- Email: [email protected]
Is fat talk more believable than self-affirming body talk?
Body Image 19 (2016) 122–125
Contents lists available at ScienceDirect
Body Image journal homepage: www.elsevier.com/locate/bodyimage
Is fat talk more believable than self-affirming body talk? Alexandra F. Corning a,∗ , Michaela M. Bucchianeri b a b
Department of Psychology, University of Notre Dame, Notre Dame, IN, United States Department of Psychological Science, Gustavus Adolphus College, St. Peter, MN, United States
a r t i c l e
i n f o
a b s t r a c t
Article history: Received 13 August 2014 Received in revised form 20 April 2016 Accepted 3 September 2016 Keywords: Fat talk Body dissatisfaction Overweight Body image Prevention
Two experiments tested the extent to which the believability of women’s body statements (fat talk or selfaffirming) depends on their body type (thin or overweight). Experiment 1 (N = 130) revealed fat talk was more believable than self-affirming talk regardless of body type. Experiment 2’s (N = 125) results showed, as hypothesized, that overweight women’s fat talk was significantly more believable than fat talk by thin women and self-affirming talk by either thin or overweight women. Consistent with Experiment 1, there was a trend in Experiment 2 toward thin women’s fat talk being more believable than their self-affirming talk. Overall, fat talk generally may be perceived as more believable than self-affirming body talk, and overweight women’s fat talk may be perceived as most authentic. These results have implications for increasing understanding of fat talk’s potential role in body dissatisfaction as well as the development of positive body image campaigns. © 2016 Elsevier Ltd. All rights reserved.
Introduction “Fat talk” – the self-abasing, conversational exchange about food, weight, or the body in which especially girls and women engage – is garnering increasing empirical attention. Fat talk is important to understand because not only is it correlated with elevated body image concerns (e.g., Arroyo & Harwood, 2012; Clarke, Murnen, & Smolak, 2010; Corning & Gondoli, 2012), but it can actually increase body dissatisfaction (Corning, Bucchianeri, & Pick, 2014; Stice, Maxfield, & Wells, 2003; Tucker, Martz, Curtin, & Bazzini, 2007) and is correlated with eating pathology (e.g., Clarke et al., 2010; Ousley, Cordero, & White, 2008). Moreover, one person’s fat talk causes conversation partners to engage in fat talk (e.g., Salk & Engeln-Maddox, 2012). To what extent, however, does the listener believe the fat-talker actually thinks herself “fat”? Nichter and Vuckovic (1994) inferred from their observations of adolescent girls that others often do not believe fat-talkers’ self-abasing body statements. They proposed that fat talk comments such as “I’m so fat” are often construed as symbolic of other needs (e.g., calls for reassurance one is not fat). The extent to which women believe that other women mean the content of their body-related statements may have both theoretical and practical implications. Believability may be important to
∗ Corresponding author at: Department of Psychology, University of Notre Dame, 118 Haggar Hall, Notre Dame, IN 46556, United States. E-mail address: [email protected] (A.F. Corning). http://dx.doi.org/10.1016/j.bodyim.2016.09.004 1740-1445/© 2016 Elsevier Ltd. All rights reserved.
better understanding the contagious nature of fat talk, the relation between fat talk and subsequent increases in body dissatisfaction, as well as how to best focus programmatic efforts to promote body positivity. First, the contagious nature of fat talk may be owed in part to the extent to which the listener perceives the fat-talk remark as authentic. Because disclosures perceived as authentic tend to prompt reciprocal responses (e.g., see Cialdini, 2001), if a woman perceives her conversation partner’s fat talk as how she actually views herself, the listener may then respond in kind. Second, that fat talk can increase others’ body dissatisfaction may be owed to the degree of credibility ascribed to others’ fat talk. Because a source’s credibility enhances its persuasive potential (Pornpitakpan, 2004), if other women’s fat talk is perceived as authentic, she may in turn scrutinize her own body for perceived “flaws,” leading thus to an increase in her own body dissatisfaction. Finally, the benefits of body positivity programs would be enhanced by better understanding the extent to which participants believe the content of body statements made by spokespersons. Compounding matters, within the context of a thin-idealizing culture, women who are overweight are expected to feel negatively about their bodies. Thus, their fat-talk statements likely would be perceived as highly believable due to the widespread, negative bias and prejudice overweight persons experience (Puhl & Heuer, 2009); this could have implications for the perpetuation of both fat talk and weight stigma. We presented women with photos of thin or overweight women, each accompanied by either a fat-talk or self-affirming body-related statement the woman in the photo ostensibly made.
A.F. Corning, M.M. Bucchianeri / Body Image 19 (2016) 122–125
Participants indicated the extent to which they believed each woman’s statement. Because body image concerns are commonplace in our thin-idealizing culture and negative bias toward fatness is pervasive, we hypothesized that the believability of women’s body statements is dependent on both the type of statement made (self-abasing or self-affirming) and the speaker’s body type (thin or overweight). Specifically, we hypothesized that fat talk, especially by overweight women, is significantly more likely to be believed than thin women fat-talking or thin and overweight women making self-affirming statements. Experiment 1 Method Participants. Participants (N = 130) were undergraduate women (Mage , SD = 19.09, 1.21). The majority was European American (72.09%) and one-fifth was Latina (20.16%); smaller numbers were Asian American (13.95%), African American (3.88%), Native American (3.1%), and Middle Eastern (1.55%). The average body mass index (BMI) fell within the normal range, M (SD) = 22.18 (3.49). Most (80%) were normal weight, whereas another 6.92% was underweight, 10.77% was overweight, and 2.31% was obese. Materials and measures. The stimulus materials and measures were a pre-piloted set of photos of women making body-related statements, each followed by a question set which included the target item assessing how believable each woman’s statements were, followed by a demographic questionnaire. Photo-statement stimuli. The substantive stimulus materials comprised eight photos of women’s bodies and eight accompanying body-related statements (Corning et al., 2014). The photos were of four thin women and four overweight women. The bodyrelated statements comprised four fat-talk and four self-affirming statements. Detailed description of their development and piloting is provided in Corning et al. (2014), and briefly summarized below. The photos and statements were crossed to derive eight, unique photo-statement pairs: two were of thin women fat-talking, two of overweight women fat-talking, two of thin women making self-affirming statements, and two of overweight women making self-affirming statements. The presentation order of the photostatement pairs was varied to avoid a patterned sequence. To increase the stimulus materials’ relatability, participants were told the photos and statements were from women who had participated in a prior study and were used with their permission. Body-type photos. Four of the photos depict a thin body type and four depict an overweight body type, with a very large difference in the pilot sample’s ratings of those judged as thin versus overweight (Cohen’s d = 9.45; Corning et al., 2014). The women in the photos were depicted full-length or from at least mid-thigh to top of head (these were distributed evenly across conditions). All photos had been piloted to ensure they depicted women who were of college age and were of equivalent facial attractiveness.
123
as representing fat talk versus self-affirming body talk (d = 33.3; Corning et al., 2014). Believability. Below each photo-statement pair was a set of items. The set included the target believability item along with distractor items asking about, for example, the target’s likeability and happiness levels. The target item read, “How much do you believe that she really means the statement she made about her body?,” from 1 (don’t believe at all) to 7 (completely believe). Demographic information. The demographic questionnaire included self-reports of age, year in school, race/ethnicity, height, and weight. Procedure. Participants were recruited from introductory psychology courses at a Midwestern university. The solicitation, posted to the department’s electronic bulletin board, described the study as examining women’s attitudes, behaviors, and feelings. Participation was voluntary and compensated with class extra-credit. All materials were administered online at a quiet location of the participants’ choosing. The study was approved by the university’s institutional review board. Upon logging on, participants were greeted by a welcome page followed by an informed consent form. They were then directed to one of three counterbalanced versions of the study materials. Before exiting, they were presented with a debriefing statement and the experimenters’ contact information. The counterbalanced study measures and materials were administered in the following way. Each page displayed a photo of a woman paired with a statement she ostensibly made, followed by the item assessing the believability of the statement, which was embedded within distractor items. (To familiarize participants with the task, prior to presentation of these stimuli, they were presented with a pre-trial, unscored, neutral photo-statement pair depicting an average-weight woman who made a neutral body statement followed by the set of questions.) Following the last question set, participants completed the demographic questionnaire. Design. Experiment 1 employed a 2 × 2 mixed-factorial design. The between-subjects factor was body type (thin vs. overweight) and the within-subject factor was body statement type (fat talk vs. self-affirming). Participants were randomly assigned to either the thin (n = 65) or overweight body type condition (n = 65). In each condition, participants were presented with four pictures of women (either thin or overweight), two of whom made fat-talk and two of whom made self-affirming statements. The two believability scores for each photo-statement pairing (e.g., the two thin women fattalking) were averaged to derive a believability mean for each of the four conditions. To provide a further check of the body-type manipulation beyond that offered by the pilot data, the substantive materials were followed by a manipulation-check question. It read, “As a group, I’d say these past participants generally had bodies that, to me, looked .” Participants selected a value ranging from 1 (thin) to 9 (overweight) (with all intervening points provided and the midpoint, 5, labeled average). Results
Body-related statements. Four of the eight statements represent fat talk (e.g., “Every time I look in the mirror, I can pick out something about every part of my body that I’d like to change. I can’t remember the last time I was happy with the way I looked.”) and four represent self-affirming body statements (e.g., “Other women talk about their weight in numbers, but I look in the mirror, and I like what I see! This is my body, and I’m going to own it.”). There was a very large difference in the pilot sample’s ratings of those judged
Prior to conducting the main analyses, we conducted several checks of the data. First, we found one missing datum. It was in response to the race/ethnicity query and thus unimputable. Second, as expected, the photos of the thin (M, SD = 2.82, 1.26) versus overweight (M, SD = 6.95, .89) women were perceived accordingly, F(1, 128) = 466.75, p < .001, d = 3.79. Finally, as expected, participant BMI did not differ across conditions, F(1, 128) = 1.47, p = .23.
124
A.F. Corning, M.M. Bucchianeri / Body Image 19 (2016) 122–125
6
5.72 (.86)
Believability Ratings
5.49 (1.17) 5
5.07 (1.09) 4.51 (1.21) Overweight Women Thin Women
4
3
Self-Affirming Statements
Fat Talk
Fig. 1. Experiment 1 employed a 2 × 2 mixed-factorial design. Condition means (SDs) show the extent to which body-related statements by overweight and thin women were perceived as believable. Higher scores indicate greater believability.
Next, we tested for the hypothesized interaction of statement type and body type via a two-way mixed ANOVA. The interaction did not reach significance, F(1, 128) = 2.12, p = .15 (see Fig. 1). There was, however, a main effect of statement type: regardless of whether speakers were thin or overweight, their body statements were significantly more believable when they fat-talked than when they made self-affirming statements, F(1, 128) = 49.31, p = .001; Cohen’s d = 0.74. Given that the data reflected the expected direction of effects (see Fig. 1), we re-designed the experiment using a new sample. Specifically, we designed Experiment 2 as fully within-subject because this approach minimizes chance variation across conditions, decreasing error variance associated with individual differences (because all participants rotate through all conditions and, in this way, serve as their own controls), and thus permits a more powerful test of the effect. Experiment 2 Method Participants. Participants (N = 125) were undergraduate women, M (SD) age = 19.04 (1.21). Most were European American (73.6%); many fewer were Latina (13.6%), Asian American (6.4%), Middle Eastern (3.2%), African American (1.6%), Native American (.8%), or “other” (.8%). The mean BMI was 22.14 (SD = 2.78) and most of the sample (79.2%) was of normal weight. Materials and measures and procedure. The materials, measures, and procedure were identical to those in Experiment 1. Design. Experiment 2 employed a 2 (thin vs. overweight body type) × 2 (fat talk vs. self-affirming body statement) fully withinsubject, counterbalanced design, through which each participant rotated twice. That is, each participant was presented with eight photo-statement pairings – two each of women of each body type making each type of body statement. The two believability scores for each photo-statement pairing (e.g., the two thin women fattalking) were averaged to derive a believability mean for each condition. Results Given no missing data, we proceeded to the substantive test. Results of a two-way, within-subject ANOVA indicated a significant interaction, F(3, 372) = 15.69, p < .001 (see Fig. 2). We then
conducted follow-up, paired mean comparisons to discern between which conditions lay significant differences, first Bonferroniadjusting alpha to reduce the family-wise error rate (˛ = .05/5 tests = .01). The comparison tests revealed that, as hypothesized, overweight women’s fat talk was perceived as more credible than: (a) thin women’s fat talk, F(1, 124) = 42.32, p < .001, d = 0.55; (b) overweight women’s self-affirming body statements, F(1, 124) = 10.52, p = .002, d = 0.43; and (c) thin women’s self-affirming body statements, F(1, 124) = 46.06, p < .001, d = 0.84. Thin women’s fat talk statements received higher believability ratings than their self-affirming statements, but this difference did not achieve statistical significance, F(1, 124) = 3.84, p = .05. Taken with finding (b), the direction of this latter finding paints a picture not inconsistent with the main effect found in Study 1, wherein fat talk statements were rated as more believable than self-affirming statements. Finally, overweight women’s self-affirming statements were found more (not less) believable than thin women’s self-affirming statements, F(1, 124) = 13.59, p < .001, d = 0.37. Discussion These experiments yielded mixed support regarding the believability of a woman’s body statements (fat talk or self-affirming) being dependent on her body type (thin or overweight). In Experiment 1, fat talk was significantly more believable than selfaffirming talk regardless of the speaker’s body type. In Experiment 2, overweight women’s fat talk was significantly more believable than fat talk by thin women and self-affirming talk by both thin and overweight women. There was a trend in Experiment 2 toward thin women’s fat talk being more believable than their self-affirming talk, but the difference did not reach statistical significance. In sum, fat talk may generally be perceived as more believable than self-affirming body talk, with there being some evidence that overweight women’s fat talk may be seen as most authentic. These results have potential theoretical and research implications for understanding the relation between fat talk and body dissatisfaction. Research shows that fat talk can actually increase body dissatisfaction in other women (Corning et al., 2014; Stice et al., 2003; Tucker et al., 2007), but the mechanism by which this occurs has not been identified. Based on the present study’s results, it may be that because other women’s body self-disparagement is generally perceived as authentic, fat talk signals to conversation partners that they too should scrutinize their own bodies for perceived flaws, and then voice them, resulting, perhaps necessarily, in their increased body dissatisfaction. That is, because greater source credibility is persuasive in gaining behavioral compliance
A.F. Corning, M.M. Bucchianeri / Body Image 19 (2016) 122–125
125
6
Believability Ratings
5.58 (1.08) 5.08 (1.26)
4.94 (1.25)
5
4.63 (1.18) Overweight Women Thin Women 4
3
Self-Affirming Statements
Fat Talk
Fig. 2. Experiment 2 employed a 2 × 2 within-subjects factorial design. Condition means (SDs) show the extent to which body-related statements by overweight and thin women were perceived as believable. Higher scores indicate greater believability.
(Pornpitakpan, 2004), it is possible that a conversant who hears fat talk and judges it to be a valid indicator of the other woman’s negative attitude toward her body, is, in turn, prompted to selfscrutinize, which results in her own increased body dissatisfaction. This potential pathway remains to be tested. These results also have practical implications for positive body image promotion efforts. The results indicate that self-affirming body statements generally were viewed as less credible than fat talk. Decades of research on social persuasion in health care, advertising, and other social influence contexts indicate that more credible sources have significantly greater influence potential (Pornpitakpan, 2004). These results suggest that self-abasing and self-affirming body messages may not have equivalent power to influence body satisfaction and weight stigma. Women may be less responsive to self-affirming statements because they may be perceived as less authentic. Those involved in body image campaigns may bolster the effectiveness of their messages by incorporating other factors known to enhance persuasion when source credibility is lower (e.g., perceived similarity between source and audience, argument strength) (see Pornpitakpan, 2004, for suggestions). Some limitations of this study can be addressed in future work. First, whereas experimental designs provide the control from which results confidently can be inferred, they lack the ecological validity field studies possess. Participants did not have access to cues such as facial expression and tone of voice to help judge sincerity. Second, these studies should be replicated in racial/ethnic minority groups because the meaning of body discourse may vary intra-culturally. Finally, future studies will benefit from inclusion of both a neutral-statement condition and a normal-weight condition to better understand how a range of body-related statements of normal-weight women are interpreted.
References Arroyo, A., & Harwood, J. (2012). Exploring the causes and consequences of engaging in fat talk. Journal of Applied Communication Research, 40, 167–187. http://dx.doi. org/10.1080/00909882.2012.654500 Cialdini, R. B. (2001). Influence: Science and practice (4th ed.). Boston: Allyn & Bacon. Clarke, P. M., Murnen, S. K., & Smolak, L. (2010). Development and psychometric evaluation of a quantitative measure of “fat talk”. Body Image, 7, 1–7. http://dx. doi.org/10.1016/j.bodyim.2009.09.006 Corning, A. F., Bucchianeri, M. M., & Pick, C. M. (2014). Thin or overweight women’s fat talk: Which is worse for other women’s body satisfaction? Eating Disorders: The Journal of Treatment and Prevention, 22, 121–135. http://dx.doi.org/10.1080/ 10640266.2013.860850 Corning, A. F., & Gondoli, D. M. (2012). Who is most likely to fat talk? A social comparison perspective. Body Image, 9, 528–531. http://dx.doi.org/10.1016/j.bodyim. 2012.05.004 Nichter, M., & Vuckovic, N. (1994). Fat talk. In N. Sault (Ed.), Many mirrors: Body image and social relations (pp. 109–131). New Brunswick, NJ: Rutgers University Press. Ousley, L., Cordero, E. D., & White, S. (2008). Fat talk among college students: How undergraduates communicate regarding food and body weight, shape and appearance. Eating Disorders: The Journal of Treatment and Prevention, 16, 73–84. http://dx.doi.org/10.1080/10640260701773546 Pornpitakpan, C. (2004). The persuasiveness of source credibility: A critical review of five decades’ evidence. Journal of Applied Social Psychology, 34, 243–281. http:// dx.doi.org/10.1111/j.1559-1816.2004.tb02547.x Puhl, R. M., & Heuer, C. A. (2009). The stigma of obesity: A review and update. Obesity, 17, 941–964. http://dx.doi.org/10.1038/oby.2008.636 Salk, R. H., & Engeln-Maddox, R. (2012). Fat talk among college women is both contagious and harmful. Sex Roles, 66, 636–645. http://dx.doi.org/10.1007/s11199011-0050-1 Stice, E., Maxfield, J., & Wells, T. (2003). Adverse effects of social pressure to be thin on young women: An experimental investigation of the effects of “fat talk”. International Journal of Eating Disorders, 34, 108–117. http://dx.doi.org/10.1002/ eat.10171 Tucker, K. L., Martz, D. M., Curtin, L. A., & Bazzini, D. G. (2007). Examining fat talk experimentally in a female dyad: How are women influenced by another woman’s body presentation style? Body Image, 4, 157–164. http://dx.doi.org/10. 1016/j.bodyim.2006.12.005
Contents lists available at ScienceDirect
Body Image journal homepage: www.elsevier.com/locate/bodyimage
Is fat talk more believable than self-affirming body talk? Alexandra F. Corning a,∗ , Michaela M. Bucchianeri b a b
Department of Psychology, University of Notre Dame, Notre Dame, IN, United States Department of Psychological Science, Gustavus Adolphus College, St. Peter, MN, United States
a r t i c l e
i n f o
a b s t r a c t
Article history: Received 13 August 2014 Received in revised form 20 April 2016 Accepted 3 September 2016 Keywords: Fat talk Body dissatisfaction Overweight Body image Prevention
Two experiments tested the extent to which the believability of women’s body statements (fat talk or selfaffirming) depends on their body type (thin or overweight). Experiment 1 (N = 130) revealed fat talk was more believable than self-affirming talk regardless of body type. Experiment 2’s (N = 125) results showed, as hypothesized, that overweight women’s fat talk was significantly more believable than fat talk by thin women and self-affirming talk by either thin or overweight women. Consistent with Experiment 1, there was a trend in Experiment 2 toward thin women’s fat talk being more believable than their self-affirming talk. Overall, fat talk generally may be perceived as more believable than self-affirming body talk, and overweight women’s fat talk may be perceived as most authentic. These results have implications for increasing understanding of fat talk’s potential role in body dissatisfaction as well as the development of positive body image campaigns. © 2016 Elsevier Ltd. All rights reserved.
Introduction “Fat talk” – the self-abasing, conversational exchange about food, weight, or the body in which especially girls and women engage – is garnering increasing empirical attention. Fat talk is important to understand because not only is it correlated with elevated body image concerns (e.g., Arroyo & Harwood, 2012; Clarke, Murnen, & Smolak, 2010; Corning & Gondoli, 2012), but it can actually increase body dissatisfaction (Corning, Bucchianeri, & Pick, 2014; Stice, Maxfield, & Wells, 2003; Tucker, Martz, Curtin, & Bazzini, 2007) and is correlated with eating pathology (e.g., Clarke et al., 2010; Ousley, Cordero, & White, 2008). Moreover, one person’s fat talk causes conversation partners to engage in fat talk (e.g., Salk & Engeln-Maddox, 2012). To what extent, however, does the listener believe the fat-talker actually thinks herself “fat”? Nichter and Vuckovic (1994) inferred from their observations of adolescent girls that others often do not believe fat-talkers’ self-abasing body statements. They proposed that fat talk comments such as “I’m so fat” are often construed as symbolic of other needs (e.g., calls for reassurance one is not fat). The extent to which women believe that other women mean the content of their body-related statements may have both theoretical and practical implications. Believability may be important to
∗ Corresponding author at: Department of Psychology, University of Notre Dame, 118 Haggar Hall, Notre Dame, IN 46556, United States. E-mail address: [email protected] (A.F. Corning). http://dx.doi.org/10.1016/j.bodyim.2016.09.004 1740-1445/© 2016 Elsevier Ltd. All rights reserved.
better understanding the contagious nature of fat talk, the relation between fat talk and subsequent increases in body dissatisfaction, as well as how to best focus programmatic efforts to promote body positivity. First, the contagious nature of fat talk may be owed in part to the extent to which the listener perceives the fat-talk remark as authentic. Because disclosures perceived as authentic tend to prompt reciprocal responses (e.g., see Cialdini, 2001), if a woman perceives her conversation partner’s fat talk as how she actually views herself, the listener may then respond in kind. Second, that fat talk can increase others’ body dissatisfaction may be owed to the degree of credibility ascribed to others’ fat talk. Because a source’s credibility enhances its persuasive potential (Pornpitakpan, 2004), if other women’s fat talk is perceived as authentic, she may in turn scrutinize her own body for perceived “flaws,” leading thus to an increase in her own body dissatisfaction. Finally, the benefits of body positivity programs would be enhanced by better understanding the extent to which participants believe the content of body statements made by spokespersons. Compounding matters, within the context of a thin-idealizing culture, women who are overweight are expected to feel negatively about their bodies. Thus, their fat-talk statements likely would be perceived as highly believable due to the widespread, negative bias and prejudice overweight persons experience (Puhl & Heuer, 2009); this could have implications for the perpetuation of both fat talk and weight stigma. We presented women with photos of thin or overweight women, each accompanied by either a fat-talk or self-affirming body-related statement the woman in the photo ostensibly made.
A.F. Corning, M.M. Bucchianeri / Body Image 19 (2016) 122–125
Participants indicated the extent to which they believed each woman’s statement. Because body image concerns are commonplace in our thin-idealizing culture and negative bias toward fatness is pervasive, we hypothesized that the believability of women’s body statements is dependent on both the type of statement made (self-abasing or self-affirming) and the speaker’s body type (thin or overweight). Specifically, we hypothesized that fat talk, especially by overweight women, is significantly more likely to be believed than thin women fat-talking or thin and overweight women making self-affirming statements. Experiment 1 Method Participants. Participants (N = 130) were undergraduate women (Mage , SD = 19.09, 1.21). The majority was European American (72.09%) and one-fifth was Latina (20.16%); smaller numbers were Asian American (13.95%), African American (3.88%), Native American (3.1%), and Middle Eastern (1.55%). The average body mass index (BMI) fell within the normal range, M (SD) = 22.18 (3.49). Most (80%) were normal weight, whereas another 6.92% was underweight, 10.77% was overweight, and 2.31% was obese. Materials and measures. The stimulus materials and measures were a pre-piloted set of photos of women making body-related statements, each followed by a question set which included the target item assessing how believable each woman’s statements were, followed by a demographic questionnaire. Photo-statement stimuli. The substantive stimulus materials comprised eight photos of women’s bodies and eight accompanying body-related statements (Corning et al., 2014). The photos were of four thin women and four overweight women. The bodyrelated statements comprised four fat-talk and four self-affirming statements. Detailed description of their development and piloting is provided in Corning et al. (2014), and briefly summarized below. The photos and statements were crossed to derive eight, unique photo-statement pairs: two were of thin women fat-talking, two of overweight women fat-talking, two of thin women making self-affirming statements, and two of overweight women making self-affirming statements. The presentation order of the photostatement pairs was varied to avoid a patterned sequence. To increase the stimulus materials’ relatability, participants were told the photos and statements were from women who had participated in a prior study and were used with their permission. Body-type photos. Four of the photos depict a thin body type and four depict an overweight body type, with a very large difference in the pilot sample’s ratings of those judged as thin versus overweight (Cohen’s d = 9.45; Corning et al., 2014). The women in the photos were depicted full-length or from at least mid-thigh to top of head (these were distributed evenly across conditions). All photos had been piloted to ensure they depicted women who were of college age and were of equivalent facial attractiveness.
123
as representing fat talk versus self-affirming body talk (d = 33.3; Corning et al., 2014). Believability. Below each photo-statement pair was a set of items. The set included the target believability item along with distractor items asking about, for example, the target’s likeability and happiness levels. The target item read, “How much do you believe that she really means the statement she made about her body?,” from 1 (don’t believe at all) to 7 (completely believe). Demographic information. The demographic questionnaire included self-reports of age, year in school, race/ethnicity, height, and weight. Procedure. Participants were recruited from introductory psychology courses at a Midwestern university. The solicitation, posted to the department’s electronic bulletin board, described the study as examining women’s attitudes, behaviors, and feelings. Participation was voluntary and compensated with class extra-credit. All materials were administered online at a quiet location of the participants’ choosing. The study was approved by the university’s institutional review board. Upon logging on, participants were greeted by a welcome page followed by an informed consent form. They were then directed to one of three counterbalanced versions of the study materials. Before exiting, they were presented with a debriefing statement and the experimenters’ contact information. The counterbalanced study measures and materials were administered in the following way. Each page displayed a photo of a woman paired with a statement she ostensibly made, followed by the item assessing the believability of the statement, which was embedded within distractor items. (To familiarize participants with the task, prior to presentation of these stimuli, they were presented with a pre-trial, unscored, neutral photo-statement pair depicting an average-weight woman who made a neutral body statement followed by the set of questions.) Following the last question set, participants completed the demographic questionnaire. Design. Experiment 1 employed a 2 × 2 mixed-factorial design. The between-subjects factor was body type (thin vs. overweight) and the within-subject factor was body statement type (fat talk vs. self-affirming). Participants were randomly assigned to either the thin (n = 65) or overweight body type condition (n = 65). In each condition, participants were presented with four pictures of women (either thin or overweight), two of whom made fat-talk and two of whom made self-affirming statements. The two believability scores for each photo-statement pairing (e.g., the two thin women fattalking) were averaged to derive a believability mean for each of the four conditions. To provide a further check of the body-type manipulation beyond that offered by the pilot data, the substantive materials were followed by a manipulation-check question. It read, “As a group, I’d say these past participants generally had bodies that, to me, looked .” Participants selected a value ranging from 1 (thin) to 9 (overweight) (with all intervening points provided and the midpoint, 5, labeled average). Results
Body-related statements. Four of the eight statements represent fat talk (e.g., “Every time I look in the mirror, I can pick out something about every part of my body that I’d like to change. I can’t remember the last time I was happy with the way I looked.”) and four represent self-affirming body statements (e.g., “Other women talk about their weight in numbers, but I look in the mirror, and I like what I see! This is my body, and I’m going to own it.”). There was a very large difference in the pilot sample’s ratings of those judged
Prior to conducting the main analyses, we conducted several checks of the data. First, we found one missing datum. It was in response to the race/ethnicity query and thus unimputable. Second, as expected, the photos of the thin (M, SD = 2.82, 1.26) versus overweight (M, SD = 6.95, .89) women were perceived accordingly, F(1, 128) = 466.75, p < .001, d = 3.79. Finally, as expected, participant BMI did not differ across conditions, F(1, 128) = 1.47, p = .23.
124
A.F. Corning, M.M. Bucchianeri / Body Image 19 (2016) 122–125
6
5.72 (.86)
Believability Ratings
5.49 (1.17) 5
5.07 (1.09) 4.51 (1.21) Overweight Women Thin Women
4
3
Self-Affirming Statements
Fat Talk
Fig. 1. Experiment 1 employed a 2 × 2 mixed-factorial design. Condition means (SDs) show the extent to which body-related statements by overweight and thin women were perceived as believable. Higher scores indicate greater believability.
Next, we tested for the hypothesized interaction of statement type and body type via a two-way mixed ANOVA. The interaction did not reach significance, F(1, 128) = 2.12, p = .15 (see Fig. 1). There was, however, a main effect of statement type: regardless of whether speakers were thin or overweight, their body statements were significantly more believable when they fat-talked than when they made self-affirming statements, F(1, 128) = 49.31, p = .001; Cohen’s d = 0.74. Given that the data reflected the expected direction of effects (see Fig. 1), we re-designed the experiment using a new sample. Specifically, we designed Experiment 2 as fully within-subject because this approach minimizes chance variation across conditions, decreasing error variance associated with individual differences (because all participants rotate through all conditions and, in this way, serve as their own controls), and thus permits a more powerful test of the effect. Experiment 2 Method Participants. Participants (N = 125) were undergraduate women, M (SD) age = 19.04 (1.21). Most were European American (73.6%); many fewer were Latina (13.6%), Asian American (6.4%), Middle Eastern (3.2%), African American (1.6%), Native American (.8%), or “other” (.8%). The mean BMI was 22.14 (SD = 2.78) and most of the sample (79.2%) was of normal weight. Materials and measures and procedure. The materials, measures, and procedure were identical to those in Experiment 1. Design. Experiment 2 employed a 2 (thin vs. overweight body type) × 2 (fat talk vs. self-affirming body statement) fully withinsubject, counterbalanced design, through which each participant rotated twice. That is, each participant was presented with eight photo-statement pairings – two each of women of each body type making each type of body statement. The two believability scores for each photo-statement pairing (e.g., the two thin women fattalking) were averaged to derive a believability mean for each condition. Results Given no missing data, we proceeded to the substantive test. Results of a two-way, within-subject ANOVA indicated a significant interaction, F(3, 372) = 15.69, p < .001 (see Fig. 2). We then
conducted follow-up, paired mean comparisons to discern between which conditions lay significant differences, first Bonferroniadjusting alpha to reduce the family-wise error rate (˛ = .05/5 tests = .01). The comparison tests revealed that, as hypothesized, overweight women’s fat talk was perceived as more credible than: (a) thin women’s fat talk, F(1, 124) = 42.32, p < .001, d = 0.55; (b) overweight women’s self-affirming body statements, F(1, 124) = 10.52, p = .002, d = 0.43; and (c) thin women’s self-affirming body statements, F(1, 124) = 46.06, p < .001, d = 0.84. Thin women’s fat talk statements received higher believability ratings than their self-affirming statements, but this difference did not achieve statistical significance, F(1, 124) = 3.84, p = .05. Taken with finding (b), the direction of this latter finding paints a picture not inconsistent with the main effect found in Study 1, wherein fat talk statements were rated as more believable than self-affirming statements. Finally, overweight women’s self-affirming statements were found more (not less) believable than thin women’s self-affirming statements, F(1, 124) = 13.59, p < .001, d = 0.37. Discussion These experiments yielded mixed support regarding the believability of a woman’s body statements (fat talk or self-affirming) being dependent on her body type (thin or overweight). In Experiment 1, fat talk was significantly more believable than selfaffirming talk regardless of the speaker’s body type. In Experiment 2, overweight women’s fat talk was significantly more believable than fat talk by thin women and self-affirming talk by both thin and overweight women. There was a trend in Experiment 2 toward thin women’s fat talk being more believable than their self-affirming talk, but the difference did not reach statistical significance. In sum, fat talk may generally be perceived as more believable than self-affirming body talk, with there being some evidence that overweight women’s fat talk may be seen as most authentic. These results have potential theoretical and research implications for understanding the relation between fat talk and body dissatisfaction. Research shows that fat talk can actually increase body dissatisfaction in other women (Corning et al., 2014; Stice et al., 2003; Tucker et al., 2007), but the mechanism by which this occurs has not been identified. Based on the present study’s results, it may be that because other women’s body self-disparagement is generally perceived as authentic, fat talk signals to conversation partners that they too should scrutinize their own bodies for perceived flaws, and then voice them, resulting, perhaps necessarily, in their increased body dissatisfaction. That is, because greater source credibility is persuasive in gaining behavioral compliance
A.F. Corning, M.M. Bucchianeri / Body Image 19 (2016) 122–125
125
6
Believability Ratings
5.58 (1.08) 5.08 (1.26)
4.94 (1.25)
5
4.63 (1.18) Overweight Women Thin Women 4
3
Self-Affirming Statements
Fat Talk
Fig. 2. Experiment 2 employed a 2 × 2 within-subjects factorial design. Condition means (SDs) show the extent to which body-related statements by overweight and thin women were perceived as believable. Higher scores indicate greater believability.
(Pornpitakpan, 2004), it is possible that a conversant who hears fat talk and judges it to be a valid indicator of the other woman’s negative attitude toward her body, is, in turn, prompted to selfscrutinize, which results in her own increased body dissatisfaction. This potential pathway remains to be tested. These results also have practical implications for positive body image promotion efforts. The results indicate that self-affirming body statements generally were viewed as less credible than fat talk. Decades of research on social persuasion in health care, advertising, and other social influence contexts indicate that more credible sources have significantly greater influence potential (Pornpitakpan, 2004). These results suggest that self-abasing and self-affirming body messages may not have equivalent power to influence body satisfaction and weight stigma. Women may be less responsive to self-affirming statements because they may be perceived as less authentic. Those involved in body image campaigns may bolster the effectiveness of their messages by incorporating other factors known to enhance persuasion when source credibility is lower (e.g., perceived similarity between source and audience, argument strength) (see Pornpitakpan, 2004, for suggestions). Some limitations of this study can be addressed in future work. First, whereas experimental designs provide the control from which results confidently can be inferred, they lack the ecological validity field studies possess. Participants did not have access to cues such as facial expression and tone of voice to help judge sincerity. Second, these studies should be replicated in racial/ethnic minority groups because the meaning of body discourse may vary intra-culturally. Finally, future studies will benefit from inclusion of both a neutral-statement condition and a normal-weight condition to better understand how a range of body-related statements of normal-weight women are interpreted.
References Arroyo, A., & Harwood, J. (2012). Exploring the causes and consequences of engaging in fat talk. Journal of Applied Communication Research, 40, 167–187. http://dx.doi. org/10.1080/00909882.2012.654500 Cialdini, R. B. (2001). Influence: Science and practice (4th ed.). Boston: Allyn & Bacon. Clarke, P. M., Murnen, S. K., & Smolak, L. (2010). Development and psychometric evaluation of a quantitative measure of “fat talk”. Body Image, 7, 1–7. http://dx. doi.org/10.1016/j.bodyim.2009.09.006 Corning, A. F., Bucchianeri, M. M., & Pick, C. M. (2014). Thin or overweight women’s fat talk: Which is worse for other women’s body satisfaction? Eating Disorders: The Journal of Treatment and Prevention, 22, 121–135. http://dx.doi.org/10.1080/ 10640266.2013.860850 Corning, A. F., & Gondoli, D. M. (2012). Who is most likely to fat talk? A social comparison perspective. Body Image, 9, 528–531. http://dx.doi.org/10.1016/j.bodyim. 2012.05.004 Nichter, M., & Vuckovic, N. (1994). Fat talk. In N. Sault (Ed.), Many mirrors: Body image and social relations (pp. 109–131). New Brunswick, NJ: Rutgers University Press. Ousley, L., Cordero, E. D., & White, S. (2008). Fat talk among college students: How undergraduates communicate regarding food and body weight, shape and appearance. Eating Disorders: The Journal of Treatment and Prevention, 16, 73–84. http://dx.doi.org/10.1080/10640260701773546 Pornpitakpan, C. (2004). The persuasiveness of source credibility: A critical review of five decades’ evidence. Journal of Applied Social Psychology, 34, 243–281. http:// dx.doi.org/10.1111/j.1559-1816.2004.tb02547.x Puhl, R. M., & Heuer, C. A. (2009). The stigma of obesity: A review and update. Obesity, 17, 941–964. http://dx.doi.org/10.1038/oby.2008.636 Salk, R. H., & Engeln-Maddox, R. (2012). Fat talk among college women is both contagious and harmful. Sex Roles, 66, 636–645. http://dx.doi.org/10.1007/s11199011-0050-1 Stice, E., Maxfield, J., & Wells, T. (2003). Adverse effects of social pressure to be thin on young women: An experimental investigation of the effects of “fat talk”. International Journal of Eating Disorders, 34, 108–117. http://dx.doi.org/10.1002/ eat.10171 Tucker, K. L., Martz, D. M., Curtin, L. A., & Bazzini, D. G. (2007). Examining fat talk experimentally in a female dyad: How are women influenced by another woman’s body presentation style? Body Image, 4, 157–164. http://dx.doi.org/10. 1016/j.bodyim.2006.12.005