Journal Clothing and Textiles Research

38 downloads 6813 Views 1MB Size Report
Email Alerts: ... Permissions: http://ctr.sagepub.com/content/19/4/171.refs.html. Citations: ..... Seid, 1989), marketing (Peterson, 1987; Richins, 1991), and.
Clothing and Textiles Research Journal http://ctr.sagepub.com/

Self-Schema or Self-Discrepancy? Which Best Explains Body Image? Jaehee Jung, Sharron J. Lennon and Nancy A. Rudd Clothing and Textiles Research Journal 2001 19: 171 DOI: 10.1177/0887302X0101900403 The online version of this article can be found at: http://ctr.sagepub.com/content/19/4/171

Published by: http://www.sagepublications.com

On behalf of:

International Textile and Apparel Association

Additional services and information for Clothing and Textiles Research Journal can be found at: Email Alerts: http://ctr.sagepub.com/cgi/alerts Subscriptions: http://ctr.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://ctr.sagepub.com/content/19/4/171.refs.html

>> Version of Record - Jan 1, 2001 What is This?

Downloaded from ctr.sagepub.com at UNIV OF DELAWARE LIB on December 28, 2013

Self-Schema or Self-Discrepancy? Which Best Explains Body Image? Jaehee Jung Sharron J. Lennon Nancy A. Rudd Abstract We investigated the discrepancy between women’s actual and ideal appearance and its relationship to body image, mood, and self-esteem. In a two-part experimental design, 88 college women completed measures of appearance self-discrepancy and appearance self-schema. Two weeks later they were exposed to 16 photos of attractive media images and subsequently completed measures of mood, body image, and self esteem. One-way between subjects multivariate and univariate analyses of variance revealed that appearance self-discrepancy was related to all the body image variables and self-esteem, but not to the mood variables. Appearance self-schema may better explain body image than appearance self-discrepancy. Results are discussed in light of self-discrepancy theory, schema theory, and the Rudd and Lennon (1994) model of body aesthetics.

Jung, J., Lennon, S. J., & Rudd, N. A. (2001). Self-schema or self-discrepancy? Which best explains body image? Clothing and Textiles Research Journal, 19(4), 171-184. Key Words: body image, body aesthetics, appearance self-

discrepancy,

appearance self-schema.

Women in many cultures, and particularly in western cultures, are socialized to be interested in beauty and fashion (Brownmiller, 1984; Freedman, 1988; Rodin, Silberstein, & Striegel-Moore, 1984; Timko, Striegel-Moore, Silberstein, & Rodin, 1987). In societies like the United States in which femininity is linked to attractiveness, people are likely to consider attractiveness, including thinness (Silverstein, Perdue, Peterson, & Kelly, 1986; Silverstein, Peterson, & Perdue, 1986), a desirable physical trait for women. Women engage in a variety of appearance-management behaviors to construct and maintain an attractive appearance (Rudd & Lennon, 1994). Those behaviors are reinforced by evaluations from family members (Levine, Smolak, Moodey, Shuman, & Hessen, 1994; Paxton et al.,1991 ), peers (Fabian & Thompson, 1989; Thompson & Heinberg, 1993), and significant others, as well as by messages from the media (Pinhas, Toner, Ali, Garfinkel, & Stuckless, 1999; Stice, Schupak-Neuberg, Shaw, & Stein, 1994). A woman’s

(Cash, 1990). It is not a physical construct but a mental image. bound criteria of what is attractive play an important part in the development of body image because body image is affected by cultural aesthetic standards (Fallon, 1990; Grogan, 1999; Hesse-Biber, 1996). Therefore, the way women evaluate themselves, including their bodies, is likely to be influenced by cultural ideology (Tseelon, 1995). When the perception of one’s body matches cultural standards,’1 strong self-esteem and positive attitudes toward and feelings about one’s body are the likely result (Rudd & Lennon, 1994). But low self-esteem and negative attitudes and feelings may result when one’s body is discrepant from these standards. Researchers have found that such discrepancies are correlated with mood, such as depression and anxiety (Cash & Szymanski, 1995) and dejection and agitation (Szymanski & Cash, 1995). Women typically fail to approximate the cultural ideal, which is often narrow and can only

Culturally

body image is her perception of her body

and, equally important, how she feels about that perception

Authors’ Addresses: Jaehee Jung, Department of Apparel Merchandising & Interior Design, Indiana University, Bloomington, IN 47405; Sharron J. Lennon and Nancy A. Rudd, Department of Consumer & Textile Sciences, Ohio State University, Columbus, OH 43210-1295.

’For example, in the.United States at the end of the 20th century such standards for women include thinness (e.g., Berscheid, Walster, & Bohrnstedt, 1973; Cash, Winstead, & Janda, 1986; Douty & Brannon, 1984; Lennon, 1988; Mahoney & Finch, 1976; Miller, Coffman, & Linke, 1980), fitness (Banner, 1983; Beuf,1990; Freedman,1986), blond hair, blue or light eyes, an attractive face, big eyes, long legs, narrow hips, and a well-defined bust (Rudolph,1991; "The Body Game," 1993).

171 Downloaded from ctr.sagepub.com at UNIV OF DELAWARE LIB on December 28, 2013

achieved by a small percentage of women. As a result, women are likely to perceive a discrepancy between their actual appearance (what they perceive) and their ideal appearance (what they desire). An attractive appearance has important social consequences (Jackson, 1992), especially for women, due to greater social pressure placed upon them (Mazur, 1986) and the influence it has on their body images (Cooper, Anastasiades, & Fairbum, 1992; Cooper & Fairbum, 1992). Since body image is a mental image, it may be construed as a mental representation of information concerning one’ss body (Altabe & Thompson, 1996); as such it is a mental construction of the self or self-schema (Markus, 1977; Markus & Sentis, 1982). The extent to which individuals are satisfied or dissatisfied with their bodies in relation to cultural ideals may be dependent on how important the domain of physical appearance is to them (i.e., its centrality, influence, and self-relevance). Thus, even if one’s physical attributes do not reflect cultural standards for an attractive appearance, body satisfaction may result if appearance is not central to the individual’s self-concept. In this way, a woman might recognize that her body differs from the cultural standard and yet be accepting of that difference due to a variety of factors (e.g., valuing competence over appearance or a belief in genetic programming regarding

realistically be



body shape). The purpose of the research reported here was (a) to investigate the relationship between appearance self-discrepancy and measures of body image, mood, and self-esteem and (b) to assess the relative usefulness of two competing perspectives (self-discrepancy and self-schema) in explaining body image. In this paper cognitive and sociocultural theories are introduced as theoretical perspectives, specific relationships among the variables are developed, and those hypothesized relationships are examined empirically. The relative utility of two cognitive theoretical perspectives (i.e., self-discrepancy and self-schema) in explaining body image is investigated empirically. Finally, the extent to which measures

of appearance self-schema and appearance self-

discrepancy

are

related is assessed.

Review of the Literature

Body Image Body image has both attitudinal and perceptual dimensions. The attitudinal dimension consists of (a) appearance self appraisals and subsequent affect and (b) body-image investment (i.e., the salience, extent, and centrality of emphasis on appearance) (Cash, 1994; Cash & Szymanski, 1995; Szymanski & Cash, 1995). The perceptual aspect of body image (i.e., what people perceive in terms of their bodies such as shape, weight, attractiveness, and skin color) is evident in studies that investigate distortions of body size (Monteath & McCabe, 1997) and weight (Cash & Hicks, 1990). For example, Cash and Hicks investigated body image, eating and dieting behaviors, and psychosocial wellbeing in a group of normal-weight men and women. People who self-labeled as overweight evaluated their appearances

negatively than people who self-labeled as normal in weight. Compared to the latter group of people, the former group were more anxious about being fat, more likely to be currently dieting, and reported both more restrained eating and more binge eating. In other words, what normal-weight people &dquo;see&dquo; is often objectively inaccurate, but yet affects subsequent feelings and behaviors. According to Cash, Wood, Phelps, and Boyd ( 1991 ), the more occupied a woman is with overestimation of her weight and a fear of being fat, the greater is her risk of body dissatisfaction, disordered eating behaviors, and depression. Thus, perceptual distortion or a discrepancy between &dquo;what is&dquo; and &dquo;what one would like to be&dquo; is associated with negative body image attitudes more

and behaviors.

According to Cash and Pruzinsky (1990), body image complex construct encompassing one’s cognitions, emotions, and actions regarding one’s body. Nevertheless, the primary focus of body image research has been on cognitive and behavioral aspects (e.g., Altabe & Thompson, is

a

1996; Cash & Hicks, 1990; Harrison & Cantor, 1997; Irving, 1990 ; Martz, Handley, & Eisler, 1995; Myers & Biocca, 1992; Stice et al.,1994; Striegel-Moore, McAvay, & Rodin,

1986) and relatively fewer studies have assessed mood and other affective aspects of body image (e.g., Cash et al.,1991; Heinberg & Thompson, 1995; Kenrick, Montello, Gutierres, & Trost, 1993; Myers & Biocca, 1992; Pinhas et al., 1999). In a study by Heinberg and Thompson, research participants who viewed a videotape containing images that reflected cultural ideals of thinness and attractiveness exhibited greater depression, anger, and appearance dissatisfaction than those who viewed neutral images unrelated to appearance. Stice and Shaw (1994) exposed college women to attractive images from magazines and found that the exposure produced a &dquo;negative affective state&dquo; (depression, stress, guilt, shame, insecurity, and body dissatisfaction), which predicted bulimic tendencies. Pinhas et al. (1999) examined women’s mood states before and after exposure to images of fashion models from fashion magazines or to neutral images of objects (no human figures in them). Research participants were both more depressed and angrier after exposure to idealized fashion images as compared to exposure to the neutral images. Thus, exposure to images depicting cultural ideals of beauty and thinness affects mood and dissatisfaction when compared to a control group or when a pretest and post-test design is used. Over time such exposure may lead to the normative discontent (Rodin et al., 1984) women experience regarding their weight.

Theoretical

Background

The body image literature can be explained using the sociocultural perspective (e.g., Fallon, 1990; Richins, 1991). Another approach to the study of body image is a cognitive

perspective; two theories that take a cognitive approach to body image are schema theory and self-discrepancy theory. A third general framework from which to study body image is the Rudd and Lennon (1994) model of body aesthetics, which builds

on

the sociocultural

172 Downloaded from ctr.sagepub.com at UNIV OF DELAWARE LIB on December 28, 2013

approach

from both

an

aesthetic and a cognitive perspective. In this section we discuss the sociocultural perspective, schema theory, selfdiscrepancy theory, the Rudd and Lennon model, and related literature. Sociocultural Perspective What is considered attractive at any given time and place is largely determined by one’s sociocultural context (Brumberg, 1997; Fallon, 1990; Polivy, Garner, & Garfinkel, 1986). In other words, just as what is considered attractive in clothing is subject to fashion change, what is considered attractive in terms of body shape and size is also subject to fashion change (Banner, 1983; Seid, 1989). Within a culture, these fashions for clothing and bodies are reflected in cultural images. In fact, we alter our bodies through diet, exercise, power slips and Wonder bras, and makeup choices to achieve the fashionable body. In this way, cultural images of what is attractive affect the development of body image because we (and others) evaluate our bodies through the lens of culture. Furthermore, others respond to us on the basis of our appearances as a function of images that are idealized in our culture. A sociocultural approach to body image has been used by researchers from journalism and communication (e.g., Harrison & Cantor, 1997; Myers & Biocca, 1992), history and women’s studies (e.g., Banner, 1983; Brumberg, 1997; Seid, 1989), marketing (Peterson, 1987; Richins, 1991), and psychology (Garner & Garfinkel, 1981; Silverstein et al.,

1986; Wiseman, Gray, Mosimann, & Ahrens, 1992). For several studies have content analyzed media imand that the ideal body type for women demonstrated ages has become thinner since the 1950s (Garner & Garfinkel, 1981; Silverstein et al., 1986; Wiseman et al., 1992). These studies provide evidence that the mass media play a role in the promotion of a thin standard of beauty for women. These media images then provide the backdrop against which body image is shaped and assessed. Researchers demonstrating (a) a normative discontent (Rodin et al., 1984) among women regarding their weights and (b) overwhelming body dissatisfaction among women (Fallon & Rozin, 1985) attest to the efficacy of the sociocultural perspective in ex-

example,

plaining body image.

Cognitive Perspectives A sociocultural perspective does not explain (a) why some people2 are more susceptible to media messages within a given sociocultural context than others (e.g., Mori & Morey, 1991; Peterson, 1987), (b) why some people seem to be happy with their physical packaging even when it deviates from the cultural ideal (Rosen, 1996), or (c) why some are dissatisfied with their bodies even when they do approximate the cultural ideal (Hesse-Biber, 1996). These issues can be addressed by using a cognitive approach from psychology; two such perspectives that have been applied to body image are schema theory (Markus, Hamill, & Sentis, 1987) and self-

Mori and Morey (1991) found that depressed individuals were at risk for 2 internalizing media messages, while Peterson (1987) found that students with anorexia or bulimia were likely to be influenced by media messages.

discrepancy theory (Strauman & Glenberg, 1994; Strauman, Vookles, Berenstein, Chaiken, & Higgins, 1991). Schema theory. A self-schema (Markus, 1977) is a cognitive generalization about the self. Self-schemas are important in processing information (Markus, Hamill, & Sentis, 1987) and they seem to function as important dimensions in one’s self-definition. According to Myers and Biocca (1992), a self-schema is a person’s construction of those traits that make the person distinctive and constitute the sense of &dquo;me.&dquo; Individuals build this sense of self from observation of their own behaviors, the reaction of others to the self, and more general social cues that suggest which attributes of the self are most important (p. 115). Self-schemas may bias individuals to selectively attend to information of greater relative importance in information processing. For example, we might expect that anorexics and bulimics would pay more attention to and be more concerned with issues related to food, body size, and shape as compared to individuals without eating disorders. In fact, some researchers have examined information processing among anorexics and bulimics and found them to selectively attend to food and shape information as compared to other types of information (Channon, Hemsley, & Silva, 1988; Cooper, Anastasiades, & Fairburn,1992; Cooper & Fairbum, 1992). Individuals with eating disorders seem to have distorted attitudes toward eating, weight, and shape, which influence the maintenance of disturbed body image. Thus, when people develop schemas for appearance such that appearance is salient, central, and influential, those people are likely to selectively attend to appearance information in themselves and others. Self-discrepancy theory. Another cognitive theory which has been applied to the study of body image is self-

concept discrepancy theory or simply self-discrepancy theory

(Higgins, 1987). According to the theory, disparities may exist between those attributes possessed by the actual (or real) self and those attributes the individual would like to possess (ideal self). These disparities may lead to an unpleasant internal psychological state, which motivates reduction of the perceived self-discrepancies. When the discrepancy is large it may reflect psychopathology such as depression (Higgins, Klein, & Strauman, 1987). In terms of content, researchers have shown that both the actual self and the ideal self contain information regarding academic performance, athletic ability, social acceptance, and physical appearance (Bybee, Glick, & Zigler, 1990; Bybee, Luthar, Zigler, & Merisca, 1997). Cultural standards for ideal beauty foster the development of discrepancies between the actual and ideal self because they are used as criteria for evaluating one’s body and general physical appearance. Cultural ideals are reflected in media images and researchers have found that women are less satisfied with their physical appearances after viewing ideal images in the media (Irving, 1990; Pinhas et al., 1999; Richins, 1991), when presumably those discrepancies are made salient. Additionally, U.S. women have become objectively heavier (Dortch, 1997). For example, among young adults aged 20 to 29, overweight rates for women increased steadily between the early 1960s and the early 1990s, while they remained stable for men (Dortch). Since cultural expectations for ideal female beauty conflict with an increase 173

Downloaded from ctr.sagepub.com at UNIV OF DELAWARE LIB on December 28, 2013

weight (Dortch; Garner, Garfinkel, Schwartz, & Thompson, 1980), it is not surprising that many women experience a discrepancy between their perceived actual and their internalized ideal physical attributes. Both men and women internalize ideals discrepant from the physical attributes they actually possess; however, the extent of the discrepancy seems to be greater in women than in average

in

men

(Fallon & Rozin, 1985; Jacobi & Cash, 1994). In

a

study by Monteath and McCabe (1997), college women recognized cultural ideals of thinness in western society as extreme; nevertheless, greater levels of body dissatisfaction associated with greater deviations from the cultural Self-discrepancy theory has been used to study and identify the kinds of discrepancies that individuals may possess in relation to body image distress and eating disorders (Altabe & Thompson, 1996; Strauman & Glenberg, 1994; Strauman et al., 1991; Szymanski & Cash, 1995). Altabe and Thompson (1996) found that discrepancy between actual and ideal appearance was positively related to body were

Although not originally proposed as part of the model, appearance self-schema may relate to the coping strategies used. People for whom appearance is central, salient, and influential (i.e., those who are self-schematic on appearance) may be likely to accept the cultural standard and either try harder to achieve it or quit trying. However, people for whom appearance is not as central, salient, and influential (i.e., those who are not schematic on appearance) may be unlikely to accept the cultural standard because they may not perceive it to be relevant. As a result, they may be likely to modify their personal standard or modify the cultural standard.

ideal.

image anxiety, body dissatisfaction, and depression. In general, people who possess discrepant actual and ideal selves are thought to be vulnerable to dejection-related emotions such as sadness or disappointment due to the inability to attain desired goals (Higgins, 1989). Thus, media images precipitate women’s appearance self-discrepancies and these discrepancies affect body image and mood distress.

Rudd and Lennon Model Rudd and Lennon (1994) integrated the sociocultural perspective with aesthetic theories from textiles and clothing (DeLong, 1987, 1998; Fiore, Kimle, & Moreno, 1996a, 1996b; Fiore, Moreno, & Kimle, 1996; Hillestad, 1980) to propose a model of body aesthetics (Figure 1) that suggests how individuals respond to cultural aesthetic ideals regarding bodies. According to their model, women (i.e., creators) actively create their appearances in accord with the cultural aesthetic ideal using appearance-enhancing products and behaviors. Although the cultural aesthetic ideal is narrowly defined, women can experience aesthetic pleasure in the process of approximating it and individuals can distinguish themselves from others on that basis. Created appearances are positively or negatively evaluated by others (e.g., peers, parents, significant others) in an appreciation process; those perceived evaluations in turn influence the appearancemanagement behaviors of the creators themselves. Positive assessments from self and others increase self-esteem, whereas negative assessments lead to a feedback loop in the model. Accordingly, if a created appearance is (perceived to be) close to the ideal appearance, assessments are positive, self-esteem is elevated, and a strong self-image can result. On the other hand, if a created appearance is (perceived to be) not close to the ideal appearance, assessments are negative, and individuals engage in one of four coping strategies. These are (a) accept the cultural standard and try harder to achieve it; (b) accept the cultural standard and quit trying to achieve it; (c) modify one’s personal standard; (d) modify the cultural standard (see Rudd & Lennon for a further discussion of the coping strategies). Internalization of others’ evaluations may also affect the coping strategies selected. Support for the Rudd and Lennon model has been reported (Lennon, Rudd, Sloan, & Kim, 1999).

Self-discrepancy theory was not originally proposed as part of the Rudd and Lennon model (1994). However, the extent to which one’s created appearance approximates the perceived ideal appearance can be construed as the discrep-

ancy between one’s actual appearance and ideal appearance. In their model, self-esteem as a function of that discrepancy was expected to motivate appearance-management behaviors. In so doing, appearances could be re-created in yet another attempt to approximate the cultural aesthetic ideal (i.e., to try to reduce the discrepancy). Since body image has been found to be positively related to self-esteem (Lennon & Rudd, 1994; Striegel-Moore, Silberstein, & Rodin, 1986), measures of body image may also be related to discrepancies between actual and ideal appearances according to the model. Thus, the discrepancy between actual and ideal appearance is likely to be related to aspects of body image such as self-appraisals of appearance, cognitive and behavioral investment in appearance, and body satisfaction or dissatisfaction. Mood and other types of affective outcomes were not originally included in the diagram of the Rudd and Lennon model (see Figure 1 for variables included in the model);. however, it is reasonable to extend that model to include them. There are several possibilities; mood and other affective outcomes may be outcomes of discrepancy assessments, of coping strategies, or of the appreciation process. Researchers have determined that self-esteem is negatively related to anxiety, unhappiness, and depression (Noles, Cash, & Winstead, 1985; Rosenberg, 1985). If self-esteem is at least partly a function of the discrepancy between actual and ideal appearance, as we have argued, then measures of mood (e.g., anxiety, dissatisfaction, or depression) may also be related to that discrepancy. Body image researchers have studied the discrepancy between actual and ideal appearance and found it to be related to anxiety, dissatisfaction, and depression (Altabe & Thompson, 1996). In other more general studies of body image, researchers (Heinberg & Thompson, 1995; Pinhas et al., 1999) have found that mood (depression, anger, dissatisfaction) is affected by exposure to fashion images, perhaps because the images triggered awareness of discrepancies between actual and ideal appearances. This review suggests that mood may be related to

self-discrepancy. Based on the literature reviewed, the following hypoth-

eses were

formulated.

HI Women more

high

in appearance

self-discrepancy

will be

distressed than those low in appearance self-dis-

crepancy. H2 Women high in appearance

174 Downloaded from ctr.sagepub.com at UNIV OF DELAWARE LIB on December 28, 2013

self-discrepancy

will be

;,

Figure

1. Rudd and Lennon model of body aesthetics

(1994,

dissatisfied with their bodies than those low in appearance self-discrepancy. H3 Women high in appearance self-discrepancy will be more dissatisfied with their overall appearance than those low in appearance self-discrepancy. H4 Women high in appearance self-discrepancy will exhibit lower self-esteem scores than those low in appearmore

ance

p.

165).

women do perceive discrepancies between their actual and ideal appearances. However, not all women develop eating disorders or other body image disturbances. This suggests to us that other individual variables may be operating such as appearance self-schema, which may lessen or heighten the effects of appearance self-discrepancies on body image. Therefore, we developed the following research

question:

self-discrepancy.

H5 Women high in appearance self-discrepancy will selfrate their appearances lower than those low in appear-

self-discrepancy. H6 Women high in appearance self-discrepancy will exhibit greater cognitive and behavioral investment in appearance than those low in appearance self-discrepancy. Finally, we were interested in the relationship between appearance self-schema and appearance discrepancy in explaining body image. Researchers have demonstrated that appearance self-discrepancy is correlated with appearance self-schema and body dissatisfaction (Szymanski & Cash,

RQ1 Which of the two (appearance self-schema or ap-

self-discrepancy) body image variables? pearance

ance

1995), as well as appearance orientation, appearance evaluation, and depression (Cash & Szymanski, 1995). It is possible that appearance self-schema and appearance discrepancy are so similar that they explain the same variance in body image. Because media depictions of ideal beauty are omnipresent in the United States, it is likely that most

would best

explain

Method A laboratory experiment was designed to test the hypotheses and examine the research question. Approval for the research was obtained from the university’s institutional review board (protocol number 98B0049). This research focused (a) on the relationship between appearance selfdiscrepancy and mood, body image, and self-esteem and (b) on the relative utility of appearance self-schema and appearance self-discrepancy in explaining body image. In =

175 Downloaded from ctr.sagepub.com at UNIV OF DELAWARE LIB on December 28, 2013

summed to obtain the measures of Joy and Distress. The Joy and Distress measures have reported internal consistencies of .80 and .90, respectively (Izard). Mood distress, body dissatisfaction, and overall appearance dissatisfaction. The five Visual Analogue Scales or VAS (Heinberg & Thompson, 1992; 1995) were used in this research as additional measures of mood and in part, as measures of body image such as body dissatisfaction and overall appearance dissatisfaction. Participants indicated their current level of anxiety, depression, anger, body dissatisfaction, and overall appearance dissatisfaction by placing a mark on a 100-mm line anchored by &dquo;no&dquo; and &dquo;extreme&dquo; (e.g., no anxiety, extreme anxiety or no depression, extreme depression). As a result, these measures were 100point scales and were scored so that higher scores indicated more distress. Heinberg and Thompson (1995) have reported convergent validity for the VAS. Social self-esteem. As a general measure of social selfesteem, we used the revised Janis-Field Self-esteem Scale (Eagly, 1967; Janis & Field, 1959). Participants rated how comfortable they are in 20 social situations using a 5-point Likert-type response format (1 = very often, 5 = never). This scale has split-half reliabilities of .72 and .88 (cited in Robinson & Shaver, 1973) and a test-retest reliability of .92 (Campbell, Chew & Scratchley, 1991). Half of the items are reverse-scored so that high self-esteem is indicated by

highlight appearance self-discrepancies to particiexposed them to idealized images of the cultural ideal of beauty. Development of stimulus materials. Because fashion images typically reflect the cultural ideal of beauty, photos of female models were used as stimuli. Color photographs (N= 40) of relatively unknown female fashion models taken from fashion magazines (e.g., Marie Claire, Glamour) were scanned to use as stimuli. The size of the figures was standardized and the backgrounds were eliminated. The 40 images were rated on attractiveness by textiles and clothing graduate students using a 7-point unipolar scale. Sixteen images with average ratings between 6 and 7 were selected for use. The 16 photos were composed of head-and-shoulders shots ( 10), order to

pants,

we

head-and-torso shots

(2), and head-to-toe shots (4).

Instruments

Appearance self-discrepancy. The Body-image Ideals Questionnaire or BIQ (Cash & Szymanski, 1995) was used to assess discrepancies between actual and ideal appearance with respect to 10 appearance characteristics (i.e., height, skin complexion, hair texture/thickness, facial features, muscle tone and definition, body proportions, weight, chest size, physical strength, and physical coordination). For each physical attribute, participants were asked to think about discrepancies between their actual and ideal attributes (how they wish or prefer to be). Respondents rated the discrep-

higher scores. Global self-esteem. Rosenberg’s (1965) Self-esteem Scale is known to be a reliable (.85) and valid (Silber & Tippett, 1965) instrument measuring global self-worth. This scale is made up of 10 items that were measured using a 5point Likert response format ( strongly disagree, 5 strongly agree). Five of the items are keyed in a positive direction and the other five are keyed in a negative direction. The items are scored such that higher scores indicate high self-esteem. Rosenberg’s measure has been used in other body image research (e.g., Akan & Grilo, 1995; Lennon &

ancy between actual and ideal for all 10 characteristics as follows: -1 (exactly as I am), +1 (almost as I am), +2 (fairly unlike me), and +3 (very unlike me). Respondents also rated the importance- of each characteristic as follows: 0 (not important), 1 (somewhat important), 2 (moderately important), and 3 (very important). Appearance self-discrepancy scores by imfor each characteristic and then summing the 10 weighted scores. Higher scores indicate greater discrepancy. Cash and Szymanski reported an internal consistency of .77 for appearance self-discrepancy. Appearance self-schema. The Appearance Schemas Inventory or ASI (Cash & Labarge, 1996) composed of 14 items was used to measure the cognitive importance, meaning, and influence of appearance in one’s life. Example items are &dquo;The only way I could ever like my looks would be to change what I look like&dquo; and &dquo;My appearance is responsible for much of what has happened to me in my life.&dquo; Items were rated on 5-point Likert scales (1 = strongly disagree, 5 = strongly agree). The ASI has an internal consistency of .79 (Cash, 1992) and a 1-month test-retest stability of .71 (Cash, 1992). Item scores were summed for an overall score. Positive and negative mood. The Joy and Distress subscales from the Differential Emotions Scale (Izard, 1972) were used to assess positive and negative mood of participants following exposure to attractive images. Participants were asked to indicate how they were feeling on each of the scales at the time of measurement using 5-point scales; the amount of affect could vary from 1 (very slightly or not at all) to 5 (very strongly). For the Joy scale, participants rated themselves using the adjectives happy, joyful, and delighted, while for the Distress scale, participants rated themselves using the adjectives discouraged, sad, and downhearted. For each subscale, ratings on the three component adjectives were was

calculated

portance

=

by multiplying discrepancy

scores

Rudd, 1994; Lennon .

et

=

al., 1999; Thomas, 1989).

Body image. In addition to the two measures of body image from the VAS (body dissatisfaction, overall appearance dissatisfaction), we also used two subscales of the Multidimensional Body-Self Relations Questionnaire (Brown, Cash, & Mikulka, 1990; Cash, 1990; Cash, Winstead, & Janda, 1986), which is a well-validated bodyimage measure containing items concerning appearance, fitness, and health. For this research we used the appearance evaluation and appearance orientation subscales which use a Likert format ( = strongly disagree, 5 = strongly agree). Appearance evaluation assesses self-evaluations of appearance using seven items. This subscale has an internal consistency of .88 and a stability of .91 (Cash & Szymanski, 1995). Appearance orientation assesses cognitive and behavioral investment in one’s appearance. This subscale contains twelve items and has an internal consistency of .85 and a stability of .90 (Cash & Szymanski).

Procedure A convenience sample of 102 college women from a major midwestern university volunteered to participate in a two-part laboratory experiment. In each session, tasks were completed in small groups (range = 5 to 10). Of this group, 88

176 Downloaded from ctr.sagepub.com at UNIV OF DELAWARE LIB on December 28, 2013

.



.

completed all measures in the questionnaire packet and their data were included in the analyses. As an incentive to participate, participants were entered into cash prize drawings. Participants were assessed twice with an interval of two weeks between sessions. The time interval between administration of the initial session and the follow-up session was necessary to ensure that any effects assessed in the followup session would not be due to priming effects from the initial session. In the initial session, participants completed an informed consent form and were asked (a) to complete the Appearance Schemas Inventory and then (b) to assess discrepancies between their actual and ideal appearances using the Body-image Ideals Questionnaire. Participants also provided information regarding age, major, ethnicity, height, and weight. In the follow-up experimental session, participants were given folders containing the 16 stimulus photos. To focus the participants on the stimulus photos, filler tasks were completed that required assessing the appearances of the women in the stimulus photos. Respondents were instructed (a) to identify the skin color, hair color, and eye color of models in the head-and-shoulder shots and in the head-andtorso shots and (b) to rate the fashionability and formality of the stimulus images in the head-to-toe shots. After completing these filler tasks, participants completed the measures of positive and negative mood (from the Differential Emotions Scale); the VAS measures (mood distress, body dissatisfaction, and overall appearance dissatisfaction); social self-esteem, global self-esteem, and the two subscales from the MBSRQ (appearance evaluation and appearance orientation). Finally, the women were debriefed and given the opportunity to ask questions.

Results

.

Reliabilities of instruments. All multi-item standardized measures were determined to have adequate to high reliabilities (internal consistencies) in our sample. All the reliabilities were determined using Cronbach’s alpha. The reliability for appearance self-discrepancy was a = .77. Reliabilities were high for the two self-esteem measures (a = .90 for social self-esteem and a = .90 for global selfesteem). Other reliabilities were a = .88 for the Joy scale, a = .82 for the Distress scale, a = .90 for appearance orientation, a = .85 for appearance evaluation, and a = .81 for appearance self-schema (Appearance Schemas Inventory). Characteristics of participants. Of the 88 women who completed all measures, 40 were textiles and clothing majors and 48 were from various other majors. The participants’ ages ranged from 18 to 28, with a mean of 21. Sixtysix participants were Caucasian Americans, nine were African Americans, twelve were Asian Americans, and one was Hispanic American. A body mass Index [weight

(kg)/height2 (m)] was calculated from self-reported heights 5’S&dquo;) and weights (M 132.5) of each participant.

(M

=

As

an

=

indicator of body fatness (Garrow & Webster, 1985), mass index has been used as an important variable

body in body image

research

(Cash, Grant, Shovlin, & Lewis,

1992; Haase & and in research

Prapavessis, 1998; Stice & Shaw, 1994) appearance self-discrepancy (Strauman

on

et al., 1991). Preliminary analyses. Preliminary analyses were conducted to determine if participants’ major or body mass index might be confounding variables in this study. There is reason to expect that participants from fashion-related majors might respond differently in research concerning dress and appearance than respondents (a) from other majors or (b) without an interest in fashion (Lennon, Bums, & Rowold, 1995). Thus, we wanted to see if major (textiles and clothing major or non-major) was related to assessed appearance discrepancy. Based on a median split of appearance discrepancy scores (Median =15), 44 respondents were assigned to the low appearance discrepancy group, and 44 were assigned to the high appearance discrepancy group. Using the chisquare statistic, major and appearance discrepancy group were found to be independent, x2 (1) = 0.73, ns. Analysis of variance using major as the independent variable and appearance discrepancy scores as the dependent variable revealed that the effect for major was non-significant, F (1, 86) = 0.27. Thus, major was unrelated to appearance discrepancy in our participants. Finally, the effects of major on the dependent variables in this study were assessed using multivariate analysis of variance; the overall effect for major on the dependent variables was non-significant, F (1, 76) = 0.96. Thus, major was unrelated to the dependent variables as well as to independent variable of interest. We were also interested in whether or not body mass index was related to appearance discrepancy in our sample. Using a simple regression analysis with body mass index as the independent variable and scores on appearance discrepancy as the dependent variable, we found that these two variables were unrelated, F (1, 86) = 0.80. Finally, we divided participants into three groups based on body mass index (Underweight, normal weight, and overweight). A body mass index under 20 was categorized as underweight, from 20 to 25 was categorized as normal weight (Garrow, 1986), and over 25 was categorized as overweight (Bray, 1986; Strauman et al., 1991). Using that categorization scheme, the effects of body mass index on the dependent variables in this study were assessed using multivariate analysis of variance; the overall effect for body mass index on the dependent variables was non-significant, F (11, 75) = 1.01. Thus, body mass index was unrelated to the dependent variables as well as to independent variable of interest. Thus, results from the preliminary analyses revealed that neither major nor body mass index was related to the variables of interest in this study.

One-way Multivariate and Univariate Analyses of Variance The data were analyzed using a one-way between subjects multivariate analysis of variance with appearance selfdiscrepancy as the independent variable and the 11 measures of mood, self-esteem, and body image as the dependent variables. There was a significant multivariate main effect for appearance self-discrepancy on the dependent variables, F (11, 76) 3.35, p < .01. Univariate analyses of variance were used to determine which of the dependent variables were affected by appearance self-discrepancy. =

177 Downloaded from ctr.sagepub.com at UNIV OF DELAWARE LIB on December 28, 2013

Table 1. Univariate

Note: BD AE

Analyses

of Variance for

Appearance Self-Discrepancy on the Dependent Variable

Body Dissatisfaction; OAD Overall Appearance Dissatisfaction; SSE Social Self-Esteem; Appearance Evaluation; AO Appearance Orientation. ’Eta’ represents the variance in the dependent variable accounted for by the independent variable. *p