –
Garner & P. E. Garfinkel (Eds.), Handbook of treatment for eating
disorders (pp. 175–177). New York: Guilford Press.
Morry, M. M., & Staska, S. L. (2001). Magazine exposure: Internalization,
Garner, D. M., Olmstead, M. P., Bohr, Y., & Garfinkel, P. E. (1982). The
self-objectification, eating attitudes, and body satisfaction in male and
Eating Attitudes Test: Psychometric features and clinical correlates.
female university students. Canadian Journal of Behavioural Sciences,
Psychological Medicine, 12, 871– 878.
4, 269 –279.
Gilbert, S. C. (2003). Eating disorders in women of color. Clinical Psy-
Noll, S., & Fredrickson, B. L. (1998). Mediational model linking self-
chology: Science and Practice, 10, 444 – 455.
objectification, body shame, and disordered eating. Psychology of
Goffman, E. (1979). Gender advertisements. Cambridge, MA: Harvard
Women Quarterly, 22, 623– 636.
University Press.
Spitzack, C. (1990). Confessing excess: Women and the politics of body
Griffiths, J. A., Beumont, P. J. V., Russell, J., Schotte, D., Thornton, C.,
reduction. Albany, NY: SUNY Press.
Touyz, S., & Varano, P. (1999). Sociocultural attitudes towards appear-
Striegel-Moore, R. H., & Cachelin, F. M. (2001). Etiology of eating
ance in dieting disordered and nondieting disordered subjects. European
disorders in women. The Counseling Psychologist, 29, 635– 661.
Eating Disorders Review, 7, 193–203.
Swim, J. K., Cohen, L. L., & Hyers, L. L. (1998). Experiencing everyday
Griffiths, J. A., Mallia-Blanco, R., Boesenberg, E., Ellis, C., Fischer, K.,
prejudice and discrimination. In J. K. Swim & C. Stangor (Eds.),
Taylor, M., & Wyndham, J. (2000). Restrained eating and sociocultural
Prejudice: The target’s perspective (pp. 37– 60). San Diego, CA: Aca-
attitudes to appearance and general dissatisfaction. European Eating
demic Press.
Disorders Review, 8, 394 – 402.
Swim, J. K., Hyers, L. L., Cohen, L. L., & Ferguson, M. J. (2001).
Heinberg, L. J., Thompson, J. K., & Stormer, S. (1995). Development and
Everyday sexism: Evidence for its incidence, nature, and psychological
validation of the Sociocultural Attitudes Towards Appearance Question-
impact from three daily diary studies. Journal of Social Issues, 57,
naire. International Journal of Eating Disorders, 17, 81– 89.
Kashubeck-West, S., & Mintz, L. B. (2001). Eating disorders in women:
Tiggemann, M., & Slater, A. (2001). A test of objectification theory in
Etiology, assessment, and treatment. The Counseling Psychologist, 29,
former dancers and non-dancers. Psychology of Women Quarterly, 25,
627– 634.
57– 64.
Kashubeck-West, S., Mintz, L. B., & Saunders, K. J. (2001). Assessment
Tylka, T. L., & Subich, L. M. (2002). Exploring young women’s percep-
of eating disorders in women. The Counseling Psychologist, 29, 662–
tions of the effectiveness and safety of maladaptive weight control
techniques. Journal of Counseling & Development, 80, 101–110.
Kilbourne, J. (Producer), & Jhally, S. (Director). (2000). Killing us softly
Wampold, B. E., & Freund, R. D. (1987). Use of multiple regression in
3: Advertising images of women [Motion picture]. (Available from
counseling psychology research: A flexible data-analytic strategy. Jour-
Media Education Foundation,
nal of Counseling Psychology, 34, 372–382.
Mazzeo, S. E. (1999). Modification of an existing measure of body image
preoccupation and its relationship to disordered eating in female college
Received March 8, 2004
students. Journal of Counseling Psychology, 46, 42–50.
Revision received August 31, 2004
McKinley, N. M. (1998). Gender differences in undergraduates’ body
Accepted September 14, 2004