Essay

10.1177/0093650205279350COMMUNICATION RESEARCH • October 2005Park • Presumed Media Influence SUNG-YEON PARK

The Influence of Presumed Media Influence on Women’s Desire to Be Thin

This study investigated the effect of magazine use on the desire to be thin within the theoretical framework of presumed influence. Structural equation modeling supported the hypothesis that reading beauty and fashion maga- zines increased the drive for thinness both directly and indirectly. The indirect pathway included the perceived prevalence of the thin ideal in mass media, the presumed influence of the thin ideal on others, and the perceived influence of the thin ideal on self. Social pressure to be thin may be based both on reality and the presumption of influence on others. Results suggest potential strategies for intervention.

Keywords: body image; eating disorders; presumed media influence

Body discontent and problematic attitudes toward eating are facets of a sin- gle issue that has gained increasing attention in our society during recent decades. As many as 10 million women and 1 million men are estimated to suffer from anorexia nervosa, and an additional 25 million people are affected by bulimia nervosa (Shisslak, Crago, & Estes, 1995). Prevalence among young women, who have traditionally been considered at the highest risk for these disorders, is estimated at 15% (Levchuck, Kosek, & Drohan, 2000). Body image disturbances also appear to start at a young age. In fact, in one study, 40% of 6-year-old girls reported wishing that they were thinner. And more often than not, these negative body images were associated with strict dieting, a potential precursor to serious eating disorders (Muller, 1998).

As worrisome as these statistics are, there have also been continuous efforts to investigate the causes of body image disturbances and eating disor- ders. Specific biological traits, psychological characteristics, and family dynamics have all been implicated as contributory factors to eating disorders

594

COMMUNICATION RESEARCH, Vol. 32 No. 5, October 2005 594-614 DOI: 10.1177/0093650205279350 © 2005 Sage Publications

(Heinberg, 1996; White, 1992). However, studies using various methods, including the analyses of patient records (Garner & Garfinkel, 1980; Striegel- Moore, McAvay, & Rodin, 1986), historical analyses (Casper, 1983), and cross- cultural comparisons (Gunewardene, Huon, & Zheng, 2001; McCarthy, 1990; Nassar, 1988), have consistently identified the sociocultural emphasis on thinness as the likely primary cause of the development of these disorders.

In response to these clinical findings and the growing public suspicion that mass media are a main contributor to this problem, social scientists have investigated the connection between thin, idealized images and eating disor- der symptomatology (e.g., Bissell & Zhou, 2004; Botta, 1999, 2000; Cusumano & Thompson, 1997; Grogan, Williams, & Conner, 1996; Harrison, 1997, 2000a, 2000b; Harrison & Cantor, 1997; Harrison & Fredrickson, 2003; Heinberg & Thompson, 1995; Irving, 1990; Levine, Smolak, & Hayden, 1994; Myers & Biocca, 1992; Posavac, Posavac, & Posavac, 1998; Stice, Schupak- Neuberg, Shaw, & Stein, 1994; Stice & Shaw, 1994; J. K. Thompson & Heinberg, 1999; Thomsen, 2002). Previous studies employing diverse theo- ries and methods have increased our understanding of the effects of mass media on a wide range of relevant psychological responses including negative affect, weight concerns, distorted perceptions of body parts, and self-esteem and on more comprehensive measures of eating disorder symptoms. At the same time, however, these studies have used a limited definition of media effects in which individuals react to media messages in social isolation. Indeed these effects appear to be primarily of an individual nature. Yet many scholars insist that the social context of media use should also be integrated into research (Bishop, 2000; Stice & Shaw, 1994).

Social influence to conform to a thin ideal can originate from a wide range of relationships ranging from family, friends, and peers to society as a whole (Levine et al., 1994; Taylor et al., 1998). Within social contexts, pressure to be thin can be explicitly communicated. Alternatively, this pressure can also merely be inferred, regardless of its actual existence. This investigation aims to explore a previously neglected part of this process by examining the influ- ence of media on women through their perceptions of media effects on others.

The Effects of Presumed Media Influence

In the field of mass communication, the concept of presumed media influence on others is hardly new. In fact, early in the history of modern mass media, people were aware that the projected opinions of others could influence an individual’s decision-making process. For example, the bandwagon effect, a long-standing strategy of propagandists, operates by convincing people that

595

Park • Presumed Media Influence

many important others are adopting a certain position and that therefore they must follow (Lee & Lee, 1939). Several decades later, Noelle-Neumann (1974) proposed a spiral of silence theory, which states that one’s perception of public opinion, as reflected in mass media, will influence the likelihood that he or she will express his or her opinion.

The third-person effect is based on a similar perspective regarding public opinion—in this case that people tend to perceive greater influence of mass media on others than on themselves. Recently, a generalized version of the third-person effect was proposed. Founded on an extensive line of research on this effect, Gunther and Storey (2003) developed a more comprehensive model of indirect media effects. This model, termed the influence of presumed influence, is based on the idea that people will perceive some influence of communication on others and as a result will change their own attitudes or behaviors accordingly.

The more inclusive nature of this model is illustrated by its two fundamen- tal departures from the original conceptualization of the third-person effect. First, whereas the third-person effect requires the a priori identification of the desirability of perceived influence to make any theoretical predictions, the new model omits this restriction. Messages with either positive or nega- tive presumed influence can be used, and their effects on either intended or unintended audiences can be examined, depending on the researcher’s inter- ests. Second, because Gunther and Storey (2003) declared that the perceived effect on self is not an integral part of this model, a comparison between the perceived effects on self and others is no longer an essential component of theory testing.

Gunther and Storey (2003) tested this broader model of indirect effects in the context of a Nepalese radio campaign aimed at health workers in clinics. This campaign used a drama format and was intended to improve the inter- personal communication, counseling skills, and technical knowledge of health workers. The model was tested by assessing the effects of the cam- paign on unintended audiences—the general Nepalese population—via their perceptions of the influence of the radio program on clinic health workers. In general, results supported the indirect effects model. Members of the public exposed to the radio campaign perceived that health workers would be influ- enced by the campaign, and in turn this presumed influence produced a more positive image of health workers that improved the interactions between the public and health workers. As expected, inclusion of the direct path linking public exposure to the radio campaign and health worker image had little influence on the magnitude of this indirect effect.

596

COMMUNICATION RESEARCH • October 2005

Media, Others, and Me: The Triangle of Influence

Exploration of the relationship between disturbed body image, which often results in disordered eating, and mass media began with the examination of circumstantial evidence. Content analyses revealed that the size of women appearing in virtually all media outlets has decreased continuously since the 1960s. And this tendency became more prominent after the 1970s, when the number of reported cases of eating disorders exploded (Garner, Garfinkel, Schwartz, & Thompson, 1980; Kaufman, 1980; Mazur, 1986; Silverstein, Perdue, Peterson, & Kelly, 1986).

Content analyses were followed by surveys and experiments that exam- ined the relationship between mass media exposure and disordered self- perceptions and body perceptions. Since the early 1990s, a substantial amount of research has been conducted on this topic. Experimental studies have typically involved the manipulation of exposure to media stimuli and subsequent measurements of self-perceptions and body perceptions and atti- tudes toward eating. Surveys have found associations between participants’ media use and similar outcomes. Using this basic model, researchers have found that exposure to the thin ideal increases the risk of body image distur- bances and eating disorders, as measured by self-esteem, weight satisfaction (Irving, 1990), body esteem (Grogan et al., 1996), body dissatisfaction, drive for thinness, attitude toward eating (Harrison & Cantor, 1997), concerns about body shape (Thomsen, 2002), self-objectification (Harrison & Fredrickson, 2003), and symptoms of anorexia and bulimia (Bissell & Zhou, 2004).

These studies established a relationship between media use and various eating disorder symptoms, but there is also extensive psychological and edu- cational research emphasizing the importance of other people in shaping women’s body images and eating attitudes. Family members, especially par- ents, may be an early and influential source of pressure to be thin. Parents exceedingly conscious of their own weight are sometimes similarly concerned with keeping their daughters in “perfect” shape (Byely, Archibald, Graber, & Brooks-Gunn, 2000; Vincent & McCabe, 2000). Also, a mother’s concern with thinness is a significant predictor of weight concerns for girls in Grades 4 to 8 (Levine et al., 1994; Taylor et al., 1998).

At the same time, the influence of peers may be equivalent to or stronger than that of family members. For example, in one study the combined influ- ence of family and peers was the strongest determinant of bulimic tendencies in high school girls with various personality and social risk factors (Pike, 1995). And during college years, the influence of family members seems to

597

Park • Presumed Media Influence

subside, whereas the influence of peers remains steady (Striegel-Moore et al., 1986).

If we combine the results from research on media effects and family and peer influence, we can be reasonably certain that both mass media and the people in one’s immediate social environment are crucial elements in individ- uals’ body images and attitudes toward eating. Also, because many previous studies have shown that mass media exposure can be nearly ubiquitously harmful to young women, we can safely assume that the actual influence of mass media on individuals and their peers is reasonably consistent. What is missing here is a body of research investigating how women are affected by their perceptions of how others are influenced by the media.

Still, three studies shed some light on women’s perceptions of the influ- ence of the thin ideal on others. In one study that examined the third-person effect, female college students estimated that other women would be more influenced by idealized images than they expected themselves to be. When self-esteem was introduced as an individual characteristic mediating the third-person perception, the difference in perceived influence was more pro- nounced in the high self-esteem group than in the low self-esteem group. Compared with their low self-esteem counterparts, people with high self- esteem reported lower levels of perceived influence on both themselves and others because of exposure to thin images (David & Johnson, 1998).

Other research, in which in-depth interviews were conducted with high school girls, found that presumed media influence can also have deleterious effects. In this study, although most participants seemed capable of criticiz- ing images in teen magazines for their unrealistic beauty and thinness, they also believed that their criticism was not widely shared by other women and men. Consequently, they still engaged in social comparison processes that made them feel abnormal and inferior relative to media images (Milkie, 1999). Results of a recent survey appear to illustrate this process more clearly. In this study, magazine reading increased college women’s beliefs that men expect women to be thin, which in turn increased their level of con- cern about body shape (Thomsen, 2002).

Taken together, these studies support the notion that mass media affects women’s attitudes toward their bodies and toward eating via presumptions about mass media’s influence on other women and men. In addition, Milkie (1999) described the nature of presumed media influence as unrelenting. In her study, irrespective of whether or not an individual personally accepted the thin ideal, everyone still felt pressured to conform to it if they perceived the thin ideal to be a norm accepted by others.

598

COMMUNICATION RESEARCH • October 2005

Construction of a Path Model

Based on this review, a model of relevant media effects was proposed. The first independent variable chosen was an objective measure of mass media use. Beauty and fashion magazine reading was selected from among the many possible types of media use because such exposure appears to be a pri- mary source of idealized thin images and a cause of unhealthy body images and attitudes toward eating (Harrison & Cantor, 1997; Thomsen, McCoy, Gustafson, & Williams, 2002).

Women’s desire to be thin was selected as the outcome variable. A variety of measures of eating disorder symptomatology have been used in previous research, no doubt because the issue of thinness is intertwined with other appearance-related issues, gender issues, and even broader social issues. Such measures have included a wide range of attitudes toward one’s body, one’s appearance, and one’s self as a whole. However, to minimize conceptual complications, this investigation focused solely on the desire to be thin.

In addition to a proposed direct path between magazine use and the desire to be thin, this research model also integrated pathways for indirect influ- ence. In a series of studies investigating the persuasive press inference, Gun- ther (1998; Gunther & Christen, 1999, 2002) proposed a presumed influence process comprised of a series of steps. These steps included (a) attention to mass media and the formation of impressions about its content, (b) an assumption of representativeness for that particular content (extrapolation), (c) an assumption that such content has broad reach, and (d) the presumption that the content will influence the opinions and attitudes of others (Gunther & Storey, 2003). Because the model is still in the conceptual stage and has not been empirically tested, only the second and fourth steps—the perceived prevalence of the thin ideal in mass media and the presumed influence of the thin ideal on others—were adopted as potential intervening variables. Also, although previous studies have not examined whether or not presumed influ- ence on other women and presumed influence on men produce different effects, a distinction between the two was considered to be a more careful and appropriate approach to this gender-specific problem.

Perceived influence on self was also included as a third intervening vari- able. The previous finding that women felt pressure to conform to the thin ideal because of its presumed influence on others, even when they resisted the media influence firsthand (Milkie, 1999), suggests the need for a clear dis- tinction between the perceived influence on self and the presumed influence on others. The final model is shown in Figure 1.

599

Park • Presumed Media Influence

Hypotheses

Because the model was comprised of one independent variable, one depend- ent variable, and three intervening variables, the effect process was exam- ined in four stages that delineated the hypothesized indirect path: Reading beauty and fashion magazines affects the perceived prevalence of the thin ideal, which affects the presumed influence on others. The level of presumed influence on others in turn affects the presumed influence on self, which affects the desire to be thin. Finally, the full model, in which direct and indi- rect paths were integrated, was examined.

Stage 1: The Effect of Media Use on Perceived Prevalence of the Thin Ideal

Repeated exposure to a certain phenomenon or behavior in the media has been shown to increase the perceived prevalence of these occurrences in both mass media and reality (Gerbner, Gross, Morgan, & Signorielli, 1994). There- fore, reading beauty and fashion magazines is expected to increase the per- ceived prevalence of the thin ideal in mass media.

Hypothesis 1: Reading beauty and fashion magazines increases the per- ceived prevalence of the thin ideal in the mass media.

600

COMMUNICATION RESEARCH • October 2005

Presumed influence on others

Perceived Prevalence of the thin ideal

Perceived influence on

self

Beauty & Fashion

magazine use Desire to be thin

Figure 1. Hypothetical Paths Illustrating the Direct Effect of Magazine Use (Dotted Line) and the Influence of Presumed Media Influence (Solid Lines)

Stage 2: The Effect of Perceived Prevalence of the Thin Ideal on Presumed Influence on Others

Assuming a broad reach for media content will increase the degree of pre- sumed influence of that content on the opinions and attitudes of others (Gun- ther, 1998; Gunther & Christen, 1999, 2002).

Hypothesis 2: The perceived prevalence of the thin ideal increases the pre- sumed influence of the thin ideal on others.

Stage 3: The Effect of Presumed Influence on Others on Perceived Influence on Self

David and Johnson (1998) reported that, compared with their low self-esteem counterparts, female college students with a high amount of self-esteem esti- mated the effects of thin images to be considerably lower on both themselves and other women. This finding suggests that women who estimate a strong influence of the thin ideal on others may also feel that they are heavily influ- enced, whereas women who estimate a weak influence of the thin ideal on others estimate the influence on themselves to be relatively small.

Hypothesis 3: The presumed influence of the thin ideal on others increases the perceived influence of the thin ideal on self.

Stage 4: The Effect of Perceived Influence on Self on Desire to be Thin

The psychological process that includes the perceived prevalence of the thin ideal, the presumed influence of the thin ideal on others, and the perceived in- fluence on self is likely to ultimately affect the desire to be thin.

Hypothesis 4: The perceived influence of the thin ideal on self increases the desire to be thin.

Full Model

The indirect effect of magazine use on the desire to be thin via presumed me- dia influence is expected to exist independently from the direct effect of mag- azine use on the desire to be thin.

601

Park • Presumed Media Influence

Hypothesis 5: Reading beauty and fashion magazines increases the desire to be thin directly and indirectly via the presumed influence on others.

Method

Sample

Many previous studies investigating the relationship between exposure to thin media images and eating disorders examined female college students, a group of people with an unusually high risk for developing body image distur- bances and eating disorders (Levchuck et al., 2000; Mazzeo & Espelage, 2002).

In this study, a total of 553 women were recruited to participate from com- munication courses at a major midwestern university. Among them, 432 (78%) completed their participation. Participants were recruited on a volun- tary basis and received extra credit for their participation. Research proce- dures closely followed the guidelines prescribed by the university’s Human Subjects Committee.

Procedure

The data-collection process was primarily managed through the Internet. Researchers are increasingly using the Internet as a data-collection tool, and it appears to have many advantages over more traditional methods (Yun & Trumbo, 2000).

The majority of participants were recruited in their classrooms. At that time they received a consent form notifying them of the personal nature of some of the questions and the fact that they could withdraw their consent at any time. After recruitment was complete, students who had agreed to partic- ipate received an e-mail containing the URL of the survey Web site. Students were allowed 3 days in which to complete the survey in a single sitting at a convenient time and place. On average, participation took approximately 30 minutes. Data for the entire sample were collected during a 10-day period.

Variables

All variables, with the exception of magazine use, were measured by 7-point Likert-type scales. The direction of these scales was varied to prevent habituation.

602

COMMUNICATION RESEARCH • October 2005

Beauty and fashion magazine use. For this variable, students were asked to report the number of issues of beauty and fashion magazines they read each month. Vogue and Cosmopolitan were provided as examples.

Perceived prevalence of the thin ideal in the mass media. Participants also estimated the prevalence of the thin ideal within two mass media channels, television and magazines: “How often do you see very thin models and actresses on television programs and commercials (in magazines)?” Answers were measured on a 7-point scale from 1 (never) to 7 (all the time). These two responses were highly correlated (r = .70, p < .001).

Presumed influence of the thin ideal on other women. This variable was measured by two questions, one assessing the direction of influence and the other estimating the intensity of that influence. Direction was measured by asking: “Do you think the images make other women in general prefer a thin- ner body type?” Responses were assessed on a 7-point scale from 1 (not at all) to 7 (very much so). Intensity was measured by asking: “How much influence do you think the images of very thin models and actresses in the mass media have on other women’s perceptions of the female body ideal?” Responses were captured on a 7-point scale from 1 (no influence at all) to 7 (a lot of influence). Again, answers to these two questions were highly correlated (r = .73, p < .001).

Presumed influence of the thin ideal on men. Presumed influence on men was measured in the same manner as presumed influence on other women but the word men was substituted for the phrase other women in each ques- tion. The correlation coefficient between these items was similarly high (r = .71, p < .001).

Perceived influence of the thin ideal on self. Again, the questions measur- ing the presumed influence of the thin ideal on other women were modified to measure the influence of the thin ideal on self. However, in this case the phrase other women was replaced with the word you. Intensity and direction were again highly correlated (r = .71, p < .001).

Desire to be thin. For this variable, students were asked to circle a number between 1 (not at all) and 7 (very much) that best reflected their level of agree- ment with each of three statements, including, “I am strongly motivated to be

603

Park • Presumed Media Influence

thinner”; “My determination to lose weight is strong”; and “I am driven to diet.” Principal component factor analysis on these three items yielded a sin- gle factor. The reliability for this index was good (α = .87).

Results

Two structural equation models that included either presumed influence on other women or presumed influence on men were constructed using LISREL 8.3. Four goodness-of-fit indices were computed for each model. The model including presumed influence on other women produced reasonably good fit indices (χ2 = 59.61, df = 28, p < .001; RMSEA = .051; TLI = .98; CFI = .98). And the fit for the model that included presumed influence on men was similarly sound (χ2 = 70.55, df = 28, p < .001; RMSEA = .059; TLI = .96; CFI = .98). Therefore, the unstandardized coefficients obtained from the maximum- likelihood estimation procedure were utilized to test hypotheses. Table 1 shows the significance of individual path coefficients.

In the first stage, Hypothesis 1 predicted that reading beauty and fashion magazines would increase the perceived prevalence of the thin ideal in mass media. Indeed, the path coefficient was positive and statistically significant (β = .14, p < .001). Therefore, Hypothesis 1 was supported.

In the next stage, Hypothesis 2 stated that the perceived prevalence of the thin ideal would increase the presumed influence of the thin ideal on others. The results were mixed. The path coefficient from perceived prevalence to the presumed influence on other women was positive and statistically significant

604

COMMUNICATION RESEARCH • October 2005

Table 1 Unstandardized LISREL Coefficient Estimates and Their Significance Tests

Coefficient SE z Statistic

Model I Magazine use → perceived prevalence .14 .04 3.34*** Perceived prevalence → (PI) on other women .23 .06 3.94*** PI on other women → PI on self .46 .06 7.60*** PI on self → desire to be thin .56 .07 7.87***

Model II Magazine use → perceived prevalence .14 .04 3.30*** Perceived prevalence → PI on men .09 .06 1.37# PI on men → PI on self .29 .06 4.69*** PI on self → desire to be thin .38 .06 6.31***

#p < .10. ***p < .001. PI = presumed influence.

(β = .23, p < .001). However, the path from perceived prevalence to the pre- sumed influence on men approached but did not meet the accepted standard for statistical significance (β = .09, p < .1). As a result, Hypothesis 2 was sup- ported for presumed influence on other women but only marginally support- ed for presumed influence on men.

In the third stage, Hypothesis 3 contended that the presumed influence of the thin ideal on others would increase the perceived influence of the thin ideal on self. In both models, the presumed influence on others was highly cor- related with the perceived influence on self (for other women: β = .46, p < .001; for men: β = .29, p < .001). Given these findings, Hypothesis 3 was fully supported.

In the final stage, Hypothesis 4 predicted that the perceived influence of the thin ideal on self would increase the desire to be thin. And as expected, the perceived influence on self was very strongly related to the desire to be thin (β = .52 and .38 for the first and second models, respectively; p < .001 for each). Hypothesis 4 was supported by the data.

Hypothesis 5, which proposed significant direct and indirect effects of reading beauty and fashion magazines on the desire to be thin, was also deci- sively supported. Figure 2 shows the structural equation model that includes the presumed influence on other women. In this model, a direct effect of mag- azine reading on the desire to be thin, the predicted indirect effect path con- necting all five variables in the predetermined order, and three additional indirect paths were significant. As expected, magazine reading predicted the desire to be thin directly (β = .09). And at the same time, an equivalent effect on the desire to be thin was produced by the combined indirect paths leading to this outcome (β = .09). The model fit appeared to deteriorate substantially when any one of the predicted indirect paths was deleted. The full model explained 24% of variance in this outcome.

The second model, which included the variable for presumed influence on men, produced similar findings. Figure 3 shows the results for this analysis. A direct path from magazine reading to the desire to be thin, the predicted indi- rect path, and two additional indirect paths were significant. In this model, though, the direct effect of magazine reading on the desire to be thin appeared to be somewhat larger (β = .12) than the combined indirect effect from all sig- nificant presumed influence pathways (β = .09). Again, model fit was optimal when all of the indirect paths were included. This model accounted for 17% of variance in the desire to be thin. Therefore, in support of Hypothesis 5, read- ing beauty and fashion magazines appeared to increase the desire to be thin directly and indirectly via the presumed influence on others.

605

Park • Presumed Media Influence

Discussion

This research found that the use of beauty and fashion magazines increases the desire to be thin among female college students not only directly but also indirectly via the presumed influence on others. The data supported Hypoth- eses 1 to 4, the building blocks of the presumed influence model. Hypothesis 5 was also supported, confirming the presence of independent direct and indi- rect paths to this outcome.

According to these results, female college students who read beauty and fashion magazines perceived that idealized thin images were prevalent in mass media. The perception of a prevalent thin ideal in mass media in turn created an assumption that other women and men are influenced by this prevalent image and prefer thin body types as a result. This presumed

606

COMMUNICATION RESEARCH • October 2005

Presumed influence on other

women (.05)

Perceived prevalence of

the thin ideal (.03)

Perceived influence on

self (.28)

Desire to be thin (.24)

Beauty & Fashion

magazine use

Magazine

TV

Intensity

Direction

Intensity

Direction

Scale 1

Scale 2

Scale 3

1

1.15

1 1.19

1

1

1

1.06

.75

.18***

.23***

.16**

.43***

-.27*** .52*** .15***

.09*

Figure 2. Structural Model Showing the Direct and Indirect Effects of Magazine Use on the Desire to be Thin

Note: The indirect path is via perceived prevalence, presumed influence on other women, and per- ceived influence on self. Parameters are standardized coefficients, and R2 values are reported in parentheses. Rectangles represent observed variables, and ovals symbolize latent variables. *p < .05. **p < .01. ***p < .001.

influence on others then reinforced the influence of the thin ideal on the female students themselves, adding to the pressure to conform to the norm of ideal thinness. Finally, the perceived media influence on self enhanced the desire to be thin in these women. And this indirect effect persisted even after controlling for the direct effect of magazine use on this outcome.

Within the broader context of research on media effects, the identification of this indirect pathway has two implications. First, it increases our ability to explicate the key processes behind media effects. In these data, direct and indirect paths accounted for comparable amounts of unique variance. The second implication lies at the junction of interpersonal and mass communica- tion. Although this study is primarily concerned with mass media effects, the presumed influence of mass media on other people could also depend heavily on interpersonal communication. What is omitted from the current study is the role of others’ actual remarks related to the particular media effect at

607

Park • Presumed Media Influence

Presumed influence on men

(.01)

Perceived prevalence of

the thin ideal (.04)

Perceived influence on

self (.19)

Desire to be thin (.17)

Beauty & Fashion

magazine use

Magazine

TV

Intensity Direction

Intensity

Direction

Scale 1

Scale 2

Scale 3

1

1.34

1 1.03

.88

1

1

1.06

.75

.19***

.08#

.19***

.27***

.38***

.19***

.12**

Figure 3. Structural Model Showing the Direct and Indirect Effects of Magazine Use on the Desire to be Thin

Note: The indirect path is via perceived prevalence, presumed influence on men, and perceived influence on self. Parameters are standardized coefficients, and R2 values are reported in parenthe- ses. Rectangles represent observed variables, and ovals symbolize latent variables. *p < .10. **p < .01. ***p < .001.

issue. For example, some women in this study may have heard friends wish aloud that they were as thin as the models and actresses in magazines or on television. Such remarks may have served as the basis for the presumed influence of mass media on other women. Perceived similarities between one- self and others, actual remarks by others, and other aspects of interpersonal communication could also conceivably amplify or diminish the presumed level of media influence on others.

In the narrower context of presumed media influence research, this study provides another piece of supporting evidence for this emerging research par- adigm. In addition, these results validate the model of presumed media influ- ence previously outlined by Gunther (1998). The presumed influence path connecting beauty and fashion magazine use, the perceived prevalence of the thin ideal in the mass media, and the presumed influence on others were sup- ported by the analyses and were parallel to the original framework of pre- sumed media influence. In particular, including the perceived prevalence of the thin ideal in mass media as an intervening variable between magazine use and the presumed influence on others was instrumental in this investiga- tion. In Figure 2, magazine use was not directly related to the presumed influence on other women. However, it was associated with this variable via the perceived prevalence of the thin ideal in media. In other words, magazine reading increased the presumed influence of the thin ideal on other women only through an increased perception that the thin ideal is rampant in our— and therefore other women’s—media environment. Similarly, in Figure 3, magazine use did not have a direct effect on the presumed influence on men. However, the association between these variables was revealed when per- ceived prevalence was included as a mediator. Based on these results, the assumed representativeness of media content should not be overlooked in future research on presumed media influence.

These findings should also be discussed in the context of media effects on body image disturbances and eating disorders. In this regard, results provide a replication of previous reports of a significant association between media use and eating disorder symptomatology, a relationship that has sometimes been elusive.

One relatively unique contribution of this study is the finding of a pre- sumed influence pathway for this outcome that appears to be separate from any direct influence of mass media. As expected, female college students received pressure to be thin directly from the media. But women also appeared to create additional pressure on themselves by assuming that oth- ers were exposed to similar images and preferred a thinner body type as a result. However, the lack of a direct positive relationship between the

608

COMMUNICATION RESEARCH • October 2005

presumed influence on others and the desire to be thin also suggests that the assumptions about other people did not increase the desire to be thin immedi- ately. Instead, it did so by bolstering the media effect on self. This finding is consistent with Milkie’s (1999) observation that teenage girls received strong pressure to be thin by assuming media influence on male and female peers. In the first model, nearly half of the perceived influence on self resulted from the presumed influence on other women (Figure 2), and in the second model, a quarter of the perceived influence on self came from the presumed influence on men (Figure 3).

A negative and statistically significant path from the presumed influence on other women to the desire to be thin was an unexpected but interesting finding. All in all, the presumed influence on other women decreased the desire to be thin directly while increasing the same dependent variable via its influence on the perceived influence on self. How is this possible?

To solve this puzzle, the relationships among the presumed influence on other women, the perceived influence on self, and the desire to be thin were examined more closely. A subgroup analysis that controlled for perceived influence on self revealed a more accurate picture. Among those women whose perceived influence on self was low, the presumed influence on other women was negatively correlated with the desire to be thin. However, for those women who reported a relatively high level of perceived influence on self, there was no relationship between the presumed influence on other women and the desire to be thin. In other words, for women who were rela- tively free from the pressure to emulate thin media images, knowing that other women were getting the same pressure had mixed effects. But for women who were already under heavy pressure, knowing that other women also suffered the pressure exerted only negative effects.

Although the two models were consistent with theory and produced acceptable goodness-of-fit indices, this correlational data set cannot speak for the direction of the influence specified by the paths. In particular, the causal order for the path between the presumed influence of the thin ideal on others and the perceived influence of the thin ideal on self could be contested. Con- trary to the direction of influence specified by the model, the perceived influ- ence on self may have instead affected the level of presumed influence on oth- ers. Consequently, experiments and longitudinal studies are needed to clarify the time order among variables.

The overall findings of this study indirectly support the feminist argu- ment that the social pressure to be thin, produced largely by the mass media industry, turns a woman into another woman’s enemy. Presumably, as a result of this competition, women waste personal resources that could have

609

Park • Presumed Media Influence

been spared for self-development (Kilbourne, 1995). But at the same time, these results could also potentially be used to inform future interventions targeting body image disturbances and eating disorders.

This study demonstrated that a presumption of influence on others appears to account for a significant proportion of the pressure that young women experience when exposed to thin images. But, many other studies are filled with the voices of individuals decrying the media images (Kates & Shaw-Garlock, 1999; Milkie, 1999; C. J. Thompson & Haytko, 1997). Taken together, it’s possible that a substantial portion of the presumed influence on other women may be presumed rather than real. And because women tend to overestimate the effect of the thin ideal on other women (David & Johnson, 1998), they might also be likely to overestimate the effect on men. Although it has not yet been established, inflated perception about the effect of the thin ideal on men may have contributed to women’s overestimation of men’s pref- erences for thin body types (Cohn et al., 1987; Jacoby & Cash, 1994). Still, it should be recognized that the presumption of influence on others is likely to be exaggerated rather than entirely invalid. Although in some previous reports women have denied any influence of mass media on themselves (Milkie, 1999), in this study the positive direct path from magazine use to the perceived influence on self was significant and independent of any competing indirect path via presumed influence.

Based on these findings, intervention efforts should start from the under- standing that the social pressure to be thin is likely to be based partly on real- ity but also partly on projection. To combat the real, external social pressure to be thin, we should identify and expunge the sources of information that encourage comparison and competition among women. However, to combat the projected social pressure to be thin, we might instead facilitate the exchange of opinions about the issue among members of our society. In this way, people could become more aware of the fact that criticisms of these images are widely shared.

References

Bishop, R. (2000). More than meets the eye: An exploration of literature related to the mass media’s role in encouraging changes in body image. Communication Yearbook, 23, 271-303.

Bissell, K. L., & Zhou, P. (2004). Must-see TV or ESPN: Entertainment and sports media exposure and body-image distortion in college women. Jour- nal of Communication, 54(1), 5-21.

Botta, R. A. (1999). Television images and adolescent girls’ body image distur- bance. Journal of Communication, 49(2), 22-41.

610

COMMUNICATION RESEARCH • October 2005

Botta, R. A. (2000). The mirror of television: A comparison of Black and White adolescents’ body image. Journal of Communication, 50(3), 144-159.

Byely, L., Archibald, A. B., Graber, J., & Brooks-Gunn, J. (2000). A prospective study of familial and social influences on girls’ body image and dieting. International Journal of Eating Disorders, 28(2), 155-173.

Casper, R. C. (1983). On the emergence of bulimia nervosa as a syndrome: A historical view. International Journal of Eating Disorders, 2(3), 3-16.

Cohn, L. D., Adler, N. E., Irwin, C. E., Millstein, S. G., Kegeles, S. M., & Stone, G. (1987). Body-figure preferences in male and female adolescents. Jour- nal of Abnormal Psychology, 96(3), 276-279.

Cusumano, D. L., & Thompson, J. K. (1997). Body image and body shape ide- als in magazines: Exposure, awareness, and internalization. Sex Roles, 37, 701-721.

David, P., & Johnson, M. A. (1998). The role of self in third-person effects about body image. Journal of Communication, 48(4), 37-58.

Garner, D. M., & Garfinkel, P. E. (1980). Socio-cultural factors in the develop- ment of anorexia nervosa. Psychological Medicine, 10, 647-656.

Garner, D. M., Garfinkel, P. E., Schwartz, D., & Thompson, M. (1980). Cultural expectation of thinness in women. Psychological Reports, 47, 483-491.

Gerbner, G., Gross, L., Morgan, M., & Signorielli, N. (1994). Growing up with television: The cultivation perspective. In J. Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and research (pp. 17-41). Hillsdale, NJ: Lawrence Erlbaum.

Grogan, S., Williams, Z., & Conner, M. (1996). The effects of viewing same- gender photographic models on body-esteem. Psychology of Women Quar- terly, 20, 569-575.

Gunewardene, A., Huon, G. F., & Zheng, R. (2001). Exposure to Western- ization and dieting: A cross-cultural study. International Journal of Eat- ing Disorders, 29, 289-293.

Gunther, A. C. (1998). The persuasive press inference: Effects of mass media on perceived public opinion. Communication Research, 25, 486-504.

Gunther, A. C., & Christen, C. T. (1999). Effects of news slant and base rate information on public opinion inferences. Journalism and Mass Commu- nication Quarterly, 76, 277-292.

Gunther, A. C., & Christen, C. T. (2002). Projection or persuasive press? Con- trary effects of person al opinion and perceived news coverage on esti- mates of public opinion. Journal of Communication, 52(1), 177-192.

Gunther, A. C., & Storey, J. D. (2003). The influence of presumed influence. Journal of Communication, 53(2), 199-224.

Harrison, K. (1997). Does interpersonal attraction to thin media personali- ties promote eating disorders? Journal of Broadcasting & Electronic Media, 41, 478-500.

Harrison, K. (2000a). Television viewing, fat stereotyping, body shape stan- dards, and eating disorder symptomatology in grade school children. Com- munication Research, 27, 617-640.

Harrison, K. (2000b). The body electric: Thin-ideal media and eating disor- ders in adolescents. Journal of Communication, 50(3), 119-143.

611

Park • Presumed Media Influence

Harrison, K., & Cantor, J. (1997). The relationship between media consump- tion and eating disorders. Journal of Communication, 47(1), 40-67.

Harrison, K., & Fredrickson, B. L. (2003). Women’s sports media, self- objectification, and mental health in Black and White adolescent females. Journal of Communication, 53(2), 216-232.

Heinberg, L. J. (1996). Theories of body image disturbance: Perceptual, devel- opmental, and sociocultural factors. In J. K. Thompson (Ed.), Body image, eating disorders, and obesity: An integrative guide for assessment and treatment (pp. 27-48). Washington, DC: American Psychological Association.

Heinberg, L. J., & Thompson, J. K. (1995). Body image and televised images of thinness and attractiveness: A controlled laboratory investigation. Jour- nal of Social and Clinical Psychology, 14, 325-338.

Irving, L. M. (1990). Mirror images: Effects of the standard of beauty on the self- and body-esteem of women exhibiting varying levels of bulimic symp- toms. Journal of Social and Clinical Psychology, 9, 230-242.

Jacoby, L., & Cash, T. F. (1994). In pursuit of the perfect appearance: Discrep- ancies among self-ideal percepts of multiple physical attributes. Journal of Applied Social Psychology, 24, 379-396.

Kates, S. M., & Shaw-Garlock, G. (1999). The ever entangling web: A study of ideologies and discourses in advertising to women. Journal of Advertising, 28(2), 33-49.

Kaufman, L. (1980). Prime-time nutrition. Journal of Communication, 30(3), 37-46.

Kilbourne, J. (Producer). (1995). Slim hopes: Advertising and the obsession with thinness [Motion picture]. Northampton, MA: Media Education Foundation.

Lee, A. M., & Lee, E. B. (Eds.). (1939). The fine art of propaganda: A study of Father Caughlin’s speeches. New York: Harcourt Brace.

Levchuck, C. M., Kosek, J. K., & Drohan, M. (2000). Eating disorders. In A. McNeill (Ed.), Healthy living: Exercise, nutrition, and other healthy habits (pp. 355-380). Farmington Hills, MI: UXL.

Levine, M. P., Smolak, L., & Hayden, H. (1994). The relation of sociocultural factors to eating attitudes and behaviors among middle school girls. Jour- nal of Early Adolescence, 14(4), 471-490.

Mazur, A. (1986). U.S. trends in feminine beauty and overadaptation. The Journal of Sex Research, 22, 281-303.

Mazzeo, S. E., & Espelage, D. L. (2002). Association between childhood physi- cal and emotional abuse and disordered eating behaviors in female under- graduates: An investigation of the mediating Role of alexithymia and depression. Journal of Counseling Psychology, 49(1), 86-100.

McCarthy, M. (1990). The thin ideal, depression and eating disorders in women. Behavior Research Therapy, 28(3), 205-215.

Milkie, M. (1999). Social comparisons, reflected appraisals, and mass media: The impact of pervasive beauty images on Black and White girls’ self-con- cepts. Social Psychology Quarterly, 62(2), 190-210.

Muller, M. A. (1998). Getting what you want from your body image. Chestnut Hills, MA: ETR.

612

COMMUNICATION RESEARCH • October 2005

Myers, P. N., & Biocca, F. A. (1992). The elastic body image: The effect of televi- sion advertising and programming on body image distortions in young women. Journal of Communication, 42(3), 108-133.

Nassar, M. (1988). Culture and weight consciousness. Journal of Psychoso- matic Research, 32, 573-577.

Noelle-Neumann, E. (1974). The spiral of silence: A theory of public opinion. Journal of Communication, 24, 43-51.

Pike, K. M. (1995). Bulimic symptomatology in high-school girls: Toward a model of cumulative risk. Psychology of Women Quarterly, 19, 373-396.

Posavac, H. D., Posavac, S. S., & Posavac, E. J. (1998). Exposure to media images of female attractiveness and concern with body weight among young women. Sex Roles, 38(3/4), 187-201.

Shisslak, C. M., Crago, M., & Estes, L. S. (1995). The spectrum of eating dis- turbances. International Journal of Eating Disorders, 18(3), 209-219.

Silverstein, B., Perdue, L., Peterson, B., & Kelly, E. (1986). The role of the mass media in promoting a thin standard of bodily attractiveness for women. Sex Roles, 13(9/10), 519-532.

Stice, E., Schupak-Neuberg, E., Shaw, H. E., & Stein, R. I. (1994). Relation of media exposure to eating disorder symptomatology: An examination of mediating mechanisms. Journal of Abnormal Psychology, 103, 836-840.

Stice, E., & Shaw, H. E. (1994). Adverse effects of the media portrayed thin- ideal on women and linkages to bulimic symptomatology. Journal of Social and Clinical Psychology, 13, 288-308.

Striegel-Moore, R., McAvay, G., & Rodin, J. (1986). Psychological and behav- ioral correlates of feeling fat in women. International Journal of Eating Disorders, 5, 935-947.

Taylor, C. B., Sharpe, T., Shisslak, C., Bryson, S., Estes, L. S., Gray, N., et al. (1998). Factors associated with weight concerns in adolescent girls. Inter- national Journal of Eating Disorders, 24(1), 31-42.

Thompson, C. J., & Haytko, D. L. (1997). Speaking of fashion: Consumers’ uses of fashion discourses and the appropriation of countervailing cultural meanings. Journal of Consumer Research, 24(June), 15-42.

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.

Thomsen, S. R. (2002). Health and beauty magazine reading and body shape concerns among a group of college women. Journalism & Mass Communi- cation Quarterly, 79(4), 988-1007.

Thomsen, S. R., McCoy, J. K., Gustafson, R. L., & Williams, M. (2002). Moti- vations for reading beauty and fashion magazines and anorexic risk in college-age women. Media Psychology, 4(2), 113-135.

Vincent, M. A., & McCabe M. P. (2000). Gender differences among adolescents in family, and peer influences on body dissatisfaction, weight loss, and binge eating behaviors. Journal of Youth and Adolescence, 29(2), 205-221.

White, J. H. (1992). Women and eating disorders, Part I: Significance and socio- cultural risk factors. Health Care for Women International, 13(4), 351-362.

613

Park • Presumed Media Influence

Yun, G. W., & Trumbo, C. D. (2000). Comparative response to a survey exe- cuted by post, e-mail, & web form. Journal of Computer-Mediated Commu- nication, 6(1). Retrieved May 5, 2004, from http://www.ascusc.org/jcmc/ vol6/issue1/yun.html

Sung-Yeon Park, Ph.D., is an assistant professor in the Department of Telecommunications at the School of Communication Studies at Bowling Green State University. Her research focuses on consumer issues in marketing communication.

614

COMMUNICATION RESEARCH • October 2005