Appetite 86 (2015) 88–95

Contents lists available at ScienceDirect

Appetite j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / a p p e t

Research report

“She got more than me”. Social comparison and the social context of eating Janet Polivy *, Patricia Pliner University of Toronto, Canada

A R T I C L E

I N F O

Article history: Received 26 June 2014 Accepted 5 August 2014 Available online 13 August 2014 Keywords: Social comparison Restrained eaters Social influence Body image Amount eaten

A B S T R A C T

Eating is a social activity for most people. Other people influence what and how much an individual chooses and eats. Such social influence on eating has long been recognized and studied, but we contend here that one important social influence factor, social comparison, has been largely overlooked by researchers. We review the literature on comparing oneself to others on eating and weight-related dimensions, which appears to have an effect not only on eating per se, but also on self-image, body dissatisfaction, and emotions. Social comparison processes may well underlie many of the social influence findings discussed in this special issue. © 2014 Elsevier Ltd. All rights reserved.

Introduction Although much of the research investigating eating behavior examines the behavior of individuals eating alone in a lab, most actual eating outside of a laboratory setting involves people eating with other people, usually people whom they know (Rozin, 1996). Even when “meals” of only 50 calories (i.e., snacks) are included, most meals are eaten with at least one other person (de Castro & de Castro, 1989; Redd & de Castro, 1992). A US survey reported that only one third of people reported eating alone during the week and even fewer claimed to eat alone on the weekends (Rodrigues & Almeida, 1996). Thus, human eating is generally a social activity. In fact, according to Sobal and Nelson (2003), “Eating alone is devalued and is not considered a ‘real’ meal for many people” and furthermore, “almost all people (who were surveyed) thought that an ideal meal should be eaten with the company of others” (p. 182). When Redd and de Castro (1992) asked people to eat their meals as they normally would with respect to eating companions, about two thirds of the meals were eaten with other people. Young adults report that they prefer to eat with others, although they do not always have time to do so (Larson, Nelson, Neumark-Sztainer, Story, & Hannan, 2009). Eating with others The presence of others has been shown to affect what and how much people order or serve themselves and what and how much

* Corresponding author. E-mail address: [email protected] (J. Polivy). http://dx.doi.org/10.1016/j.appet.2014.08.007 0195-6663/© 2014 Elsevier Ltd. All rights reserved.

they consume and thus exerts a powerful effect on their behavior; hence, the current special issue of Appetite. In their systematic review of eating and social influence, Herman, Roth, and Polivy (2003) identified three separate sub-literatures, all demonstrating that people eat differently in the company of others: social facilitation, impression management, and conformity or modeling. Research on the social facilitation of eating shows that people eating with others generally eat more food than those eating alone (e.g., de Castro & de Castro, 1989). Additionally, it has been shown that the more people present, the more each eats (e.g., Hetherington, Anderson, Norton, & Newson, 2006), possibly because a meal with many eaters tends to have a longer duration (Pliner, Bell, Hirsch, & Kinchla, 2006). (See Herman’s review in this issue.) As for impression management, Mori, Chaiken, and Pliner (1987) demonstrated that when women are motivated to make a positive impression, they eat less than when they are not. And it is not simply the quantity of food consumed that is affected by impressionmanagement concerns, but also the type of food selected. For example, women eating in a dyad with a male eat foods that are lower in calories than women eating with another woman, and women’s caloric intake in general is negatively related to the number of men in the group (Young, Mizzau, Mai, Sirisegaram, & Wilson, 2009). It is assumed that what underlies this impression-management effect is women’s desire to behave in accordance with social norms regarding appropriate levels of food intake so as to garner social approval. (See Vartanian’s review in this issue. Vartanian also discusses impression management in men.) In addition, there is a large literature showing that people tend to eat more or less depending on the amounts eaten by their eating companions (e.g., Herman et al., 2003; Robinson, Blisset, & Higgs,

J. Polivy, P. Pliner/Appetite 86 (2015) 88–95

2013; Roth, Herman, Polivy, & Pliner, 2001). It is thus abundantly clear that what and how much individuals eat are strongly influenced by the people around them. (See Cruwys, Bevelander, and Hermans’s review in this issue.) In the present paper, we will focus on the effects on eating and other behaviors of another social process – namely, social comparison. Although it is rarely discussed, social comparison plays a more important role in eating than is ordinarily assumed. To begin, we contend that at least two of the other social influence processes described above ultimately stem from social comparisons. Both modeling and impression management imply an initial social comparison, with the individual comparing his/her own eating either to an existing abstract norm or ideal or to the actual current or previous behavior of some other individual(s). So, social comparison is a prerequisite for other forms of social influence to occur. However, there are also social comparison effects that cannot be neatly described as the precursors to modeling and impression management, and it is those effects that constitute our main focus in this paper. In either case, the basic tenets of social comparison theory, as described below, apply. Social comparison and eating Historically, social comparison was considered to be the outcome of a basic and ubiquitous human drive to evaluate the correctness of one’s opinions and the “goodness” of one’s abilities (Festinger, 1954). The study of social comparison soon expanded to include the evaluation of other personal qualities and behaviors, including emotions (Schachter, 1959). Social comparison theory posits that in the absence of concrete, physical standards against which to measure these characteristics, people rely on comparison of their own characteristics with those of other people (Festinger, 1954). The earliest studies generated situations in which the “need” for self-evaluation was aroused, often by providing the individual with ambiguous feedback about his or her standing on some valued characteristic. These studies focused on the question of with whom (which person or group) the individual chose to compare. Subsequent research, including much recent work, created situations in which a comparison was foisted on an individual and examined the behavioral and affective outcomes of such comparisons. Although Festinger (1954) emphasized that accurate evaluation required comparing oneself with a similar other, he also acknowledged that comparisons could occur with better-off or superior others (upward comparisons) or with worse-off or inferior others (downward comparisons). In terms of the affective consequences of such upward or downward comparisons, the simplest view was that upward comparisons would result in negative feelings and decreased self-esteem whereas downward comparisons would result in positive feelings and enhanced self-esteem. Basically these are contrast effects in which, one’s own standing appears worse than that of a superior other and better than that of an inferior other (e.g., Tesser, 1991). It is now widely accepted that both upward and downward comparisons can produce both positive and negative effects (e.g., Buunk, Collins, Taylor, VanYperen, & Dakof, 1990). For example, Lockwood and Kunda (1997) note the often-positive effects on individuals of successful role models or “superstars.” According to their analysis, superstars promote self-enhancement and inspiration provided that they are perceived as relevant to the self and, importantly, provided that achieving a comparable success seems attainable; however, they elicit self-deflation and discouragement when a comparable success seems unattainable. Thus, if a relevant superstar’s achievements appear attainable, the individual assimilates to or literally feels similar to that superstar, making the individual feel better; however, if the achievements do not appear to be attainable, the contrast between that other and the self makes one feel worse. Examining downward rather than upward comparison, Wood, Taylor,

89

and Lichtman (1985) found that some breast cancer patients found downward comparisons threatening, presumably because they instantiated the idea that things might well become worse for them. Similarly, Markus and Nurius (1986) talk about feared possible selves. In this article, we examine social comparison in the context of eating. We identify two areas in which social comparison can occur in the context of eating and influence an individual. Specifically, we examine studies in which 1) social comparison occurs on a dimension related to eating (i.e., body weight) and, if food is available, affects eating, and 2) social comparison occurs in terms of amount of food eaten or supplied and affects other feelings and/or behaviors (which may include eating behavior). Social comparison on dimensions related to eating or body weight In this section, we will discuss what happens when one compares oneself to others on an eating- or weight-related dimension such as physique, and then is given an opportunity to eat. To understand this phenomenon, we will begin with a brief review of a voluminous literature on the effects of mass media exposure to thin images on body satisfaction and related variables. A meta-analysis by Groesz, Levine, and Murnen (2002) examined experimental studies that compared exposure to magazine or video images of thin models with control exposure to average or heavy models or to objects such as cars or houses, and assessed women’s body satisfaction or self-rated attractiveness. Their main finding was that body image was significantly more negative after women viewed thin images than after they viewed control images (although a few studies found the opposite). Examining moderating variables, Groesz et al. (2002) found that younger women as well as women who had significant body image issues were more adversely affected by the media exposure. In discussing their findings, the authors noted that social comparison is a construct that has received insufficient attention in understanding these results. A more recent meta-analysis by Myers and Crowther (2009) cast a broader net by examining both experimental and correlational studies and tested the hypothesis that social comparison with thin others, directly or as represented in media images, is associated with body dissatisfaction. While the correlational studies typically use self-report measures to assess social comparison, the experimental studies are more likely to simply assume that exposure to thin people (or their images) induces social comparison; however, a few experimental studies do ask participants to report whether they had made such comparisons. The overall effect size was .77 – quite a large effect size by Cohen’s (1992) criterion. Examining moderator variables, Myers and Crowther (2009) found a larger effect size for studies that measured (vs. inferred) social comparison, for younger (vs. older) participants, and for women (vs. men). Thus, social comparison is a construct that researchers have employed to understand the effects of exposure to thin images on individuals’ body satisfaction. It seems that social comparison does occur and that the comparison is associated (in most cases) with negative effects on people’s satisfaction with their bodies. If we assume that thin comparison others are considered to be superior others (as they would be according to the Duchess of Windsor’s dictum that “one cannot be too thin. . .”), then individuals are showing contrast effects. That is, people are making upward comparisons, and from the perspective of the Lockwood–Kunda approach, the success of the comparison other is not seen as attainable. These two meta-analyses are relevant for our purposes because they provide background for the description of a much smaller set of studies assessing actual eating after exposure to thin images. If we assume that social comparison processes are engaged by such exposure, then any effect on eating could be considered to be a direct or indirect outcome of such comparison.

90

J. Polivy, P. Pliner/Appetite 86 (2015) 88–95

Before we describe the relevant studies, we must acknowledge that it is difficult to make perfect sense of them all given how few of them there are and how much they differ on various dimensions. We cannot say with any certainty what contributed most to their varied empirical outcomes. We have chosen to describe them in terms of the following: 1) medium for presenting thin images (television ads, magazine ads, etc.); 2) control images (images of average weight people, images of overweight people, images of objects, etc.); 3) characteristics of thin images (model, peer, etc.); 4) any other manipulations; 5) context of eating (taste test, incidental eating, etc.). Because nearly all the studies assessed dietary restraint and most tested its effects as a moderator of the relation between image exposure and eating, we have indicated which measure of dietary restraint was used (Herman–Polivy RS (Herman & Polivy, 1980); DEBQ (Van Strien, Frijters, Bergers, & Defares, 1986); TFEQ (Stunkard & Messick, 1985)). We have also described other dependent variables that were intended to assess the “psychological” effects of the exposure manipulations (e.g., body satisfaction, mood). Eating after exposure to thin female images: images embedded in movies We begin with five studies employing roughly similar methodology and designed in part to investigate a “re-inhibition” hypothesis: namely, the idea that exposure to thin images (under the assumption that exposure would cause women to compare themselves to these images) might enable or motivate restrained eaters whose eating had been disinhibited by some factor such as a preload or emotional arousal to bring their eating back under control. Strauss, Doyle, and Kreipe (1994) offered snacks to restrained and unrestrained (using the Herman–Polivy RS) women as they viewed a sad movie following what was described as a high calorie preload. For some participants, ads with thin models advertising diet products were embedded in the movie; for others, the ads were neutral; and for the remainder, there were no ads. Contrary to the “re-inhibition” hypothesis, restrained eaters ate more in the thin/diet ads condition than in the other two conditions, whereas unrestrained eaters were unaffected by ad type. Although all participants showed an increase in MAACL depression from before to after the movie, there were no effects involving restraint status or ad type. Warren, Strauss, Taska, and Sullivan (2005) more or less replicated the Strauss et al. (1994) study, albeit without the preload and with just thin/diet ads and neutral ads and employing high school students as participants (in contrast to most of the other studies described in this section, which employed college students). They found that the restrained eaters ate more in the thin/diet ad condition than participants in the other three conditions, who ate equivalent amounts. Participants felt sadder after the movie, irrespective of restraint status or advertisement condition. Three studies by Anschutz and colleagues also employed a snacking-while-watching-a-movie paradigm. In the first (Anschutz, Engels, Becker, & van Strien, 2008), restrained and unrestrained women (assessed by the DEBQ) snacked while they watched a cheerful movie with beautiful characters. In one condition the movie was shown in a normal screen format and the characters appeared thin; in the other, the movie was shown in a broad screen format and the body size of the actresses was slightly stretched breadthways, making them appear heavier. Restrained eaters ate more when the actresses appeared thinner whereas unrestrained participants ate more when the actresses appeared heavier. Further, although mood was not differentially affected by the screen manipulation, restrained eaters showed greater body dissatisfaction in the thin body condition than in the heavier body condition, whereas unrestrained participants showed the opposite pattern. Anschutz, Van Strien, and Engels (2008) exposed restrained and unrestrained eaters

(DEBQ) to either a sad movie or a neutral movie in which ads with thin models promoting diet products or neutral ads were embedded. In this study, when offered snacks, restrained eaters ate less when they were exposed to the thin/diet ads, whereas unrestrained eaters ate more. Although the mood measure showed the expected effects of movie type, there were no interactions with restraint or ad type. In the third study in this series, Anschutz, Engels, Becker, and van Strien (2009) showed restrained and unrestrained (DEBQ) eaters a neutral movie in which there were embedded neutral ads (primarily product ads with no female characters), ads for Nivea beauty products with thin models, or ads for Dove products with normal or slightly oversized models and with an “. . . explicit focus on the use of realistic models.” In this study, there was a main effect of ad type; all participants ate more when they viewed the Nivea (thin) ads than in either of the other two conditions; restraint did not interact with ad type. In addition, ad condition had a significant main effect on sadness; participants in the Dove condition felt sadder than did participants in the Nivea condition after watching the movie. They authors note that they “. . . checked whether sadness mediated the relation between commercial exposure and food intake, but this appeared not to be the case” (p. 274). Summarizing the food intake results of these five studies in the simplest possible “box score” terms, three showed interaction effects such that restrained eaters ate more during exposure to thin images whereas unrestrained eaters showed no effect or the reverse effect. One study found that restrained eaters ate less during exposure to thin images and unrestrained participants ate more. Finally, the last study found that both restrained and unrestrained eaters ate more while viewing thin images. Eating after exposure to thin female images: television commercial images Strahan, Spencer, and Zanna (2007) conducted a set of studies testing the hypothesis that women would eat less after exposure to thin images. This hypothesis was based on the notion that thin media images convey the message that thinness is highly valued in society and women would feel that one means of attaining a thin body is to limit how much food they eat. Under the guise of a memory task, participants viewed a set of neutral commercials, advertising products such as cell phones, or the same set of commercials plus other commercials featuring thin models advertising products such as lingerie and beauty products. They then participated in an “unrelated” taste test. In three studies, their hypothesis was confirmed; women ate less after exposure to the thin images, unless the cultural norm in favor of thinness was experimentally undermined. In these studies, the interactive effect of dietary restraint was not assessed. In a study by Seddon and Berry (1996), restrained and unrestrained eaters (Herman–Polivy RS) viewed either television ads with thin images or matched ads that did not contain such images. Participants high in dietary restraint who viewed the video with thin images ate more in a subsequent taste task. There were no effects on state self-esteem as a function of the image manipulation, restraint, or their interaction. Eating after exposure to thin female images: magazine images In another set of studies, participants viewed magazine ads. Mills, Polivy, Herman, and Tiggemann (2002) had restrained and unrestrained eaters (Herman–Polivy RS) view ads with attractive thin or plus-size models or product-only ads and then participate in a “taste test” during which their intake was surreptitiously measured. Restrained eaters ate more after the thin ads than after the other two types, whereas there was no effect of ad type on unrestrained eaters. There were no ad type or restraint status effects on

J. Polivy, P. Pliner/Appetite 86 (2015) 88–95

mood; however, restrained eaters had marginally higher state appearance self-esteem and reported themselves to be thinner following the thin (vs. the large-body) ads. It should be noted, however, that restrained eaters rated themselves as more similar to the heavy (vs. thin) models, whereas unrestrained participants felt equally similar to the two types. Monro and Huon (2006) showed their participants either magazine ads with thin images or the same ads with the images digitally removed and then staged a “taste test.” These authors did their analyses using self-objectification rather than restraint as a moderator, but because these two measures were highly correlated (r = .60 with the Herman–Polivy RS) we may assume that participants high in self-objectification were also high in restraint. High self-objectifiers ate more when the thin images were present, whereas low self-objectifiers ate more when the thin images were absent. These authors did not include measures of body dissatisfaction or mood in their study. In a study by Boyce, Kuijer, and Gleaves (2013), participants watched a slide show with either thin models advertising products such as perfume or make-up or the same advertisements with the models digitally removed. Restraint was assessed by means of the Concern for Dieting subscale of the Herman–Polivy measure. There was no effect of ad condition, restraint or their interaction on food intake. However, ad condition and restraint interacted significantly on the measure of negative mood such that participants higher in restraint had a more negative mood when thin images were present than when they were absent, whereas participants lower in restraint showed the reverse pattern. Similarly, on a measure of weight satisfaction, participants higher in restraint had lower weight satisfaction when thin images were present, whereas participants lower in restraint showed the reverse pattern. Durkin, Hendry, and Stritzke (2012) showed their restrained and unrestrained (Herman–Polivy RS) eaters sets of magazine photographs of either thin or overweight models eating chocolate, accompanied by either negative or positive statements about chocolate consumption (e.g., “Chocolate: good mood food!” “Chocolate: a moment on the lips, forever on the hips”). After rating the “likely impact” of each ad, participants were invited to help themselves to chocolate. Restrained eaters ate more chocolate in the thin (vs. overweight) model condition, whereas unrestrained eaters ate similarly in the two conditions. There were no effects of message valence. To summarize this set of studies, three of four found that restrained (or self-objectifying) participants ate more when exposed to thin (vs. control) images whereas unrestrained eaters showed either no effect or the opposite effect. Only the Boyce et al. (2013) study found no effects of restraint, exposure to thin images, or their interaction on food intake. In the two studies in this group that assessed mood and/or weight satisfaction, one showed positive effects of thin exposure on restrained eaters whereas the other showed negative effects; both found that restrained eaters ate more after thin exposure. Eating after exposure to thin female images: “real” people We turn now to a set of studies whose methodology is diverse but which have in common the fact that participants were exposed, not to images of models or actors, but to “ordinary” people. Trottier, Polivy, and Herman (2007) created a scenario in which restrained and unrestrained eaters read information about a peer who was thin, of average weight, or overweight; they were informed that they would be required to recall the information later in the experiment. They then participated in a “taste test” supposedly as a distracter prior to the memory task. There were no differences in amount eaten as a function of restraint, peer weight, or their interaction. Restrained eaters in the thin peer condition, however, perceived themselves as more overweight and had higher body dissatisfaction and lower appearance self-esteem than did those in the

91

average-weight condition. Green and Saenz (1995) used a videotape to allow restrained and unrestrained (Herman–Polivy RS) eaters to view a peer with whom they would be remotely interacting later in the session. The peer was either extremely well-dressed in formfitting clothes, with a fashionable hairdo and make-up, or dressed in shabby, baggy clothes, made to look overweight, not wearing makeup, and with her hair sloppily done. In addition, participants received written information confirming the peer’s weight and also manipulating a third independent variable, the peer’s academic performance, which was described as high or low or was not described. A taste test followed. Restrained eaters ate less after viewing the thin/attractive peer than after viewing the heavy/unattractive target, whereas the reverse was true for unrestrained eaters. In addition, restrained eaters had more negative mood after exposure to the heavy/unattractive target than after exposure to the thin/attractive target, whereas unrestrained participants showed no mood effect. After computing a mediational model, these authors suggest that for restrained eaters, the social comparison leads to negative affect which in turn leads to decreases in dieting self-efficacy which in turn leads to increased eating. Thus, these authors are suggesting that restrained participants are making a threatening social comparison with the heavy/unattractive target and the resulting negative affect increases their eating. Another interpretation would be that they are making a self-enhancing social comparison with the slim/ attractive target and are inspired to eat less. Reasoning from a literature demonstrating that people assimilate behaviorally to an activated stereotype (e.g., Dijksterhuis & Bargh, 2001), Campbell and Mohr (2011) hypothesized that people would eat more after being primed with a photograph of an overweight person. Researchers claiming they were “calibrat[ing] pictures for later studies” approached participants on campus, showed them a photograph of an overweight person, a person of normal weight, or a lamp, and then invited them to help themselves from a bowl of candy as a reward. In five studies, participants (unclassified as to restraint status) ate more after an overweight prime than after a control prime. Similarly, McFerran, Dahl, Fitzsimons, and Morales (2010a, 2010b) conducted a simple study in which restrained and unrestrained (unspecified measure) eaters took part in a taste test and which manipulated the perceived weight of the person serving the food. This manipulation was accomplished by having the normally thin server wear a professionally constructed “obesity prosthesis” as she served half the participants. In a perfect crossover interaction, restrained eaters ate less after receiving their food from the thin server, whereas unrestrained eaters showed precisely the opposite pattern. There were no measures of mood or body dissatisfaction. Eating after exposure to thin female images: explaining the findings We can now summarize the 21 studies that we have just reviewed in terms of the effects of exposure to thin (vs. control) images on eating in restrained and unrestrained women. To provide a box score synopsis: seven studies found that restrained eaters ate more following exposure to thin (vs. control) images, with unrestrained eaters showing either no effect or the opposite effect; three studies found that restrained eaters ate less following exposure to thin (vs. control) images, with unrestrained eaters showing either no effect or the opposite effect; two studies found that restrained (and unrestrained) eaters ate equivalent amounts following exposure to thin (vs. control) images; and one study found that both restrained and unrestrained participants ate more after exposure to thin (vs. control) images. The remaining eight studies did not analyze their data with restraint as an independent variable; all found that participants ate less after exposure to thin (vs. control) images. Thus, when restraint is not included as a variable, the modal finding is that

92

J. Polivy, P. Pliner/Appetite 86 (2015) 88–95

individuals decrease their intake following exposure to thin images; but when restraint is included as a variable, the modal finding is that the intake of restrained eaters, in contrast to that of unrestrained eaters, increases after exposure to thin images; of 13 studies in which restraint was included as a moderator variable, seven produced this finding (and an eighth found that everyone ate more when exposed to thin images). So, is there any pattern or consistency in the data? As was noted above, the modal finding of the studies in which restraint was included as a variable is that restrained but not unrestrained eaters increase their intake following exposure to thin images, with one additional study finding that restrained and unrestrained eaters both increase their intake following exposure to thin images. Let us then focus on the five studies in which restrained eaters did not increase their intake following exposure to thin images. What three of these studies have in common in contrast to all of the remainder is that the thin images were not described as media images or images of models. So, restrained eaters increased their intake when the thin images they were exposed to were images of beautiful and glamorous models and actresses but were less likely to do so when they were not so described. This generalization is extremely tenuous, and it would be undermined by even one or two additional studies that are not consistent with it. However, in the absence of disconfirming data, let us proceed as though it were correct. It is possible in the latter cases (when the images were not of media images of models or actresses), participants were much more likely to feel that thinness was, in Lockwood and Kunda’s (1997) terms, attainable, resulting in assimilation. Thus, we may suggest that when restrained eaters are exposed to others who are thin, and thinness appears to be attainable and/ or they feel similar to those thin others, they eat relatively sparingly. When thinness appears to be unattainable and/or they do not feel similar to those thin others, they increase their intake upon exposure to thin others. Are these findings consistent with what we know about social comparison? We begin with the question of why restrained eaters might be more affected by such comparisons than unrestrained eaters would be. Much of the research on social comparison processes suggests that people are more likely to compare themselves with others on dimensions that are personally relevant or important to them (Papies & Nicolaije, 2012). Pliner, Chaiken, and Flett (1990) found that concern with weight and appearance were high in women in general, and Cachelin, Striegel-Moore, and Paget (1997) showed that they were even higher in women who were high in restraint. Now, let us consider why women (and in particular restrained eaters) might potentially eat more after viewing thin images. First, we assume that exposure to a thin image constitutes an upward social comparison. That is, in contemporary North American culture, a thinner body is a better body. One possible outcome of such a comparison would be a contrast effect in which one’s own status appears worse after exposure to the better-off other, and one would experience negative affect and lowered self-esteem. One of the most robust findings in the literature on restraint is that negative affect disinhibits the eating of restrained eaters (Heatherton, Striepe, & Wittenberg, 1998; Polivy & Herman, 1999); hence, following an upward (and unattainable) social comparison, restrained eaters would be expected to eat more. Eating by unrestrained eaters would be expected to be unaffected or less affected for two reasons: 1) weight is a less relevant dimension for them and 2) unrestrained eaters’ eating is not disinhibited by negative affect. On the other hand, what if social comparison with a thin person produced an assimilation effect, in which the individual felt thinner and experienced positive affect and enhanced self-esteem? This comparison might be inspirational and might increase the dieter’s motivation to stick to his/her diet and cause him/her to eat less in the experimental situation.

Returning to the data, we look next to see whether there is any hint that the three studies in which the thin images were not of models and/or actresses (and in which eating in restrained eaters did not increase) produced assimilation rather than contrast effects. This we can do by looking to see whether the measures of selffeelings are consistent with assimilation (positive mood, selfenhancement) or contrast (negative mood, self-disparagement). Unfortunately, the data do not help us here – in one study (Green & Saenz, 1995) participants reported more positive self-feelings following exposure to the thin peer; but in a second (Trottier et al., 2007), participants reported more negative self-feelings, and in the third, there were no measures of self-feelings. So, we are left with a tenuous generalization with nothing to back it up. What we do have, however, is strong evidence that the effects of social comparison with respect to body weight/shape can and often do affect eating behavior, especially in chronically-dieting restrained eaters. Social comparison of food provided and/or eaten We turn now to social comparisons that occur in terms of the amount of food provided to an individual or the amount of food consumed by an individual. How do such social comparisons affect feelings and/or behaviors (which may include eating behavior)? For example, children frequently complain that a sibling received a bigger portion of a favorite food than they did. In such cases, the effect of the comparison might affect how the child feels (e.g., aggrieved, harddone-by, less-favored, angry, or sad) and might also influence the child’s subsequent eating behavior (e.g., he/she might eat more of something else, feeling deserving of compensation with more of another food). We contend and will show that it is not only children who compare their own plates to those of others. This special issue is devoted to examining the effects of social influences on eating behavior, often with respect to future consequences of what and how much we eat when with others. We contend, however, that at least some of these social influences stem purely from social comparisons of our food with someone else’s food at the moment when eating is occurring. We will thus examine some research on how these immediate social comparisons regarding food affect our own feelings and behaviors. Modeling/matching the intake of co-eaters It has been argued convincingly that others set the norms that we use to decide how much to eat in a given situation (e.g., see Herman et al., 2003; Robinson et al., 2013; Robinson, Thomas, Aveyard, & Higgs, 2014; see reviews by Higgs and by Cruwys, Bevelander, & Hermans in this issue). There is substantial evidence that people model the eating behaviors of co-eaters, eating more when the model eats more and less when she/he eats less, especially in situations that are in any way ambiguous (Herman et al., 2003). Modeling of another’s eating behavior will be reviewed thoroughly in another paper in this volume (Cruwys et al.), but there are some aspects of modeling effects on eating that we believe are relevant to our thesis that comparing one’s own food to what others have or eat has a significant influence on our own eating behavior and feelings about ourselves, so we will discuss these briefly. Hermans, Larsen, Herman, and Engels (2008) found that modeling occurred in normal-weight young women only if the model’s appearance was also “normal weight” (as opposed to being visibly thin). As Cruwys et al. pointed out earlier in this issue, the extent to which the model is seen as similar to the self is a significant moderator of the degree of modeling that will occur. This finding of differentially strong or weak responses to how much a model is eating, however, relies on one’s attention to aspects of the model and on comparing oneself to the model in order to decide whether or not the model is an appropriate person upon whom to base one’s own

J. Polivy, P. Pliner/Appetite 86 (2015) 88–95

behavior. So modeling occurs more or less strongly because people compare themselves to other people and choose which ones to “follow.” Obviously, this is social comparison, not simple copying of another person’s behavior (although people may be unaware that they are using such comparative information to control their own eating – see Vartanian, Sokol, Herman, & Polivy, 2013 or Cruwys et al. in this issue). We may be more inclined to compare ourselves to others under some conditions than others and coordinate our eating to the other only when the other person seems relevant, important, or similar to us (as, for example, when she announces that she is on a diet (Polivy, Herman, Younger, & Erskine, 1979)). We may also be more likely to compare ourselves to those who are similar to us. Likewise, our state of mind may determine the extent to which we compare our own intake to the eating behavior of others. Florak, Palcu, and Friese (2013) showed that inducing a preventive regulatory focus (as opposed to a promotion focus) increased the degree to which participants matched the eating intake of others. It seems, then, that we compare ourselves more to certain people (e.g., other dieters, other overweight peers) and when we are in a particular state of mind (e.g., prevention focused). Effects of the presence of others when we eat The mere presence of another person in the room when one is eating has an effect on the eater (Herman et al., 2003). For example, if the other person is not eating, but simply observing, the eater eats much less (e.g., Roth et al., 2001). Being watched while we eat seems to minimize how much we are willing to consume, possibly because we feel self-conscious, as if our eating becomes excessive when compared to that of another person who is not eating at all. The influence of the presence of others on eating behavior thus may well reflect a social comparison process. For example, overweight girls who eat with a peer eat very differently if that peer is normal weight versus overweight; they eat significantly more when eating with an overweight peer, especially if the peer also eats a lot (Salvy, Romero, Paluch, & Epstein, 2007). Normal-weight girls were unaffected by the weight of their co-eater, so only overweight girls seem to compare themselves to their eating companion. This influence is especially strong if the co-eater is a friend (because everyone eats more when eating with friends than when eating alone); overweight girls eat substantially more with an overweight friend than with a normal-weight friend, although they still eat more even with a stranger if she is overweight than if the coeater is not overweight (Salvy, Howard, Read, & Mele, 2009). As we discussed earlier, people tend to compare themselves more to similar others, and emulate their behavior when it reflects a goal that is attainable and appealing. If an important eating goal (especially for restrained eaters and overweight individuals) is to eat as much as possible without seeming to eat excessively (Herman et al., 2003), then overweight girls would be more inclined to compare themselves to and increase their intake when they are with their overweight peers. Social comparison-induced contrast/context effects of food People tend to look at each other while engaging in social interactions such as eating together (Wu, Bischof, & Kingstone, 2013) and thus are likely to notice what each other is eating. We compare our eating options to those of others, and this can make us more or less satisfied with our own options. In a recent study, participants were given a standard slice of pizza as a “light lunch” before a taste perception study in which they would share their opinions with another (actually nonexistent) participant (Polivy, Herman, & Deo, 2010). Although all participants were given a slice of pizza of the same size, one third of them saw a smaller slice about to be

93

given to the fictitious other participant, one third saw a larger slice going to the other, and the remaining one third received no information about the amount the other was to receive. So although all actual participants received a slice of pizza of the same size, because of its comparative size one third of them thought that they had been given a “big” piece and one third thought that they had received a “small” piece. Restrained and unrestrained eaters reacted quite differently to this perceived difference in portions; unrestrained eaters reported more (PANAS) negative affect if they got a larger piece than if they got a smaller piece. Restrained eaters, however, reported more negative affect than did unrestrained eaters if they got a smaller piece of pizza (with a trend in the opposite direction for those receiving the “larger” slice). We interpreted these results as suggesting that unrestrained eaters did not want to overeat and were unhappy if they perceived that they had received a large piece of pizza; in contrast, restrained eaters were disappointed if they thought they had missed out on an opportunity to (be required to) eat more of a favorite food. Moreover, in a subsequent taste-rating task, restrained and unrestrained eaters behaved quite differently when the pizza slice had been perceived as large (although not if they saw it as small or got no information about the other person’s slice). Unrestrained eaters tended to eat fewer cookies if they thought they had eaten more pizza whereas restrained eaters ate more cookies. Thus, comparing their portion to someone else’s had significant (and opposite) effects on how restrained and unrestrained eaters felt and behaved. When they compared their portion to that of the other participant and thought they had eaten more than she did, restrained eaters were disinhibited and ate more cookies. Interestingly, however, their affect was actually less negative in this condition than when they thought that they were getting less to eat than the other person, suggesting that they may have felt short-changed when given the smaller piece even though it left their diets intact, and might conceivably have been pleased to have their diets broken by the experimenter, allowing them to eat more of their preferred foods. In a recent as yet unpublished study from our lab, we allowed students the opportunity to in effect correct the inequity in portions given to them by making more pizza available after they had eaten their original slice. Male and female restrained eaters who were led to believe that they had received the “smaller” slice compensated for having been “short-changed” by eating more of the “extra” pizza available to them than did unrestrained eaters. In a conceptual replication of this study, we also found that participants given a vegetarian sub sandwich ate very differently depending on what they believed another participant was getting to eat. Restrained eaters who thought the other participant was getting a more preferred meal (pizza) ate less of their sub sandwich than the other participants did, and both restrained and unrestrained eaters ate more sandwich than the control group did when they thought the other person was given a less preferred meal (plain cheese sandwich). Moreover, participants reported liking the sub less when they thought another participant was getting pizza than if they thought that they were both eating subs, although only unrestrained eaters increased their ratings of the sub when they thought the other person was eating a cheese sandwich. Comparing our food to what others are getting to eat thus affects both how much we like our own meal and how much we eat. When we compare our food to what others are eating, downward comparisons (i.e., the other person’s food is worse than ours is) allow us to value our meal more highly and thus encourage us to consume more of it, whereas upward comparisons have the reverse effect. Moreover, if we can compare our food to that of others, and emulating them is desirable and “attainable” (i.e., we can increase our portion size to match that of another eater), we will change our intake accordingly (especially if we are dieters who are highly concerned with how much we and those around us are eating).

94

J. Polivy, P. Pliner/Appetite 86 (2015) 88–95

Social comparison of actual amounts eaten We have described some of the consequences of comparisons of the quantity and nature of the food received by another individual with the food received by the participant himself or herself. We now describe two studies in which researchers examined the consequences of social comparisons of the amount actually eaten by another individual and the amount eaten by the participant herself. In both studies, female participants found themselves in a situation that was ambiguous with respect to the appropriate amount to eat; nevertheless, they were encouraged to eat and did so. Subsequently, they received experimenter-generated information about how much was eaten by a fictitious fellow participant who was tested at the same time as they were. In Pliner et al.’s study (unpublished manuscript, 2009), the fictitious other participant had eaten either half as much as the real participant or had eaten the same amount. We assume that the former case would constitute an upward comparison, as eating lightly is met with approval and is considered to be a sex role – appropriate behavior for women (Chaiken & Pliner, 1987). Furthermore, because the eating had already occurred, there was no way for the participant herself to attain the status of being a “light” eater in that situation; indeed, in relative terms, the participant was now a “heavy” eater. Thus, one would expect that participants would experience the self-deflation and discouragement that accompany an upward social comparison when attainability is low. The results showed exactly that; participants who had been “undereaten” by their fellow participant showed lower ratings on an omnibus measure of self-positivity which included a measure of self-esteem and selfratings of likeability. Using a similar methodology, Leone, Herman, and Pliner (2008) created a situation in which participants believed they had eaten either twice as much or half as much as their fictitious partners had eaten. Once again, participants in the former condition found themselves forced to make an adverse, upward social comparison – their partner had “performed” better than they had in the eating domain by eating lightly. In this study, the outcome of interest was the participant’s rating of the partner, assessed by means of a composite “liking” variable that included, among others, items tapping liking for, perceived similarity to, and willingness to work with the partner. Participants whose partners had eaten less than they had eaten gave lower ratings to those partners than did participants whose partners had eaten more than they had eaten. This finding is reminiscent of Pleban and Tesser’s (1981) finding that participants forced into an upward social comparison on a dimension that was relevant to them (as we assume that eating was for the women in the Leone et al. study) “distanced” themselves from the comparison other, rating him as less similar and expressing less willingness to work with him than did those for whom relevance was low and those who were not forced into an upward comparison. Thus, in two studies upward social comparisons produced by causing the participant to perform less well in the eating domain than another person under conditions in which attainability was low resulted in classic social comparison findings: negative affect and distancing the self from the better-performing other. Conclusion We are proposing in this paper that social influences on eating in many cases reflect the operation of the underlying mechanism of social comparison processes. Festinger (1954) posited that humans have a basic drive or need to evaluate and assess themselves, and the means of evaluation is comparing oneself to other people, some of whom are more or less similar to oneself. The degree of similarity to the other determines the extent to which we will compare with that person. Thus, as we have indicated, some social “models”

will affect our eating more than others will, or will affect some of us and not others, depending on our similarity to the model. When confronted with a comparison on an attainable dimension (e.g., amount of food eaten when food is available), we can change our own eating; if the comparison involves an unattainable quality (e.g., eating has already occurred, or we are required to eat a large amount) we may react negatively, with consequences for our evaluation of the other person. We are thus comparing ourselves to others to determine what and how much to eat, how we feel about our own food, how we feel about ourselves, and how we feel about others who eat similarly to or differently from us. Much of the literature on social influences on eating focuses on amounts eaten by individuals eating with real or putative other people (e.g., modeling, social facilitation), or on the impressions that we make on others by what or how much we eat (e.g., impression management). We suggest that the driving process for these effects is the comparison we are making between ourselves and the others in the eating situation. The drive to compare ourselves serves our need to bolster our own identity or feelings about ourselves. We may ultimately turn this comparison into a means of impressing others or matching social norms, but the process begins with the attempt to serve our own needs to feel that we are good people and/or correct in our own behavior. References Anschutz, D. J., Engels, R. C. M. E., Becker, E. S., & van Strien, T. (2008). The bold and the beautiful. Influence of body size of televised media models on body dissatisfaction and actual food intake. Appetite, 51, 530–537. Anschutz, D. J., Engels, R. C. M. E., Becker, E. S., & van Strien, T. (2009). The effects of TV commercials using less thin models on young women’s mood, body image and actual food intake. Body Image, 6, 270–276. Anschutz, D. J., Van Strien, T., & Engels, R. C. M. E. (2008). Exposure to slim images in mass media. Television commercials as reminders of restriction in restrained eaters. Health Psychology, 27, 401–408. Boyce, J. A., Kuijer, R. G., & Gleaves, D. H. (2013). Positive fantasies or negative contrasts. The effect of media body ideals on restrained eaters’ mood, weight satisfaction, and food intake. Body Image, 10, 535–543. Buunk, B. P., Collins, R. L., Taylor, S. E., VanYperen, N. W., & Dakof, G. A. (1990). The affective consequences of social comparison. Either direction has its ups and downs. Journal of Personality and Social Psychology, 59, 1238–1249. Cachelin, F. M., Striegel-Moore, R., & Paget, W. B. (1997). Comparison of women with various levels of dietary restraint on body image, personality, and family environment. Eating Disorders, 5, 205–215. Campbell, M. C., & Mohr, G. S. (2011). Seeing is eating. How and when activation of a negative stereotype increases stereotype-conducive behavior. Journal of Consumer Research, 38, 431–444. Chaiken, S., & Pliner, P. (1987). Women, but not men, are what they eat. The effect of meal size and gender on perceived masculinity and femininity. Personality and Social Psychology Bulletin, 13, 166–176. Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159. Cruwys, T., Bevelander, K. E., Hermans, R. C. J. (2014). Social Modeling of Eating: A Review of When and Why Social Influence Affects Food Intake and Choice. Appetite, doi:10.1016/j.appet.2014.08.035 de Castro, J. M., & de Castro, E. S. (1989). Spontaneous meal patterns of humans. Influence of the presence of other people. American Journal of Clinical Nutrition, 50, 237–247. Dijksterhuis, A., & Bargh, J. A. (2001). The perception–behavior expressway. Automatic effects of social perception on social behavior. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 33, pp. 1–40). San Diego, CA: Academic Press. Durkin, K., Hendry, A., & Stritzke, W. G. K. (2012). Mixed selection. Effects of body images, dietary restraint, and persuasive messages on females’ orientations towards chocolate. Appetite, 60, 95–102. Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117–140. Florak, A., Palcu, J., & Friese, M. (2013). The moderating role of regulatory focus on the social modeling of food intake. Appetite, 69, 114–122. Green, B. L., & Saenz, D. S. (1995). Tests of a mediational model of restrained eating. The role of dieting self-efficacy and social comparisons. Journal of Social and Clinical Psychology, 14, 1–22. Groesz, L. M., Levine, M. P., & Murnen, S. K. (2002). The effect of experimental presentation of thin media images on body satisfaction. A meta-analytic review. International Journal of Eating Disorders, 31, 1–16. Heatherton, T. F., Striepe, M., & Wittenberg, L. (1998). Emotional distress and disinhibited eating. The role of self. Personality and Social Psychology Bulletin, 24, 301–313. Herman, C. P., & Polivy, J. (1980). Restrained eating. In A. F. Stunkard (Ed.), Obesity (pp. 208–225). Philadelphia: W.B. Saunders.

J. Polivy, P. Pliner/Appetite 86 (2015) 88–95

Herman, C. P., Roth, D., & Polivy, J. (2003). Effects of the presence of others on food intake. A normative interpretation. Psychological Bulletin, 129, 873–886. Hermans, R. C. J., Larsen, J. K., Herman, C. P., & Engels, R. C. M. E. (2008). Modeling of palatable food intake in female young adults. Effects of perceived body size. Appetite, 51, 512–518. Hetherington, M. M., Anderson, A. S., Norton, G. N. M., & Newson, L. (2006). Situational effects on meal intake. A comparison of eating alone and eating with others. Physiology & Behavior, 88, 498–505. Larson, N. I., Nelson, M. C., Neumark-Sztainer, D., Story, M., & Hannan, P. J. (2009). Making time for meals. Meal structure and associations with dietary intake in young adults. Journal of the American Dietetic Association, 109, 72–79. Leone, T., Herman, C. P., & Pliner, P. (2008). Perceptions of undereaters. A matter of perspective? Personality and Social Psychology Bulletin, 34, 1737–1746. Lockwood, P., & Kunda, Z. (1997). Superstars and me. Predicting the impact of role models on the self. Journal of Personality and Social Psychology, 73, 91–103. Markus, H., & Nurius, P. (1986). Possible selves. American Psychologist, 41, 954–969. McFerran, B., Dahl, D. W., Fitzsimons, G. J., & Morales, A. C. (2010a). I’ll have what she’s having. Effects of social influence and body type on the food choices of others. Journal of Consumer Research, 36, 915–929. McFerran, B., Dahl, D. W., Fitzsimons, G. J., & Morales, A. C. (2010b). Might an overweight waitress make you eat more? How the body type of others is sufficient to alter our food consumption. Journal of Consumer Psychology, 20, 146–151. Mills, J. S., Polivy, J., Herman, C. P., & Tiggemann, M. (2002). Effects of exposure to thin media images. Evidence of self-enhancement among restrained eaters. Personality and Social Psychology Bulletin, 28, 1687–1699. Monro, F. J., & Huon, G. F. (2006). Media-portrayed idealized images, selfobjectification, and eating behavior. Eating Behaviors, 7, 375–383. Mori, D., Chaiken, R., & Pliner, P. (1987). “Eating lightly” and the self-presentation of femininity. Journal of Personality and Social Psychology, 53, 693–702. Myers, T. A., & Crowther, J. H. (2009). Social comparison as a predictor of body dissatisfaction. A meta-analytic review. Journal of Abnormal Psychology, 118, 683–698. Papies, E. K., & Nicolaije, K. A. H. (2012). Inspiration or deflation? Feeling similar or dissimilar to slim and plus-size models affects self-evaluation of restrained eaters. Body Image, 9, 76–85. Pleban, R., & Tesser, A. (1981). The effects of relevance and quality of another’s performance on interpersonal closeness. Social Psychology Quarterly, 44, 278– 285. Pliner, P., Bell, R., Hirsch, E. S., & Kinchla, M. (2006). Meal duration mediates the effect of “social facilitation” on eating in humans. Appetite, 46, 189–198. Pliner, P., Chaiken, S., & Flett, G. L. (1990). Gender differences in concern with eating, weight, and physical appearance over the lifespan. Personality and Social Psychology Bulletin, 16, 263–273. Pliner, P., Royal, S., Pelletier, M., Leone, T., Herman, C. P., & Polivy, J. (2009). Selfevaluations following overeating. Unpublished manuscript. Polivy, J., & Herman, C. P. (1999). Distress and eating. Why do dieters overeat? International Journal of Eating Disorders, 26, 153–164. Polivy, J., Herman, C. P., & Deo, R. (2010). Getting a bigger slice of the pie. Effects on eating and emotion in restrained and unrestrained eaters. Appetite, 55, 426–430. Polivy, J., Herman, C. P., Younger, J., & Erskine, B. (1979). Effects of a model on eating behavior. The induction of a restrained eating style. Journal of Personality, 47, 100–117. Redd, E. M., & de Castro, J. M. (1992). Social facilitation of eating. Effects of instructions to eat alone or with others. Physiology & Behavior, 52, 749–754.

95

Robinson, E., Blisset, J., & Higgs, S. (2013). Social influences on eating. Implications for nutritional interventions. Nutrition Research Reviews, 26, 166–176. Robinson, E., Thomas, J., Aveyard, P., & Higgs, S. (2014). What everyone else is eating. A systematic review and meta-analysis of the effect of informational eating norms on eating behavior. Journal of the Academy of Nutrition and Dietetics, 114, 414–429. Rodrigues, S. S. P., & Almeida, M. D. V. (1996). Food habits. Concepts and practices of two different age groups. In J. S. A. Edwards (Ed.), Culinary arts and sciences. Global and national perspectives (pp. 387–396). Boston, MA: Computational Mechanics. Roth, D., Herman, C. P., Polivy, J., & Pliner, P. (2001). Self-presentational conflict in social eating situations. A normative perspective. Appetite, 36, 165–171. Rozin, P. (1996). The socio-cultural context of eating and food choice. In H. L. Meiselman & H. J. H. Macfie (Eds.), Food choice acceptance and consumption (pp. 83–104). Glasgow: Blackie Academic and Professional. Salvy, S. J., Howard, M., Read, M., & Mele, E. (2009). The presence of friends increases food intake in youth. American Journal of Clinical Nutrition, 90, 282–287. Salvy, S. J., Romero, N., Paluch, R., & Epstein, L. H. (2007). Peer influence on preadolescent girls’ snack intake. Effects of weight status. Appetite, 49, 177–182. Schachter, S. (1959). The psychology of affiliation. Experimental studies of the sources of gregariousness. Palo Alto, CA: Stanford University Press. Seddon, L., & Berry, N. (1996). Media-induced disinhibition of dietary restraint. British Journal of Health Psychology, 1, 27–33. Sobal, J., & Nelson, M. K. (2003). Commensal eating patterns. A community study. Appetite, 41, 181–190. Strahan, E. J., Spencer, S. J., & Zanna, M. P. (2007). Don’t take another bite. How sociocultural norms for appearance affect women’s eating behavior. Body Image, 4, 331–342. Strauss, J., Doyle, A. E., & Kreipe, R. E. (1994). The paradoxical effect of diet commercials on reinhibition of dietary restraint. Journal of Abnormal Psychology, 103, 441–444. Stunkard, A. J., & Messick, S. (1985). The three-factor eating questionnaire to measure dietary restraint, disinhibition and hunger. Journal of Psychosomatic Research, 29, 71–83. Tesser, A. (1991). Emotion in social comparison and reflection processes. In J. Suls & T. A. Wills (Eds.), Social comparison. Contemporary theory and research (pp. 115–145). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Trottier, K., Polivy, J., & Herman, C. P. (2007). Effects of exposure to thin and overweight peers. Evidence of social comparison in restrained and unrestrained eaters. Journal of Social and Clinical Psychology, 26, 155–172. Van Strien, T., Frijters, J. E. R., Bergers, G. P. A., & Defares, P. B. (1986). The Dutch Eating Behaviour Questionnaire (DEBQ) for assessment of restrained, emotional and external eating behaviour. International Journal of Eating Disorders, 5, 295–315. Vartanian, L. R., Sokol, N., Herman, C. P., & Polivy, J. (2013). Social models provide a norm of appropriate food intake for young women. PLoS ONE, 8, e79268. Warren, C. S., Strauss, J., Taska, J. L., & Sullivan, S. J. (2005). Inspiring or dispiriting? The effect of diet commercials on snack food consumption in high school and college-aged women. International Journal of Eating Disorders, 37, 266–270. Wood, J. V., Taylor, S. E., & Lichtman, R. R. (1985). Social comparison in adjustment to breast cancer. Journal of Personality and Social Psychology, 49, 1169–1183. Wu, D. W. L., Bischof, W. F., & Kingstone, A. (2013). Looking while eating. The importance of social context to social attention. Scientific Reports, 3, 2356. Young, M. E., Mizzau, M., Mai, N. T., Sirisegaram, A., & Wilson, M. (2009). Food for thought. What you eat depends on your sex and eating companions. Appetite, 53, 268–271.