Persistent high self-focus after failure and low self-focus after success: the depressive self-focusing style.
Dieting and nondieting subjects were given either failure or neutral performance feedback on a problem-solving task. Failure subjects were then assigned to one of three self-awareness conditions: One group was forced to watch a video clip of themselves failing on the problem-solving task, one group was asked to watch a distracting video clip on bighorn sheep, and the final group was asked to sit quietly for 10 minutes. Subjects were then allowed to eat as much ice cream as tbey wanted. Only in those conditions which allowed—or promoted—low self-awareness (the failure/distraction and simple failure condition) did dieters sbow disinhibited eating. In the failure/videotape condition, which enforced high levels of self-awareness, eating in dieters remained itihibited. This supports the proposal that a reduction in self-awareness is necessary for lifting of inhibitions. Eating in nondieters was reduced in the failure/videotape and simple failure conditions, possibly because of the autonomic correlates of distress. This research was supported by the National Sciences and Engineering Research Council of Canada and the Social Sciences and Humanities Research Council of Canada. We thank Lynn Kozlowski, Patricia Pliner, and Dianne Tice for their comments on this manuscript. Address correspondence to Todd F. Heatherton, Department of Psychology, Harvard University, 33 Kirkland Street, Cambridge, MA 02138. Journal of Personality 61:1, March 1993. Copyright © 1993 by Duke University Press. CCC 0022-3506/93/51.50 50 Heatherton et al. Psychology's understanding of the effects of emotional distress on eating has evolved through several stages. Initially it was believed that distress suppressed eating for everyone because of its attendant autonomic correlates (e.g., increased levels of sugar in the bloodstream; Cannon, 1915). However, Schachter, Goldman, and Gordon (1968) showed that obese individuals failed to show this reduction of eating in response to emotional distress, and Herman and Polivy (1975) later demonstrated that distress actually increased eating among dieters while suppressing it among nondieters.' The rapid accumulation of findings since then has supported that pattern, although it appears to depend on the form of emotional distress involved (see Herman, Polivy, & Heatherton, 1991, for a review). Physical fear reduces eating among nondieters but fails to produce significant increases in eating among dieters; only self-image threats produce significant increases in eating among dieters (Heatherton, Herman, & Polivy, 1991). Thus, emotional distress produces significant increases in eating among dieters only if the distress involves some unfiattering implications about the self. This suggests that the causal mechanism involves self-awareness or some related set of motivations involving the self (such as self-presentational concems; see Schlenker & Leary, 1982, on self-presentation and self-awareness). Failure presumably activates motivations and attentional shifts involving the self, and the resulting processes may aflfect eating. A recent review by Heatherton and Baumeister (1991) proposed that eating binges among dieters, bulimics, and others occur when these individuals escape from aversive high selffocus by achieving a cognitively restricted state that has the effect of disinhibition. The impact of failure on the self has been illuminated by a variety of studies. Recent work by Greenberg and Pyszczynski (1986) has shown that ego-relevant failure leads to a state of high self-awareness. This state is presumably quite unpleasant, for it involves awareness of the 1. The term "dieter" is used to refer to individuals who chronically restrain their food intake. Such individuals may differ from casual dieters (Heatherton & Polivy, in press; Polivy & Herman, 1987). However, the majority of individuals who identify themselves as dieters are likely to display the sorts of behaviors that we describe herein. There is some controversy over the use of various dieter descriptions; those interested in this issue should consult Heatherton, Herman, Polivy, King, and McGree (1988) for a full discussion of these issues. In this article we use the terms dieter and restrained eater interchangeably. In all cases we are referring to those individuals who chronically attempt to restrain their intake of food in order to lose weight. Attentional Focus and Eating 51 self as incompetent or otherwise deficient. The aversiveness of this state thus motivates people to want to escape from it (e.g., Duval & Wicklund, 1972; Greenberg & Musham, 1981). High self-awareness in general has been linked to the strong and effective functioning of inhibitions (see Diener, 1979). For example, Diener and Wallbom (1976) showed that mirrors and other self-focusing cues substantially reduced the likelihood of cheating on a test. Beaman, Klentz, Diener, and Svanum (1979) found that the presence of a mirror reduced the stealing of Halloween candy substantially. In contrast, the process of escape from self-awareness has been associated with disinhibition (Baumeister, 1991; Diener, 1979; Wicklund, 1982). Thus, for example, the effects of alcohol include both a reduction in self-awareness (Hull, 1981) and the removal of certain inner inhibitions, leading to behavioral excess (Steele & Josephs, 1988; Steele & Southwick, 1987). Similarly, in group settings, an increase in group size is associated with a decrease in self-awareness (Mullen, 1987), as well as disinhibited aggressive behavior such as lynch mob atrocity (Mullen, 1986). Ego-relevant failure should therefore produce a high state of selfawareness that will strengthen preexisting inhibitions, such as those against eating in dieters. Escape from this aversive self-aware state, however, should have a disinhibiting effect that could result in increased eating among dieters. The present study tested these predictions in dieting and nondieting subjects (restrained and unrestrained eaters, respectively) by exposing them to a distressing failure, manipulating their subsequent attention, and examining the consequences for their subsequent eating behavior. The study began with two conditions comparing failure and a nofailure control. Following ego-relevant failure, some subjects were allowed to sit by themselves for 10 minutes, in order to allow them to escape from their aversive self-awareness. Consistent with past findings (cf. Baucom& Aiken, 1981 ;Heathertonetal., 1991;Ruderman, 1985), we predicted that dieters would show disinhibited (i.e., increased) eating following failure as compared with dieters in the control condition. To shed light on the mechanism, however, we added two further conditions. The first of these was designed to force people to remain in the highly self-aware state; these subjects were videotaped during their failure and then were required to watch the tape of their own poor performance while being questioned about their failure by the experimenter. The presence of the camera, the self-observation, and the re52 Heatherton et al. suiting self-presentational dilemma in interacting with the experimenter should make it difficult for these subjects to escape from awareness of themselves, and so their eating was predicted to remain inhibited. In contrast, the final condition sought to facilitate the subjects' escape from self-awareness by showing them a film that had no relevance to their performance or self-concept. These subjects should therefore find it easy to escape from self-awareness and hence were predicted to show the pattern of disinhibited eating. For comparison purposes, we extended the design of the study to include nondieters. Nondieters, by definition, lack the set of inhibitions regarding eating that dieters have, and perhaps for that reason their eating patterns follow very different pattems and principles (e.g., Heatherton et al., 1991; Herman et al., 1991). We predicted only that they would fail to conform to the pattems of dieters' responses, since selfimage threats usually have weak effects on nondieters' eating (Herman etal., 1991).