Sociocultural variables and socialization processes that render girls and women vulnerable to the development of body image and eating disturbances have been the focus of extensive theorizing and research over the past 30 years. The sociocultural model holds that social and cultural pressures towards thinness in women are reflected in "normative" dissatisfaction with weight, body size, and appearance among women in industrialized, Western countries (Fallon, Katzman, & Wooley, 1994; Striegel-Moore, Silberstein, & Rodin, 1986). Furthermore, according to self-objectification theory, the sociocultural emphasis on women's bodies leads them to adopt an observer's objective perspective and to treat themselves as objects to be evaluated on the basis of appearance (see Fredrickson & Roberts, 1997 for a review of objectification theory).
There is indisputable evidence that media messages in Western cultures emphasize a thin body ideal for women and actively promote weight loss methods (e.g., Thompson & Heinberg, 1999). There is also good evidence that these messages are internalized and are associated with increased self-objectification and body shape dissatisfaction in women (Morry & Staska, 2001). However, whereas most women express normative discontent with body size and appearance and many engage in weight-loss behaviours, only a minority go on to develop the more severe disturbances in body image, eating attitudes and behaviors that meet the criteria for eating disorders. The specific mechanisms that mediate the process by which normative discontent with body shape and weight becomes eating pathology have yet to be clearly identified, although some recent theoretically based research has yielded intriguing results.
Gender Socialization and Self-Worth
Gilligan (1982, 1990) proposed that whereas males are socialized to derive their sense of self-worth from instrumental achievements, self-worth in girls and women is tied to their self-perceived ability to form and maintain caring relationships. Building on this work from Gilligan, Jack and Dill (1992) identified four cognitive patterns that are often used by girls and women to maintain their focus on relationships with others. First, when judging themselves, girls and women are highly and emotionally influenced by peer evaluations, or "externalized self-perceptions." Second, girls and women are likely to "silence" or "hold in" their thoughts, needs, feelings, and opinions in order to avoid confrontations with others. Third, they are prone to solidify their relationships by putting the needs of others ahead of their own. Finally, they are likely to suppress anger and hostility and present themselves as nurturing, caring, and compliant.
Adherence to these socialized patterns has been linked to depressive symptoms in women and has been offered as an explanation of known gender differences in the prevalence of depression (Jack & Dill, 1992). More recently, the cognitive patterns described by Jack and Dill have also been found to relate to disturbed eating practices. For example, Geller, Cockell, and Goldner (2000) found that women with anorexia scored significantly higher on all four cognitive patterns than do women in psychiatric and normal control groups. Of the four cognitive schemas proposed by Jack and Dill, externalized self-perceptions and self-silencing appear to have particularly strong theoretical ties to eating pathology. Furthermore, Striegel-Moore, Silberstein, and Rodin (1993) detail clinical observations of both self-silencing and externalized self-perceptions in bulimics.
Externalized Self-Perceptions
Lieberman, Gauvin, Bukowski, and White (2001) found externalized self-perceptions to be negatively correlated with body-esteem and positively related to eating pathology in a sample of adolescent girls. In their examination of an adult nonclinical sample of women, Striegel-Moore et al. (1993) observed a relationship between low body-esteem and public self-consciousness, suggesting that women with more negative self-perceptions of body appearance are more concerned about what others think of them.
Other theorists have proposed that women are socialized to assess and think about their own bodies from the perspective of an outsider (i.e., in the third person.) Fredrickson, Roberts, Noll, Quinn, and Twenge (1998) found this kind of self-objectification to be associated with increased experience of body shame, self-consciousness, feelings of worthlessness and powerlessness, and problem eating behaviours.
Externalized Self-Perceptions and the Importance of Body Shape and Weight
The extent to which feelings of overall self-worth are based on body shape and weight is thought to be a core cognitive element in the etiology of eating disorders (Cooper & Fairburn, 1993; Fairburn, Peveler, Jones, Hope, & Doll, 1993). However, the relative importance of body shape and weight versus other self-concept domains has not been the focus of much research.
Geller and her colleagues demonstrated that the importance that women assign to their shape and weight is an important predictor of eating pathology in adolescent (Geller, Srikameswaran, Cockell, & Zaitsoff, 2000) and adult female samples (Geller, Johnston, & Madsen, 1997). Interestingly, whereas (Geller, SriKameswaran, et al., 2000) found that self-perceived importance of shape and weight was related to both body dissatisfaction and actual body weight in their adolescent sample, it was correlated only with body dissatisfaction in an adult sample (Geller et al., 1997). It may be that as women mature and it becomes more difficult to effect significant or permanent changes in weight, they learn to discount the importance of weight in favour of other self-concept domains (e.g., relationship success, vocational and educational achievements).
The tendency of women to base their judgments of self-worth on externalized or objectified self-perceptions has been examined in relation to body dissatisfaction, but it has not been looked at in relation to perceived importance of body shape and weight. Considering the importance placed on weight status by society, it is reasonable to speculate that a tendency to be influenced by the standards of peers would relate to perceptions of the importance of weight and shape.
Self-Silencing
Lerner, Hertzog, and Hooker (1988) postulated that self-silencing often emerges in the form of internalized anger, as girls and women learn that their expressions of anger are judged by others to be inappropriate and tend to push others away. Hooker and Convisser (1983) argued that a tendency to suppress anger may put women at risk of disturbed eating, as women may learn to "swallow" their anger with food when they feel distressed. More specifically, they propose that overeating gives women a reason to feel guilty and badly about themselves (for lacking willpower) thereby providing them with a path to direct their anger inwards, instead of towards others. Such hypotheses are difficult to test. However, symptomatic eating behaviour has been shown to relate to suppressed anger and self-silencing in both adolescent (Frank, Buchholz, & White, 1995; Zaitsoff, Geller, & Srikameswaran, 2002) and adult (Balfour, 1997) samples.
The Current Study
The aim of the current study was to examine the utility of externalized self-perceptions and self-silencing as predictors of eating pathology, over and above the traditionally powerful predictors of weight and body dissatisfaction. We also wanted to assess the extent to which the perceived importance of weight and body shape relates to body dissatisfaction and contributes to the prediction of eating pathology. This study differs from previous research in the following ways: a) We examined the utility of gender-socialized aspects of relational behaviour as predictors of eating pathology in a large sample of young women; b) We assessed the relative importance of weight and body shape as determinants of self-worth and looked at the relationship between externalized self-perceptions and perceived importance of body shape and weight; and c) We used a criterion measure of eating pathology that separates the cognitive and behavioural aspects of anorexic and bulimic eating pathology. These factors have been lumped together in many previous research. However, empirical evidence suggests that cognitions related to eating disturbance (i.e., distorted beliefs and negative attitudes about weight, body shape, food intake, dieting) are relatively common among young women in the general population (i.e., "normative"), whereas the behavioural components of these disorders (i.e., severe and sustained caloric restriction, regular purging or other compensatory behaviours) are observed much less frequently in nonclinical samples.
Based on the theoretical ideas and empirical research cited above, the following hypotheses were articulated. First, we hypothesized that externalized self-perceptions would be positively correlated with the self-perceived importance of body shape and weight. Second, we hypothesized that perceived importance of weight and body shape would account for unique variance in eating pathology over and above the traditionally powerful predictors of weight and body dissatisfaction variables. Finally, we hypothesized that (a) externalized self-perceptions and (b) self-silencing would improve the prediction of cognitions and behaviours associated with eating disorders over and above weight and body dissatisfaction and the perceived importance of body weight and shape.
Method
Participants
Participants were 236 undergraduate women aged 18 to 24 (M = 19.39, SD = 1.08) who were randomly selected from first-year students in the Psychology Department participant pool at the University of Windsor. Their Body Mass Index(1) (BMI, weight in kg / height in metres(2)), based on self-reported weight(2) and height, ranged from 16.15 to 34.49 (M = 22.61, SD = 3.45). Participants provided written consent and were treated in accordance with ethical principles for research with human subjects (Sales & Folkman, 2000). They earned course credit for their participation.
Measures and Procedure
Participants provided written consent, responded to demographic questions, and completed the following measures in small groups:
The Figure Rating Scale. The Figure Rating Scale (Stunkard, Sorenson, & Schlusinger, 1983) consists of nine silhouettes of female bodies ranging from very thin to very large. Participants indicate the silhouette that most closely resembles their perceived body shape, and the silhouette which best represents their ideal figure. The discrepancy score (between current and ideal figures) is taken as a measure of overall body dissatisfaction. Scores may range from -80 to 80, with higher absolute scores corresponding to greater body dissatisfaction. The scale has been shown to have adequate validity and reliability, and has been widely used in a variety of populations (Fallon & Rozin, 1985; Thompson, 1995).
The Weight-Esteem Subscale of the Body Esteem Scale for Adolescents and Adults (BESAA). The weight-esteem subscale of the BESAA (Mendelson, White, & Mendelson, 1996) is an 8-item Likert-type self-report instrument designed to measure the degree of satisfaction with body weight. Sample items from the BESAA include, "I am preoccupied with trying to change my body weight," "I am satisfied with my weight," and "my weight makes me unhappy." Scores are derived by taking the mean rating on the eight items, and can range from 1 to 5, with higher scores corresponding to greater satisfaction with body weight. The scale has been shown to have good test-retest reliability and validity in a college-age sample, and to discriminate between feelings related to appearance and feelings related to body weight (Mendelson et al.).
The Stirling Eating Disorder Scales (SEDS). The SEDS (Williams et al., 1994) is an 80-item self-report measure designed to assess a broad range of eating pathology. This is the only assessment tool to date that separately assesses the cognitions and behaviours associated with eating disorders. Four subscales provide measures of anorexic dietary cognitions, anorexic dietary behaviours, bulimic dietary cognitions, and bulimic dietary behaviours. Sample items on the anorexic dietary cognitions subscale include, "I find myself preoccupied with food," "When I eat anything I feel guilty," and "High carbohydrate foods make me feel nervous." Sample items on the anorexic dietary behaviours subscale include, "I eat the same food day after day," "I eat low calorie food all the time," and "I often hide food rather than eat it." Sample items on the bulimic dietary cognitions subscale include, "When I binge I have a sense of unreality," "When I binge, I feel disgusted with myself," and "I am not worried about my bingeing." Sample items on the bulimic dietary behaviours subscale include, "I eat a lot of food, even when I'm not hungry," "I never eat controllably," and "I intentionally vomit after eating." Items on the SEDS have been preassigned a weight, determined statistically during the construction of the scales. Subscale scores are derived by adding the weights of each of the endorsed items, with higher scores reflecting greater eating pathology. The SEDS has demonstrated excellent reliability and construct validity (Williams et al.).
The Shape and Weight-Based Self-Esteem (SAWBS) Inventory. Perceptions of the relative importance of body shape and weight were assessed with the Shape and Weight-Based Self-Esteem (SAWBS) Inventory (Geller et al., 1997). Completion of this instrument is a three-stage process. First, participants choose from a number of personal attributes that are important to how they have felt about themselves in the past four weeks. The list of attributes includes intimate or romantic relationships, body shape and weight, competence at school/work, personality, friendships, face, personal development, competence at activities other than school/work, and other. Second, participants rank order the chosen attributes according to their importance in determining self-opinion. Finally, participants are asked to divide a circle into pieces, such that the size of each piece reflects the extent to which the respondent's self-opinion is based on each chosen attribute. The score derived from the SAWBS is the angle (in degrees) of the piece allocated by the subject to body shape and weight. The instrument has demonstrated good convergent validity, discriminant validity, and reliability in a nonclinical adolescent sample, as well as in nonclinical and eating disordered adult samples (Geller et al., 1997; Geller, SriKameswaran, et al., 2000).
The Silencing the Self Scale: Externalized Self-Perceptions and Silencing the Self subscales. The Externalized Self-Perceptions subscale of the Silencing the Self Scale (Jack & Dill, 1992) was used to assess the tendency to base self-evaluations on the standards of peers. The Silencing the Self subscale of the same measure was used to assess self-silencing behaviours and attitudes. Given the age of participants in the current sample, we chose to use the adolescent version (adapted by Sippola & Bukowski, as cited in Buchholz & White, 1996). The adolescent version is identical to the adult version with one exception. Whereas the adult version asks about a participant's romantic relationship (i.e., "my partner"), the adolescent version asks about a participant's friendships (i.e., "my friend"). Sample items on the Externalized Self-Perceptions subscale include, "I tend to judge myself by how I think my friends see me," and "When I make decisions, my friends' thoughts and opinions influence me more than my own thoughts and opinions." Sample items on the Silencing the Self subscale include, "I don't tell my friends how I feel about some things when I know it will cause a conflict between us," and "I avoid getting into arguments with my friends." Scores on each subscale are determined by adding the ratings given for each of six Likert-type items that yield total scores ranging from 6 to 30. Higher scores reflect a greater tendency to rely on peer evaluations or more self-silencing behaviours and attitudes. The scale has demonstrated good construct validity and reliability (Buchholz & White, 1996; Jack & Dill, 1992).
Results
Means and standard,s deviations for all study variables are reported in Table 1. Although the SEDS is not designed to formally diagnose participants with an eating disorder, Williams et al. (1994) have suggested cut-off scores for identifying probable clinical cases. Using these cut-off scores, 33.7% of the current sample scored in the clinical range on anorexic dietary cognitions. Another 4% of the sample scored in the clinical range on anorexic dietary behaviours (an additional 7% scored in the subclinical range). With respect to bulimic dietary cognitions, 16% of the sample scored in the clinical range, and 18% of the sample scored in the clinical range on bulimic dietary behaviours. These data suggest that there is enough variability on these criterion measures for multivariate analyses to be appropriate since there is a reasonable range and distribution of disturbed levels of eating-disordered cognitions and behaviours in the current sample.
As expected, externalized self-perceptions were significantly correlated with both figure rating scores (r =.20, p <.01) and the BESAA-weight-esteem questionnaire scores (r = -.23, p <.01). However, contrary to our first hypothesis, externalized self-perceptions were not significantly correlated with perceived importance of body shape and weight (r =.05, p >.05).
[a]p <.05;
[b]p <.01.
The second and third study hypotheses were tested in a series of four hierarchical multiple regression analyses to predict (a) anorexic dietary cognitions, (b) anorexic dietary behaviours, (c) bulimic dietary cognitions, and (d) bulimic dietary behaviours, as assessed by the Stirling Eating Disorder Scales. In each of these analyses, Body Mass Index and the two body dissatisfaction variables (figure rating and BESAA-weight-esteem scores) were entered on the first step as control variables. Perceived importance of body shape and weight was entered on the second step, in order to examine its unique contribution to eating pathology when body weight and body dissatisfaction are accounted for. Externalized self-perceptions and self-silencing behaviours and attitudes were entered on the third step, in order to examine their unique contributions to eating pathology when body weight, body dissatisfaction, and perceived importance of body shape and weight are held constant. All possible two-way, three-way, and four-way interactions were tested on the fourth, fifth, and sixth steps, respectively. None of the interactions in any of the analyses were found to be significant. Therefore, the regression analyses are presented without the interaction terms.
Predicting Anorexic Dietary Cognitions
The R[Symbol Not Transcribed] was significantly different from zero at all three steps of the regression. After the final step, with all variables in the equation, adjusted R[Symbol Not Transcribed] =.38, multiple R =.63, F(6, 229) = 25.46, p <.001. See Table 2.
At step one, Body Mass Index and body dissatisfaction accounted for 35% of the variance in anorexic dietary cognitions (R[Symbol Not Transcribed] =.35, F(3, 232) = 41.36, p <.001). Body mass index (t = -4.12, p <.001), the BESAA weight-esteem measure of body dissatisfaction (t = -7.30, p <.001), and the figure rating measure of body dissatisfaction (t = 2.46, p <.05) were each found to be significant predictors of anorexic dietary cognitions (explaining 5%, 15%, and 2% unique variance, respectively). At step two, perceived importance of body shape and weight significantly improved the prediction of anorexic dietary cognitions, explaining 2% incremental unique variance (R[Symbol Not Transcribed][Symbol Not Transcribed] =.02, F[Symbol Not Transcribed] (4, 231) = 8.71, p <.01). At step three, self-silencing behaviours and attitudes and externalized self-perceptions accounted for an additional 3% of the variance in anorexic dietary cognitions (R[Symbol Not Transcribed][Symbol Not Transcribed] =.03, F[Symbol Not Transcribed](6, 229) = 5.35, p <.01). Externalized self-perceptions were found to be a significant predictor (t = 2.31, p <.05), whereas self-silencing behaviours and attitudes was not found to be a unique contributor (t = 1.04, p >.05).
Predicting Anorexic Dietary Behaviours
[a]p <.01.
The R[Symbol Not Transcribed] at step one was only significantly different from zero, but not significantly different from zero on subsequent steps. After the final step, with all variables in the equation, adjusted R[Symbol Not Transcribed] =.18, multiple R =.45, F(6, 229) = 9.54, p <.001. See Table 3.
At step one, Body Mass Index and body dissatisfaction accounted for 18% of the variance in anorexic dietary behaviours (R[Symbol Not Transcribed] =.18, F(3, 232) = 17.35, p <.001). Whereas Body Mass Index (t = -3.45, p <.001) and the BESAA weight-esteem measure of body dissatisfaction (t = -4.66, p <.001) were significant predictors of anorexic dietary behaviours (each explaining 4% and 8% unique variance, respectively), the figure rating measure of body dissatisfaction was not a unique contributor. At step two, perceived importance of body shape and weight did not improve the prediction of anorexic dietary behaviours (R[Symbol Not Transcribed][Symbol Not Transcribed] =.01, F[Symbol Not Transcribed] (4, 231) = 2.87, p >.05). At step three, self-silencing behaviours and attitudes and externalized self-perceptions also did not improve the prediction of anorexic dietary behaviours (R[Symbol Not Transcribed][Symbol Not Transcribed] =.01, F[Symbol Not Transcribed] (6, 229) =.96, p >.05).
Predicting Bulimic Dietary Cognitions
The R[Symbol Not Transcribed] was significantly different from zero at all three steps of the regression. After the final step, with all variables in the equation, adjusted R[Symbol Not Transcribed] =.33, multiple R =.59, F(6, 229) = 20.10, p <.001. See Table 4.
At step one, Body Mass Index and body dissatisfaction accounted for 29% of the variance in bulimic dietary cognitions (R[Symbol Not Transcribed] =.29, F(3, 232) = 31.94, p <.001). Body Mass Index (t = -2.03, p <.05), the BESAA weight-esteem measure of body dissatisfaction (t = -5.60, p <.001), and the figure rating measure of body dissatisfaction (t = 2.57, p <.01) were each found to be significant predictors of bulimic dietary cognitions (explaining 1%, 10%, and 2% unique variance respectively). At step two, perceived importance of body shape and weight significantly improved the prediction of bulimic dietary cognitions, explaining 2% incremental unique variance (R[Symbol Not Transcribed][Symbol Not Transcribed] =.02, F[Symbol Not Transcribed] (4, 231) = 6.41, p <.01). At step three, self-silencing behaviours and attitudes and externalized self-perceptions accounted for an additional 3% of the variance, (R[Symbol Not Transcribed][Symbol Not Transcribed] =.03, F[Symbol Not Transcribed] (6, 229) = 5.86, p <.01). Externalized self-perceptions was found to be a significant predictor (t = 2.45, p <.05), whereas self-silencing behaviours and attitudes was not found to be a unique contributor (t = 1.04, p >.05).
Predicting Bulimic Dietary Behaviours
[a]p <.05;
[b]p <.01.
[a]p <.05;
[b]p <.01.
The R[Symbol Not Transcribed] was significantly different from zero for steps one and three of the regression. After the final step, with all variables in the equation, adjusted R[Symbol Not Transcribed] =.27, multiple R =.54, F(6, 229) = 15.47, p <.001. See Table 5.
At step one, Body Mass Index and body dissatisfaction accounted for 25% of the variance in bulimic dietary behaviours (R[Symbol Not Transcribed] =.25, F(3, 232) = 26.39, p <.001). Body Mass Index (t = -2.68, p <.01), the BESAA weight-esteem measure of body dissatisfaction (t = -4.79, p <.001), and the figure rating measure of body dissatisfaction (t = 2.99, p <.01) were each found to be significant predictors of bulimic dietary behaviours (explaining 2%, 7%, and 3% unique variance, respectively). At step two, perceived importance of body shape and weight did not improve the prediction of bulimic dietary behaviours (R[Symbol Not Transcribed][Symbol Not Transcribed] =.00, F[Symbol Not Transcribed] (4, 231) = 0.49, p >.05). At step three, self-silencing behaviours and attitudes and externalized self-perceptions accounted for an additional 3% of the variance (R[Symbol Not Transcribed][Symbol Not Transcribed] =.03, F[Symbol Not Transcribed] (6,229) = 5.21, p <.01). Self-silencing behaviours and attitudes was found to be a significant predictor (t = 2.02, p <.05), whereas externalized self-perceptions was not found to be a unique contributor (t = 1.33, p >.05).
Discussion
Our first hypothesis that externalized self-perceptions would be correlated with self-perceived importance of body shape and weight was not confirmed. It may be that by the time women reach early adulthood, perceptions of the importance of body shape and weight have already been incorporated into personal value systems and are no longer readily influenced by the evaluations of others. It may also be that, while postadolescent women do perceive that societal standards emphasize thinness, they do not necessarily perceive those standards to emphasize body shape and weight over and above other personal attributes (e.g., interpersonal skills, academic competence).
Our second hypothesis, that perceived importance of weight and body shape would predict eating pathology, after accounting for the contributions of weight and body dissatisfaction was partly supported. The importance placed by a woman on her body shape and weight contributed significantly to predictions of both anorexic and bulimic cognitions, but did not improve the predictions of either anorexic or bulimic behaviours. Because our sample was a nonclinical one, we did not expect to see the degree of symptomatic behaviour that would be characteristic of a clinical sample. Looking at these results in the context of a continuum model of eating disorders (Scarano & Kalodner-Martin, 1994; Stice, Killen, Hayward, & Taylor, 1998), it appears conceivable that the relative importance a women places on body weight and shape may lead to distorted thinking and attitudes characteristic of eating disturbances, but that additional triggers must be present in order for such thinking to impact actual behaviour.
Third, we assessed the utility of both externalized self-perceptions and self-silencing as predictors of anorexic and bulimic cognitions and behaviours once body weight, body dissatisfaction, and perceived importance of body weight and shape had been accounted for. As hypothesized, externalized self-perceptions improved the prediction of both anorexic and bulimic dietary cognitions. However, externalized self-perceptions did not significantly predict anorexic or bulimic behaviours. Again, it may be that externalized self-perceptions do not impact on behaviour in the absence of other trigger variables. Geller et al. (1998) found that perceived importance of body shape and weight effectively discriminated eating disordered individuals from psychiatric and normal controls. It would be interesting to assess the extent to which externalized self-perceptions could also discriminate between such groups. We could not conduct this comparison in the current study, as it was not possible to identify an adequate number of participants who met DSM-IV criteria for clinical eating disorders (American Psychiatric Association, 1994).
Self-silencing did not significantly improve the prediction of anorexic cognitions or behaviours after controlling for weight, body dissatisfaction, and perceived importance of body shape and weight. The observation that self-silencing did not improve the prediction of anorexic dietary behaviours in the current study is inconsistent with results reported by Geller, Cockell, et al. (2000), who demonstrated that women with anorexia nervosa score higher on self-silencing behaviours and attitudes than controls. One possible explanation for the discrepancy is that anorexic pathology was relatively infrequent in our sample (i.e., only 4% met the cut-off score).
Neither did self-silencing improve the prediction of bulimic dietary cognitions. However, self-silencing was a modestly significant predictor of bulimic dietary behaviours. Zaitsoff et al. (2002) found that eating symptom scores correlated with both suppressed anger and self-silencing scores in an adolescent female sample. Suppressed anger has also been shown to be related to emotional eating in both college-age women (Balfour, 1997) and adolescent girls (Frank et al., 1995). Together, the current and previous findings are consistent with the idea introduced by Hooker and Convisser (1983) that a tendency to suppress feelings (anger-related emotions in particular) may predispose women to cope by "swallowing" their pent up feelings with food.
We speculate that the importance women place on body shape and weight and their tendencies to base self-evaluations on the standards of others (i.e., externalized self-perceptions) may be prodromal factors in the development of eating disturbances. Such perceptions may put women at risk for maladaptive eating-related cognitions that increase their future vulnerability to disturbances in eating behaviour (i.e., excessive dieting, bingeing, and purging). Self-silencing may lead women to cope with bottled up feelings and negative thoughts by engaging in problematic eating behaviours. Further research is necessary to more fully assess the role of self-silencing behaviours and attitudes in the maintenance of binge eating and other bulimic behaviours.
Anorexic and bulimic cognitions and behaviours were assessed only by self-report measures, and are subject to the usual response biases characteristic of such assessment methods. For example, it is possible that participants underreported eating pathology. Moreover, the sample used in the current study was a nonclinical sample. It is not clear to what extent the present findings would generalize to women with eating disorder diagnoses. It may be most useful to consider the present findings in the context of a prevention model. Eating-disordered cognitions and body dissatisfaction may in fact be relatively normative for undergraduate women, whereas eating-disordered behaviours are more rare, and may only emerge with the presence of other specific triggers. While further research is needed to identify such trigger variables, the current results suggest that self-silencing tendencies may act as such a trigger.
There is a complex web of relationships between the variables examined that makes it difficult to tease out which variables are influencing which. An alternative way to explore these data would be through path analyses or structural equation modeling. Such endeavours could provide further insight into the directionality of the correlations. Ideally, further research could examine these variables longitudinally, in an effort to answer questions related to the temporal or causal relationships between these variables. It is certainly important that attitudinal or cognitive symptoms and behavioural manifestations be assessed and researched as related but differentiated characteristics of eating pathology.
Overall, the results of this study suggest that further research on relational variables shaped by gender role socialization processes (i.e., externalized self-perceptions and self-silencing) could usefully contribute to our efforts to understand the etiology of eating disturbances and develop more effective prevention and treatment programs. Women with eating disorders and those who are at risk could be taught to recognize and accept their emotional experiences; effectively express their wants, needs, and feelings; de-emphasize the importance of their shape and weight, and internalize a more balanced view of personal attributes.
The authors are grateful to Dr. Stewart Page, Dr. Vicky Paraschack, Dr. Josie Geller, Dory Becker, Sophie Beugnot, Gabriella Geller, Alicia Hendley, Melanie Kelly, Rolan Koifman, Jillian Leggatt, Dan Taylor, and Jane Walsh, all of whom provided critical feedback about this project from its inception to completion. Dr. Stewart Page and Dr. Josee Jarry made helpful comments and suggested revisions to the manuscript.
Correspondence concerning this article should be addressed to Jeremy Frank, Department of Psychology, University of Windsor, 401 Sunset Avenue, Windsor, Ontario, Canada N9B 2T9 (E-mail: jfrank@sent.com).
References
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.
Balfour, L. (1997). Examining how avoidant coping and anger suppression relate to emotional eating in young women. Dissertation-Abstracts-International-B, 57(7-B), 4693.
Bowman, R. L., & Delucia, J. L. (1992). Accuracy of self-reported weight: A meta-analysis. Behavior Therapy, 23, 637-655.
Buchholz, A., & White, D. R. (1996). Living as images, appearance self-esteem and disordered eating in adolescent girls. Paper presented as part of a symposium at the XIVth Biennial Meeting of the International Society for the Study of Behavioural Development, Quebec City, Canada.
Cooper, P. J., & Fairburn, C. G. (1993). Confusion over the core psychopathology of bulimia nervosa. International Journal of Eating Disorders, 13, 385-389.
Fairburn, C. G., Peveler, R. C., Jones, R., Hope, R. A., & Doll, H. A. (1993). Predictors of 12-month outcome in bulimia nervosa and the influence of attitudes to shape and weight. Journal of Consulting and Clinical Psychology, 61, 696-698.
Fallon, A. E., & Rozin, P. (1985). Sex differences in perceptions of desirable body shape. Journal of Abnormal Psychology, 94, 102-105.
Fallon, P., Katzman, M. A., & Wooley, S. C. (1994). Feminist perspectives on eating disorders. New York: Guilford Press.
Frank, J. B., Buchholz, A., & White, D. R. (1995, May). Internalized anger and eating behavior in adolescent girls. Paper presented at the Quebec Child Psychiatry Research Day, Montreal Children's Hospital, Montreal, Canada.
Fredrickson, B. L., & Roberts, T. A. (1997). Objectification theory: Toward understanding women's lived experiences and mental health risks. Psychology of Women, 21, 173-206.
Fredrickson, B. L., Roberts, T., Noll, S. M., Quinn, D. M., & Twenge, J. M. (1998). That swimsuit becomes you: Sex differences in self-objectification, restrained eating, and math performance. Journal of Personality and Social Psychology, 75, 269-284.
Geller, J., Cockell, S. J., & Goldner, E. M. (2000). Inhibited expression of negative emotions and interpersonal orientation in anorexia nervosa. International Journal of Eating Disorders, 28, 8-19.
Geller, J., Johnston, C., & Madsen, K. (1997). A new measure of the role of shape and weight in self-concept: The Shape and Weight-Based Self-esteem Inventory. Cognitive Therapy and Research, 21, 5-24.
Geller, J., Johnston, C., Madsen, K., Goldner, E., Remic, R. A., & Birmingham, C. L. (1998). Shape-and weight-based self-esteem and the eating disorders. International Journal of Eating Disorders, 24, 285-298.
Geller, J., Srikameswaran, S., Cockell, S. J., & Zaitsoff, S. L. (2000). Assessment of shape- and weight-based self-esteem in adolescents. International Journal of Eating Disorders, 28, 339-345.
Gilligan, C. (1982). In a different voice: Psychological theory and women's development. Cambridge, MA: Harvard University Press.
Gilligan, C. (1990). Teaching Shakespeare's sister: Notes from the underground of female adolescence. In C. Gilligan, N. Lyons, & T. J. Hanmer (Eds.), Making connections: The relational worlds of adolescent girls at Emma Willard School (pp. 6-29). Cambridge, MA: Harvard University Press.
Hannan, W. J., Wrate, R. M., Cowen, S. J., & Freeman, C. P. L. (1995). Body mass index as an estimate of body fat. International Journal of Eating Disorders, 18, 91-97.
Hooker, D., & Convisser, E. (1983). Women's eating problems: An analysis of a coping mechanism. Personnel-and-Guidance-Journal, 62, 236-239.
Jack, D. C., & Dill, D. (1992). The Silencing the Self Scale: Schemas associated with depression in women. Psychology of Women Quarterly, 16, 97-106.
Lerner, J. V., Hertzog, C., & Hooker, K. A. (1988). A longitudinal study of negative emotional states and adjustment from early childhood through adolescence. Childhood Development, 59, 356-366.
Lieberman, M., Gauvin, L., Bukowski, W., & White, D. R. (2001). Interpersonal influence and disordered eating behaviours in adolescent girls: The role of peer modelling, social reinforcement, and body-related teasing. Eating Behaviours, 2, 215-236.
Mendelson, B. K., White, D. R., & Mendelson, M. J. (1996). Self-esteem and body-esteem: Effects of sex, age, and weight. Journal of Applied Developmental Psychology, 17, 321-346.
Morry, M. M., & Staska, S. L. (2001). Magazine exposure: Internalization, self-objectification, eating attitudes, and body satisfaction in male and female university students. Canadian Journal of Behavioural Science, 33, 269-279.
Sales, B. D., & Folkman, S. (Eds.). (2000). Ethics in research with human subjects. Washington, DC: American Psychological Association.
Scarano, G. M., & Kalodner-Martin, C. R. (1994). A description of the continuum of eating disorders: Implications for intervention and research. Journal of Counseling and Development, 72 (4), 356-361.
Stice, E., Killen, J. D., Hayward, C., & Taylor, C. B. (1998). Support for the continuity hypothesis of bulimic pathology. Journal of Consulting and Clinical Psychology, 66 (5), 784-790.
Striegel-Moore, R. H., Silberstein, L. R., & Rodin, J. (1986). Toward an understanding of risk factors for bulimia. American Psychologist, 41, 246-263.
Striegel-Moore, R. H., Silberstein, L. R., & Rodin, J. (1993). The social self in bulimia nervosa: Public self-consciousness, social anxiety, and perceived fraudulence. Journal of Abnormal Psychology, 102, 293-303.
Stunkard, A., Sorenson, T., & Schlusinger, F. (1983). Use of the Danish Adoption Register for the study of obesity and thinness. In S. Kety, L. P. Rowland, R. L. Sidman, & S. W. Matthysse (Eds.), The genetics of neurological and psychiatric disorders (pp. 115-120). New York: Raven Press.
Thompson, J. K. (1995). Assessment of body image. In D. Allison (Ed.), Handbook of assessment methods for eating behaviors and weight-related problems: Measures, theory and research (pp. 119-148). Thousand Oaks, CA: Sage.
Thompson, J. K., & Heinberg, L. J. (1999). The media's influence on body image disturbance and eating disorders: We've reviled them, now can we rehabilitate them? Journal of Social Issues, 55, 339-353.
Williams, G. J., Power, K. G., Miller, H. R., Freeman, C. P., Yellowless, A., Dowds, T., Walker, M., & Parry-Jones, W. L. (1994). Development and validation of the Stirling Eating Disorder Scales. International Journal of Eating Disorders, 16, 35-43.
Zaitsoff, S. L., Geller, J., & Srikameswaran, S. (2002). Silencing the self and suppressed anger: Relationship to eating disorder symptoms in adolescent females. European Eating Disorders Review, 10, 51-60.
Received February 10, 2002
Revised September 8, 2002
Accepted September 18, 2002
(1) Body Mass Index (BMI) has been shown to be a more accurate measurement of body fat than weight measures, and is highly related to percentage body fat (Hannan, Wrate, Cowen, & Freeman, 1995).
(2) Self-reported weight has been shown to be a sufficiently accurate measurement of actual weight in nonclinical weight-loss samples (Bowman & Delucia, 1992).
