The effects of meal size and body size on individuals' impressions of males and females

The effects of meal size and body size on individuals' impressions of males and females

Eating Behaviors 5 (2004) 117 – 132 The effects of meal size and body size on individuals’ impressions of males and females Yolanda Martins 1,*, Patr...

392KB Sizes 0 Downloads 58 Views

Eating Behaviors 5 (2004) 117 – 132

The effects of meal size and body size on individuals’ impressions of males and females Yolanda Martins 1,*, Patricia Pliner* , Corrie Lee 2 University of Toronto, Canada Accepted 14 January 2004

Abstract Male and female participants provided impression ratings for either a normal-weight or overweight male or female target, who was portrayed as eating either small or large meals. Males rated normal-weight targets as more physically attractive than overweight targets, whereas ratings of physical attractiveness were unaffected by the body size manipulation among female participants. In addition, among male targets, the overweight large eater was rated the least socially attractive. For female targets, males rated the normal-weight large eater as the most socially attractive, whereas females rated the normal-weight small eater as the most socially attractive. Results are discussed in terms of how body and meal sizes interact to affect impressions of others. D 2004 Elsevier Ltd. All rights reserved. Keywords: Meal size; Body size; Physical attractiveness

1. Introduction Much research and anecdotal evidence exist to support the notion that women in contemporary American culture are more concerned about body weight, food, and eating than are men. This is not surprising, given that women with thin bodies are perceived to be more successful in both personal and professional domains, in comparison to those with heavier bodies (Harris, Harris, & Bochner, 1982; Way, 1995). Essentially, it’s a halo effect for thinness: what is thin is also beautiful, good, and successful. * Corresponding authors. Yolanda Martins is to be contacted at School of Psychology, Flinders University, GPO Box 2100, Adelaide, South Australia 5001, Australia. Patricia Pliner, University of Toronto at Mississauga, 3359 Mississauga Road, Mississauga, Ontario, Canada L5L 1C6. E-mail addresses: [email protected] (Y. Martins), [email protected] (P. Pliner). 1 Now at Flinders University. 2 Now at Toronto, Canada. 1471-0153/$ - see front matter D 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.eatbeh.2004.01.008

118

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

The resulting pressure to be thin is associated with increased weight concern, body dissatisfaction, and dieting behavior (Fallon & Rozin, 1985; Herman & Polivy, 1980; Mintz & Betz, 1986; Polivy & Herman, 1987; Rozin & Fallon, 1988; Serdula et al., 1993). Chaiken and Pliner (1987) have hypothesized that being slim and eating lightly are sex-role appropriate behaviors for women; these behaviors may influence how others perceive them, thereby contributing to females’ greater concerns surrounding eating and body weight. These researchers asked male and female participants to provide their impressions of male or female targets who were portrayed (by means of an eating diary) as eating either a small or a large breakfast and lunch. Their results demonstrated that both male and female participants rated females who ate small meals as more physically attractive, more feminine, and less masculine than females who ate large meals. The ratings of male targets on these variables were unaffected by the meal size manipulation. In a subsequent study by Bock and Kanarek (1995), male and female participants rated male and female targets who were portrayed as eating small, medium, or large breakfasts and lunches. Consistent with the findings of Chaiken and Pliner (1987), both male and female participants judged the female targets as more feminine and physically attractive as meal size decreased. Male targets were perceived by both male and female participants as more feminine as meal size decreased, but the ratings of male targets’ physical attractiveness were unaffected by the meal size manipulation. The Ratings of both the male and female targets’ perceived social attractiveness3 were unaffected by the meal size manipulation. In contrast to this last finding, a similar study by Basow and Kobrynowicz (1993), using female targets only, demonstrated that males and females perceive women who are portrayed as eating small ‘‘feminine’’ meals as significantly more socially attractive than women who consume large ‘‘masculine’’ meals. These investigators also replicated the previous findings of greater perceived femininity and physical attractiveness in women eating smaller meals. While the studies described have focused on the amount of food eaten, other researchers have evaluated the effects of the type of food eaten on others’ perceptions of male and female eaters. Typically, in these studies, participants read dietary profiles that vary in terms of their fat content or general healthiness. These studies have found that both male and female participants rate male and female targets who consume low-fat or healthy diets as significantly more feminine, more physically attractive, and more socially attractive than male and female targets who consume high-fat or unhealthy diets (Mooney, DeTore, & Malloy, 1994; Stein & Nemeroff, 1995). Moreover, in a recent study by Mooney and Lorenz (1997), female targets who were portrayed as eating ‘feminine’ meals (as rated by pretest participants) were perceived as possessing greater social attractiveness than those eating ‘masculine’ meals, while type of meal had no such effect on the perceptions of male targets. Although the studies described so far have manipulated several different independent variables, all of the manipulations have focused on some aspect of the foods that targets were portrayed as eating— amount (Basow & Kobrynowicz, 1993; Bock & Kanarek, 1995; Chaiken & Pliner, 1987), healthiness/fat content (Mooney et al., 1994; Stein & Nemeroff, 1995), or femininity (Basow & Kobrynowicz, 1993; Mooney & Lorenz, 1997). Although these attributes are distinct, they co-occur naturally in foods; for example, the foods eaten by the ‘‘small’’ eaters in the Chaiken and Pliner (1987) study (i.e., a salad) tend to be associated with women, while the foods eaten by the ‘‘large’’ eaters (e.g., bacon cheeseburger) tend 3

For the purposes of this paper, the term social attractiveness will be used in lieu of the many terms written in the literature (i.e., social desirability, social appeal, social likeability, etc.). These terms are essentially trying to capture the same concept: how likeable others are perceived to be and how desirable it is to spend time with them.

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

119

to be associated with men. Similarly, the healthy foods in the Stein and Nemeroff (1995) study (chicken and fruit) tend to be foods more strongly associated with women and had a lower caloric value than the unhealthy foods used in this study. In light of the natural relationship between these attributes and the extreme difficulty encountered when trying to separate them experimentally,4 it makes sense to think of all of these studies as having manipulated independent variables that, if not exactly equivalent, are highly correlated. Moreover, these studies have also used a variety of dependent measures including perceived femininity (Basow & Kobrynowicz, 1993; Bock & Kanarek, 1995; Chaiken & Pliner, 1987; Stein & Nemeroff, 1995), physical attractiveness (Basow & Kobrynowicz, 1993; Bock & Kanarek, 1995; Chaiken & Pliner, 1987; Mooney et al., 1994; Stein & Nemeroff, 1995), and social attractiveness (Basow & Kobrynowicz, 1993; Bock & Kanarek, 1995; Mooney & Lorenz, 1997; Stein & Nemeroff, 1995). Research on the ‘‘what is beautiful is good’’ stereotype indicates that individuals tend to perceive those whom they consider physically attractive as possessing positive personality characteristics (Eagly, Ashmore, Makhijani, & Longo, 1991; Feingold, 1992; Wheeler & Kim, 1997). In particular, physically attractive people are thought to be more sociable, friendly, and likeable than unattractive ones, suggesting that physical and social attractiveness are highly correlated. Given that the person perception studies described earlier have basically manipulated the same independent variable, and have given the correlation between physical and social attractiveness as dependent variables, they all lead to the same conclusion: eating lightly/healthily/femininely increases the likelihood that individuals will be perceived as feminine and possessing positive physical and personality characteristics. This relationship may be one of the motives underlying females’ ‘light’ eating behavior (at least in social situations). In contrast to the ‘‘what is beautiful is good’’ stereotype, there is the ‘‘fat is bad’’ stereotype. Weight stigma research repeatedly demonstrates that those who are heavier5 are viewed more negatively on a variety of personality traits (Cahnman, 1968; Jasper & Klassen, 1990; Maddox, Back, & Liederman, 1968; Tiggemann & Rothblum, 1988). Harris et al. (1982) have found that male and female participants rate overweight targets as less physically attractive, intelligent, hardworking, popular, and successful than their normal-weight counterparts. In general, although both Australian and American college students perceive obese people as warmer and friendlier, they perceive them as less happy, less self-confident, more self-indulgent, less self-disciplined, lazier, and less attractive. Moreover, obese females are perceived as less sexually attractive, warm, and responsive than are non-obese females or males of any weight (Regan, 1996). Taken together, these studies have been interpreted as showing that eating lightly may be socially advantageous for both males and females because it increases the likelihood that they will be perceived as possessing positive personality characteristics and decreases the likelihood that they will be perceived as possessing negative personality characteristics. However, a closer examination of this literature reveals that several studies have reported that targets who are portrayed as eating small or healthy meals are also perceived as thinner than their large and/or unhealthy-eating counterparts, even if the presence of the same height and weight information, indicative of normal body weight, is present in all conditions (Bock & Kanarek, 1995; Chaiken & Pliner, 1987; Mooney et al., 1994; Stein & Nemeroff, 1995). Thus, 4

We (Martins and Pliner) have spent several months pilot testing meals (for a related study) that would unconfound these variables, to no avail, providing further support for the idea that these attributes are not orthogonal to one another. 5 Typically, overweight status is portrayed in these studies by describing the target as overweight or by presenting participants with the target’s weight and/or picture.

120

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

it may be that the finding that small or healthy eaters are perceived more positively than large or unhealthy eaters is mediated by the participants’ perceptions of body size. That is, no research to date has examined the independent effects of body size and food consumption variables (and possible interactions between them) on the impressions of others. In the world outside the laboratory, body size information is often available to individuals when they develop impressions of others. In addition, in social situations where food consumption is involved, these two types of information are both present. Thus, the addition of body size as an independent variable in these types of investigations would greatly enhance the ecological validity of these studies. Accordingly, in the present study, we examined separately the effects of body size and the amount eaten on perceptions of physical and social attractiveness. Specifically, we manipulated the target gender (male or female), meal size (large or small), and target body size (overweight or normal-weight) and assessed their effect on the ratings by male and female participants of the target’s physical and social attractiveness. Because some of the earlier studies obtained different responses to male and female targets (and others used female targets only), we decided to consider male and female targets separately in making our predictions. Beginning with males, the target group for whom we had the fewest expectations, our hypotheses were as follows: we expected that the overweight targets would be rated, by both male and female participants, as less physically attractive and less socially attractive than the normal-weight male targets. This expectation was based on findings from the stigma literature showing that overweight individuals are perceived less positively (Harris, 1990; Harris et al., 1982; Regan, 1996). We were unsure of what to expect for the mealsize manipulation because mealsize effects have not consistently been found for male targets, and it was unclear to us whether those found in previous investigations were mediated by perceived body size. For the female targets, we expected, based on the findings from the weight stigma literature, that both males and females would rate those who were overweight as less physically attractive and less socially attractive than those who were normal-weight. Although we were unsure of exactly what to predict for the meal-size manipulation, we suspected that it would have an effect over and above that expected for the body size manipulation. The results of prior research (Bock & Kanarek, 1995; Chaiken & Pliner, 1987; Harris et al., 1982; Regan, 1996) in this area suggests that male and female participants should rate the normal-weight, small eater as the most socially attractive, in comparison with the normal-weight large eater or the overweight targets. However, anecdotal evidence led us to speculate that males might rate the normal-weight, large eater as the most socially and physically attractive. Many males participating in other studies in our laboratory have remarked that they prefer women who eat heartily, in comparison with women who eat lightly, although their preference for such women appears to be confined to those who are thin or normal-weight. These kinds of comments lead us to consider the possibility that, at least among male participants, the ratings of female targets may not be consistent with the findings from previous empirical investigations.

2. Method 2.1. Design and overview Male and female participants rated their perceptions of the physical attractiveness and social attractiveness of a male or female target who was portrayed as being either normal-weight or overweight

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

121

and who reported eating (on two occasions) either small or large meals. Thus, the overall design of the study was a 2  2  2  2 (participant sex  target sex  target body size  meal size) factorial. 2.2. Participants Participants were 160 undergraduate students, ranging in age from 18–28, enrolled in various disciplines at the University of Toronto, with equal numbers participating in each of the 16 experimental conditions. Participation was voluntary; they did not receive course credit or remuneration for participating in the study. 2.3. Materials 2.3.1. Target questionnaires Target gender, target body size, and meal size were manipulated through the use of eight questionnaires (each including a two-meal food diary), ostensibly completed by the target person in a study in the previous year. To bolster the participants’ belief that a previous participant had completed the questionnaire, responses on this questionnaire were handwritten. The first page of each questionnaire introduced either a 21-year old male or female who was a 3rd year Anthropology student. The questionnaire also included a food preference measure, describing the target’s degree of liking for 15 foods, and an exercise measure that depicted the target as a person who exercised moderately and regularly. These aspects of the questionnaire were identical across the various conditions. 2.3.2. Manipulation of independent variables The body silhouettes, designed by Stunkard, Sorenson, and Schlusinger (1980), were utilized to convey target body size. For each gender, these consist of nine figure drawings, ranging from extremely thin to extremely obese. In the normal-weight conditions, the target had indicated that his/her current body size corresponded to Figure Drawing 4, whereas in the overweight conditions, the target had expressed that his/her current body size corresponded to Figure Drawing 7. The body silhouettes can be seen in Fig. 1. This scale has been utilized as a dependent variable in prior related research; findings related to it have proven to be robust (Fallon & Rozin, 1985; Mooney et al., 1994; Rozin & Fallon, 1988). Meal size was manipulated through the food diaries, which asked targets to describe their two most recent meals. In the smallmeal conditions, targets reported eating one of two small breakfasts and one of two small lunches, with an average total caloric content of 334. In the largemeal conditions, targets reported eating one of two large breakfasts and one of two large lunches, with an average caloric content of 2142. These meals had been previously used in the study of Chaiken and Pliner (1987; see Table 1 for examples of these meals). We recognize that this manipulation confounds meal size/healthiness/ femininity, a problem faced by prior investigations in this area, as noted in the Introduction. Since these variables are often confounded in the real world, we have chosen to maintain this confounding and recognize that we have sacrificed some internal validity for increased external validity. The target gender was manipulated in three ways. The targets’ first names (either Michael or Marianne), as well as their gender (male or female, respectively), were recorded on the cover page of the questionnaire, and, of course, targets had conveyed their current body size by choosing a silhouette from either the male or the female figures.

122

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

Fig. 1. Body silhouettes (adapted from Stunkard et al., 1980) used to convey the targets’ body size. In the normal-weight conditions, targets indicated that their current body size corresponded to Figure Drawing 4, and in the overweight conditions, targets indicated that their current body size corresponded to Figure Drawing 7.

2.3.3. Target rating questionnaire The dependent variables were assessed by means of 30 five-point bipolar adjective scales, which participants completed in terms of ‘‘the kind of person you think the target person is.’’ Of these, eight items are relevant for the purposes of this paper. Two served as manipulation checks for the meal and body size manipulations and one assessed the perceived physical attractiveness of the target, with higher scores representing greater perceived physical attractiveness. The remaining five items (likeable, desirable as a roommate, fun to spend time with, enjoyable to have a meal with, and pleasing as a friend) were averaged and served as our main dependent measure of social attractiveness (alpha coefficient of .73). Scores could range from one to five, with higher scores indicating greater social attractiveness. 2.4. Procedure The experimental session began with an explanation to participants that the participants in a previous study had rated their food preferences and indicated what they had eaten in their last two meals. The experimenter explained that she was interested in the ‘‘possibility that people form impressions of others on the basis of their food preferences.’’ The participants were informed that they would read a randomly selected questionnaire, filled out by one of the previous year’s participants, and then answer some questions about their impressions of that person.

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

123

Table 1 Examples of meals used in the food diaries Breakfast

Lunch

Small meals Bran muffin Butter Coffee with cream

Green salad Oil and vinegar dressing Coffee with cream

Large meals Fried eggs and sausages Toast with jam and butter Orange juice Coffee with cream and sugar

Vegetable soup Green salad French dressing Spaghetti w/ meat sauce Chocolate brownie Banana Large Coke

The participants were randomly assigned to read one of eight food autobiographies, which varied in target sex, target body size, and meal size, and then completed the target rating questionnaire. Upon completion of this package, the participants were thoroughly debriefed and thanked for their participation.

3. Results 3.1. Overview Since we had separate hypotheses for each target gender, all analyses were carried out separately for male and female targets. The main analyses were 2  2  2 ANOVAs, with sex of participant, target’s body size, and meal size as the independent variables, and physical and social attractiveness as the dependent variables. In addition, we examined the relationship between our main dependent variables because prior evidence illustrates that a positive relationship exists between physical and social attractiveness (Eagly et al., 1991; Feingold, 1992; Wheeler & Kim, 1997). 3.2. Manipulation checks First, the analyses described above were carried out on the manipulation check items assessing target body size (i.e., bipolar adjective scale with endpoints ‘‘thin’’ and ‘‘fat’’) and meal size (i.e., bipolar adjective scale with endpoints ‘‘eats very little’’ and ‘‘eats a lot’’) for the male and female targets. First, the main effects for target body size, Fmale targets(1,72) = 73.71, P < .01; Ffemale targets(1,72) = 28.94, P < .01, indicated that normal-weight male and female targets were rated as significantly thinner (Mmale targets = 2.68, Mfemale targets = 2.28) than their overweight counterparts were (Mmale targets = 4.03, Mfemale targets = 3.35). For the female targets, there were no other significant effects, but the analysis for the male targets also yielded a significant meal size by body size interaction, Fmale targets(1,72) = 4.96,

124

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

P < .05. An examination of the relevant means indicated that normal-weight male targets were rated as thinner when they were portrayed as eating small meals (vs. large), whereas the ratings of the overweight male targets did not differ as a function of meal size. The ANOVAs assessing the effectiveness of the meal size manipulation yielded significant main effects for meal size; both male and female targets who ate large meals were rated as eating significantly more (Mmale targets = 3.70, Mfemale targets = 3.48) than their do counterparts who ate small meals (Mmale targets = 2.72, Mfemale targets = 2.75), Fmale targets(1,72) = 18.32, P < .01, Ffemale targets(1,72) = 11.24, P < .01. There were no other significant effects in either of these analyses. Overall, these results illustrate that our body and meal size manipulations were successful. 3.3. Dependent variables 3.3.1. Perceived physical attractiveness The 2  2  2 ANOVA on the perceived physical attractiveness of the male targets produced a main effect for body size, F(1,72) = 11.98, P < .01, which was qualified by an interaction between the participant’s sex and body size F(1, 72) = 6.44, P < .05. An examination of the main effect revealed that, overall, the normal-weight male targets were rated as significantly more attractive (M = 3.50) than their overweight counterparts (M = 2.75). An examination of the interaction indicated that this effect was carried primarily by the behavior of male participants, who rated overweight targets as less attractive than normal-weight targets, whereas female participants’ ratings of attractiveness were unaffected by the body size manipulation (see Fig. 2). For female targets, the ANOVA on perceived attractiveness yielded the same effects as those found for the male targets; a main effect for body size, F(1,72) = 19.19, P < .01, and an interaction between the

Fig. 2. Mean ratings of perceived physical attractiveness of male targets as a function of body size and sex of participant.

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

125

Fig. 3. Mean ratings of perceived physical attractiveness of female targets as a function of body size and sex of participant.

Fig. 4. Mean ratings of perceived social attractiveness of male targets as function of meal and body sizes.

126

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

sex of the participant and the body size of the target, F(1,72) = 5.38, P < .05. Overall, normal-weight female targets were rated as significantly more attractive (M = 3.70) than their overweight counterparts were (M = 2.85). As seen in Fig. 3, the interaction revealed that males rated normal-weight targets as

Fig. 5. Mean ratings of perceived social attractiveness of female targets as a function of sex of participant, body size, and meal size.

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

127

more attractive than overweight targets, whereas ratings of physical attractiveness by female participants did not differ as a function of the female target’s body size. 3.3.2. Social attractiveness For the male targets, the 2  2  2 ANOVA on the main dependent measure of target social attractiveness yielded two significant effects, a main effect for body size, F(1,72) = 7.32, P < .05, and an interaction between body and meal size, F(1,72) = 4.94, P < .05. An examination of the means for the main effect revealed that participants rated the normal-weight target as more socially attractive (M = 3.61) than the overweight target (M = 3.28). The interaction is depicted in Fig. 4 and illustrates that participants rated overweight male targets who ate large meals as less socially attractive than overweight male targets who ate small meals or normal-weight male targets, regardless of their meal size. For the female targets, the only significant effect was an interaction involving all three independent variables (sex of participant, meal size, and target body size), F(1,72) = 4.58, P < .05. To understand the interaction, we examined the data separately for male and female participants; we found that males rated normal-weight female targets who ate large meals as more socially attractive than normal-weight female targets who ate small meals or overweight female targets who ate large or small meals (see Fig. 5). In contrast, the pattern of ratings for the female participants revealed that females rated normalweight, small eaters as the most socially attractive, followed by female targets of either weight who ate large meals. The overweight female target who ate a small meal was seen as the least socially attractive female target among female participants (see Fig. 5).6 3.3.3. Relationship between perceived physical attractiveness and social attractiveness Pearson r correlations between ratings of perceived physical attractiveness and ratings of social attractiveness were calculated for each of the four possible combinations of target and participant genders. For male targets, there was a significant positive correlation between male participants’ ratings of perceived physical and social attractiveness, r(40)=.32, P < .05, indicating that as male participants’ perceptions of male targets’ physical attractiveness increased, their ratings of their social attractiveness also increased. Among the female participants rating male targets, there was no correlation between the ratings of perceived physical and social attractiveness, r(40)=.19, ns. For the female targets, the same pattern of results obtained for the Male targets was found. The male participants’ ratings of perceived physical and social attractiveness were significantly positively correlated, r(40)=.47, P < .01, illustrating that as male participants’ perceptions of physical attractiveness increased, their ratings of female targets’ social attractiveness also increased. Among the female participants, there was no correlation between the ratings of perceived physical and social attractiveness, r(40)=.18, ns. 6

To rule out the possibility that the three-way interaction found above is mediated by the participants’ perceptions of meal size, an ANCOVA was conducted, using the same independent variables as described above and using the participants’ ratings of meal size as the covariate. This analysis yielded the same results as the analysis described in the body of this paper, F(1,71) = 4.40, P < .05, indicating that the interaction was not mediated by the participants’ perceptions of meal size. This analysis provides further support for the original finding described and indicates that males’ higher ratings of social attractiveness for female targets who are normal-weight and who ate large meals was not just due to the fact that the males in this group perceived these meals as smaller than the meals eaten by the other female targets.

128

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

4. Discussion The results of this study supported some of our hypotheses and failed to support others. For the physical attractiveness variable, our expectation that overweight male and female targets would be rated as less physically attractive by both male and female participants was confirmed. Interestingly, however, there were also interactions between sex of the participant and body size on the ratings of perceived physical attractiveness for both male and female targets. In both cases, male participants rated the overweight targets as less physically attractive than the normal-weight ones, whereas female participants’ ratings of attractiveness did not differ as a function of body size. These results suggest that among male participants, body size is an important component of physical attractiveness, whereas female participants are not influenced by body size when rating the physical attractiveness of others. Given that prior research has demonstrated the link between physical attractiveness and desirability (Eagly et al., 1991; Feingold, 1992; Wheeler & Kim, 1997), it is interesting to note that in our study, the rated physical and social attractiveness of the targets are significantly correlated for male participants but not for female participants. This suggests that the role of physical appearance in social attraction may be greater among males than females, a finding that is consistent with the result of the meta-analysis by Feingold (1990), which indicated that while both sexes value physical attractiveness in members of the opposite sex, men appear to value it more than women. Moreover, our finding that females are not influenced by body size when rating the physical attractiveness of others is particularly intriguing in light of previous research on perceptions of desirable body shapes. These investigations have revealed that women strongly take their own body size into account when rating their own physical attractiveness and that they often report that being thinner would make them more attractive to members of the opposite sex (Fallon & Rozin, 1985; Pliner, Chaiken, & Flett, 1990; Rozin & Fallon, 1988). The results of the current study suggest that women appear to be able to dissociate the variables of attractiveness and body size, at least when they are rating others. With respect to the variable of social attractiveness, the main hypothesis for male targets was confirmed: overweight male targets were perceived, by both male and female participants, as significantly less socially attractive than normal-weight male targets. Additionally, it was found that overweight male targets who consumed large meals were rated by both male and female participants as significantly less socially attractive than any of the other male targets. Previous research findings have demonstrated that individuals are likely to attribute negative personality traits to targets who have been portrayed as being overweight, and those who are overweight are thought to lack self-control and will power (Crandall & Biernat, 1990; Harris & Hopwood, 1983; Harris et al., 1982; Maddox et al., 1968; Regan, 1996). Maddox et al. (1968) have demonstrated that individuals typically perceive obese targets as the least likeable and as most responsible for their ‘disability’ (i.e., being overweight) in comparison with other non-obese disabled targets. Our findings for the male targets were consistent with these stereotypes; overall, our overweight male targets were viewed more negatively. Moreover, it may be that our overweight small eater was perceived as attempting to control his obesity through the regulation of his eating behavior. Alternatively, he may have been seen as less responsible for his body size since he appeared to be eating lightly, possibly suggesting to participants that his obesity was not related to his eating habits. In contrast, our overweight large eater may have been perceived as lacking in self-control and will power, and as contributing to his obesity through his ‘ravenous’ appetite. Under these circumstances, the overweight large eater may be viewed more negatively than the overweight small eater is, which is exactly what we found.

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

129

For female targets, we expected to find a main effect for body size, revealing that overweight targets are perceived as less socially attractive than normal-weight targets. More interestingly, we found a threeway interaction involving participant sex, body size, and meal size. The interaction that emerged indicates that for male participants, normal-weight large eaters were perceived as more socially attractive than any other female target, whereas female participants rated the normal-weight small eater as the most socially attractive. This pattern of results is not entirely consistent with previous work in this area and suggests some interesting future research possibilities. Additionally, it demonstrates that although meal size does not have any independent effects on impressions of either male or female targets, it does moderate the effects of perceived body size on the impressions of others. As noted earlier, prior investigations in this area have found that both males and females rate small eaters as more socially attractive than large eaters, when both targets are normal-weight (Basow & Kobrynowicz, 1993; Mooney & Lorenz, 1997; Stein & Nemeroff, 1995), and in this study, the female participants’ ratings of the female targets’ social attractiveness are congruent with the results of these earlier studies. Our results for the male participants are discrepant with these findings but are consistent with the anecdotal evidence encountered in debriefing sessions with male participants from other investigations in our laboratory. When rating female targets, male participants rated the normal-weight large eater as more socially attractive than the normal-weight small eater. Thus, males appear to prefer hearty eaters, as long as they are not overweight. Future research should examine what underlies the pattern of results obtained by males rating female targets (on social attractiveness). For example, it may be that males’ impressions of the normal-weight large eater differ from their perceptions of the normal-weight small eater or the overweight targets. Perhaps, normal-weight females who eat heartily are perceived as more comfortable with themselves (both physically and socially), more confident, more honest (i.e., they are not attempting to hide their appetite), and less concerned with others’ impressions of them. In contrast, eating lightly may be viewed as behavior exhibited by a normal-weight individual who is overly concerned about their body shape/size or as a tactic designed to increase the likelihood that one will convey a positive first impression. Alternatively, it is plausible that males exhibit this preference for normal-weight large eaters simply because this intake pattern is more similar to their own pattern of intake (given that males typically consume more than females) and/or is less likely to make them (i.e., males) feel guilty if they themselves are large eaters. Clearly, this study does not provide us with a complete picture of the relationships between gender, body size, meal size and the perceptions of physical and social attractiveness. Our current research in this area aims to improve upon this work by using a more realistic body size manipulation and gathering more individual difference information about participants. More specifically, we have manipulated body size through the use of digitally altered photographs and have included a thin target to determine what effect, if any, a thin body has on our variables of interest. We believe that our decision to use digitally altered photographs, instead of the ubiquitous Stunkard et al. (1980) body figures, significantly improves upon the existing study because it vastly increases the external validity of our body size manipulation. In short, these photos are much more representative of the way that body size/shape information is conveyed in the real world. In addition, we are now collecting more data about our participants because it is possible that their characteristics may moderate the relationships among our variables of interest. For example, in this ongoing investigation, we are measuring the participants’ height and weight to calculate their body mass index (BMI). Similarly, we are assessing the participants’ perceptions of their own body weight (i.e., do they view themselves as thin, normal-weight, or overweight, etc.), weight concern, body dissatisfaction, and restrained eating. Doing this will allow us to determine whether any of these

130

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

variables moderate the participants’ impressions of the target’s physical and social attractiveness.7 We are also evaluating how similar the meal sizes portrayed in the study are to the participants’ typical intake for an equivalent meal. Most importantly, our ongoing research seeks to obtain more information about the individuals’ perceptions of the targets’ personality characteristics. If the findings presented herein are replicated, this additional information will allow us to examine whether the impressions of others’ personalities vary as a function of their eating behavior and body size, and will allow us to begin to understand why males prefer females of normal-weight who are large eaters. 4.1. Implications for eating behavior We believe these results have implications for eating behavior, particularly for women. As noted throughout the paper, overweight individuals often attract negative personality attributions such as laziness and lack of self-control (Cahnman, 1968; Jasper & Klassen, 1990; Maddox et al., 1968). In contrast, thin individuals are perceived as possessing such positive attributes as physical attractiveness, success, and power (Harris et al., 1982; Way, 1995). Thus, people’s bodies become vehicles for projecting these qualities (Brownell, 1991). With that knowledge, individuals’ regulation of their body weight by means of eating behavior may be used to serve self-presentational concerns. In other words, the desire to project a positive image may be one factor contributing to the popularity of dieting behavior, particularly among women (Leary, Tchividijian, & Kraxberger, 1994). The role of self-presentational concerns has been demonstrated in a program of research by Chaiken and Pliner, et al. In two studies, they showed that women ate less when eating with a desirable male than when eating with another woman; a third study revealed that women’s eating was suppressed when their femininity was threatened in the presence of a male partner (Chaiken & Pliner, 1987; Mori, Chaiken, & Pliner, 1987; Pliner & Chaiken, 1990). In these studies, it was assumed that impression management concerns would be greater with opposite- (vs. same-) sex partners and that these concerns would be served by reducing the amount consumed. A subsequent examination of the self-reported social motives underlying eating behavior confirmed these assumptions (Pliner & Chaiken, 1990, Study 2). Thus, for some females, one function of eating lightly in the presence of males is to increase their perceived social attractiveness. The data from the present study suggest that, for normal-weight women, eating lightly may not be producing the desired effect of increasing social attractiveness. If anything, our data suggest that women should eat more when they eat with males, but only if they can do so while maintaining a normal-weight. While this may be achievable for many females outside of the laboratory, for other groups, this finding has different implications and may be sending mixed messages. For example, obese females (in this study) were viewed less positively regardless of whether they ate small or large meals. For these individuals in the real world, the normal-weight large eating ideal may be unrealistic. First, they must lose weight (possibly not likely if they are genetically predisposed to obesity) and keep it off. Dietary restriction and exercise are often used as tools when attempting to control weight. Dietary restriction, 7 Although weight-concerned eaters are known to differ in some of their food-related perceptions, we did not assess these variables in the present study because prior research that has included measures of weight concern has found no effects of these variables (main effects or interactions) on the participants’ perceptions of the targets (Mooney & Lorenz, 1997; Stein & Nemeroff, 1995). Moreover, if there are effects of weight concern on the findings of the present study, these effects would be equally distributed across groups and, therefore, are not likely to be accounting for our findings.

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

131

however, is not a viable option for these individuals if they want to increase the likelihood that they will be viewed as socially attractive by males and, in turn, may lead to the adoption of maladaptive behaviors such as purging or excessive exercise. For women who are excessively thin because of restricted eating practices and/or voracious exercising, these results imply that they can eat more and exercise less, and that the adoption of such behaviors is beneficial to them both physically (in terms of their own health) and socially (at least in social situations involving males).

5. Conclusion Brownell (1991) has noted that the number of people who need to diet for health-related reasons is far smaller than the actual number of people who diet in America. Since the desire to convey a positive social image is one of the motives underlying restrained eating among women, future studies examining this phenomenon in normal-weight or excessively thin women should explore whether knowledge that this behavior is unsuccessful at achieving its goal alters food intake. Research on smoking, another behavior where self-presentation factors play an important role, shows that as those engaging in the behavior come to be increasingly stigmatized rather than admired, smoking rates have declined (Jeffrey, 1989). If such knowledge is successful at lessening women’s restrictive eating behavior, it could be used in programs aimed at preventing or overcoming anorexic behavior in those at risk.

References Basow, S. A., & Kobrynowicz, D. (1993). What is she eating? The effects of meal size on impressions of a female eater. Sex Roles, 28, 335 – 344. Bock, B. C., & Kanarek, R. B. (1995). Women and men are what they eat: The effects of gender and reported meal size on perceived characteristics. Sex Roles, 33, 109 – 119. Brownell, K. D. (1991). Personal responsibility and control over our bodies: When expectation exceeds reality. Health Psychology, 10, 303 – 310. Cahnman, W. J. (1968). The stigma of obesity. Sociological Quarterly, 9, 283 – 299. Chaiken, S., & Pliner, P. (1987). Women, but not men, are what they eat: The effect of meal size and gender on perceived femininity and masculinity. Personality and Social Psychology Bulletin, 13, 166 – 176. Crandall, C., & Biernat, M. (1990). The ideology of anti-fat attitudes. Journal of Applied Social Psychology, 20, 227 – 243. Eagly, A. H., Ashmore, R. D., Makhijani, M. G., & Longo, L. C. (1991). What is beautiful is good, but. . .: A meta-analytic review of research on the physical attractiveness stereotype. Psychological Bulletin, 110, 109 – 128. Fallon, A. E., & Rozin, P. (1985). Sex differences in perceptions of desirable body shape. Journal of Abnormal Psychology, 1, 102 – 105. Feingold, A. (1990). Gender differences in effects of physical attractiveness on romantic attraction: A comparison across five research paradigms. Journal of Personality and Social Psychology, 59, 981 – 993. Feingold, A. (1992). Good-looking people are not what we think. Psychological Bulletin, 111, 304 – 341. Harris, M. B. (1990). Is love seen as different for the obese? Journal of Applied Social Psychology, 20, 1209 – 1224. Harris, M. B., Harris, R. J., & Bochner, S. (1982). Fat, four-eyed, and female: Stereotypes of obesity, glasses, and gender. Journal of Applied Social Psychology, 12, 503 – 516. Harris, M. B., & Hopwood, J. (1983). Attitudes toward the obese in Australia. Journal of Obesity and Weight Regulation, 2, 107 – 120. Herman, C. P., & Polivy, J. (1980). Restrained and unrestrained eating. In A. J. Stunkard (Ed.), Obesity: Basic mechanisms and treatment ( pp. 208 – 225). Philadelphia, PA: Saunders.

132

Y. Martins et al. / Eating Behaviors 5 (2004) 117–132

Jasper, C. R., & Klassen, M. L. (1990). Perceptions of salespersons’ appearance and evaluation of job performance. Perceptual and Motor Skills, 71, 563 – 566. Jeffrey, R.W. (1989). Risk behaviors and health: Contrasting individual and population perspectives. American Psychologist, 44, 1194 – 1202. Leary, M. R., Tchividijian, L. R., & Kraxberger, B. E. (1994). Self-presentation can be hazardous to your health: Impression management and health risk. Health Psychology, 13, 461 – 470. Maddox, G. L., Back, K. W., & Liederman, V. R. (1968). Overweight as social deviance and disability. Journal of Health and Social Behavior, 9, 287 – 298. Mintz, L. B., & Betz, N. E. (1986). Sex differences in the nature, realism, and correlates of body image. Sex Roles, 15, 185 – 195. Mooney, K. M., DeTore, J., & Malloy, K. A. (1994). Perceptions of women related to food choice. Sex Roles, 31, 433 – 442. Mooney, K. M., & Lorenz, E. (1997). The effects of food and gender on interpersonal perceptions. Sex Roles, 36, 639 – 653. Mori, D., Chaiken, S., & Pliner, P. (1987). ‘‘Eating lightly’’ and the self-presentation of femininity. Journal of Personality and Social Psychology, 53, 693 – 702. Pliner, P., & Chaiken, S. (1990). Eating, social motives, and self-presentation in women and men. Journal of Experimental Social Psychology, 26, 240 – 254. Pliner, P., Chaiken, S., & Flett, G. L. (1990). Gender differences in concern with body weight and physical appearance over the life span. Personality and Social Psychology Bulletin, 16, 263 – 273. Polivy, J. P., & Herman, C. P. (1987). Diagnosis and treatment of normal eating. Journal of Consulting and Clinical Psychology, 55, 635 – 644. Regan, P. C. (1996). Sexual outcasts: The perceived impact of body weight and gender on sexuality. Journal of Applied Social Psychology, 26, 1803 – 1815. Rozin, P., & Fallon, A. (1988). Body image, attitudes to weight, and misperceptions of figure preferences of the opposite sex: A comparison of men and women in two generations. Journal of Abnormal Psychology, 97, 342 – 345. Serdula, M. K., Cikkubs, M., Williamson, D. F., Anda, R. F., Pamuk, E., & Byers, T. E. (1993). Weight control practices of U.S. adolescents and adults. Annals of Internal Medicine, 119(7 pt 2), 667 – 671. Stein, R. I., & Nemeroff, C. J. (1995). Moral overtones of food: Judgments of others based on what they eat. Personality and Social Psychology Bulletin, 21, 480 – 490. Stunkard, A. J., Sorenson, T., & Schlusinger, F. (1980). Use of the Danish adoption register for the study of obesity and thinness. In S. Kety (Ed.), The genetics of neurological and psychiatric disorders ( pp. 115 – 120). New York: Raven Press. Tiggemann, M., & Rothblum, E. D. (1988). Gender differences in social consequences of perceived overweight in the United States and Australia. Sex Roles, 18, 75 – 86. Wheeler, L., & Kim, Y. (1997). What is beautiful is culturally good: The physical attractiveness stereotype has different content in collectivist cultures. Personality and Social Psychology Bulletin, 23, 795 – 800. Way, K. (1995). Never too rich. . .or too thin: The role of stigma in the social construction of anorexia nervosa. In J. Sobal, & D. Maurer (Eds.), Eating agendas. New York: Aldine de Gruyter.