Impact of dysphoria and self-consciousness on perceptions of social competence: Test of the depressive realism hypothesis

Clinical Psychologist, Vol. 10, No. 3, November 2006, pp. 99 108. Impact of dysphoria and self-consciousness on perceptions of social competence: Test of the depressive realism hypothesis PHUONG M. CHAU & LEONARD S. MILLING Department of Psychology, University of Hartford, West Hartford, Connecticut, USA Abstract Depressive realism refers to a cognitive style wherein depressed people sometimes have more accurate perceptions of reality than nondepressed people. The notion of depressives being sadder yet wiser was controversial when first presented, and continues to be heavily debated. Self-perception studies provide maximum external validity, but have been characterised by methodological limitations and consequently have yielded contradictory results. This study was designed to address limitations of past research. Seventy-two female participants were paired into dysphoric, nondysphoric or dysphoric nondysphoric dyads to engage in 20-min social interactions. Thereafter, they rated their partners on social competence, as well as how they believed their partners would rate them. Accuracy was determined by comparing participants perceptions of their conveyed impressions with actual ratings given to them by their partners. Results challenge the depressive realism hypothesis. Keywords: Depressive realism, perceptions of social competence Depressive realism refers to a cognitive style in which depressed people sometimes have more accurate perceptions of reality than nondepressed people. The notion of depressives being sadder yet wiser than nondepressives was controversial when first presented (Alloy & Abramson, 1979), and continues to be debated in the research literature on depression (Ackerman & DeRubeis, 1991; Colvin & Block, 1994; Haaga & Beck, 1995) despite the dozens of published studies and several reviews of the depressive realism phenomenon in the last two decades. As Alloy, Albright, Abramson, & Dykman (1990) note, much of the controversy on depressive realism stems from the fact that it violates several longstanding perspectives on mental health and psychopathology. Humans have long been considered to be rational decision-makers who process information in a logical and unbiased manner, and the predominant view in mental health has been that accurate perceptions of the self, world, and future are essential for psychological adjustment and wellbeing (Taylor & Brown, 1988). It makes sense from an intuitive standpoint that mental health entails maintaining close contact with reality. In addition, Beck s (1967) cognitive theory of depression posits that depressed people have a negatively distorted view of themselves, the world, and the future. Thus, it seems especially counterintuitive that depressive symptomology would be associated with having a better grip on reality. Furthermore, Beck s cognitive therapy for depression, which has had much empirical validation as an efficacious treatment method for depression (Dobson, 1989a), is believed to achieve its effect by helping depressed people think in a more realistic and rational manner. Research on depressive realism implies that cognitive therapy works not by making depressed clients think more realistically, but perhaps by training depressed people to engage in the positive biases and unwarranted optimism typically exhibited by nondepressives (Alloy & Abramson, 1988; Baumeister, 1989; see Taylor & Brown, 1988, for a review). Upon closer inspection, the notion of depressive realism may not be entirely surprising. Indeed, there is an abundance of research suggesting that rationality is not always characteristic of normal thought. The social psychology literature is replete with Correspondence: Dr L. S. Milling, University of Hartford, Department of Psychology, 200 Bloomfield Avenue, West Hartford, CT 06117, USA. E-mail: milling@hartford.edu Phuong M. Chau now attends the Ph.D. program in clinical psychology at the University of Kansas. ISSN 1328-4207 print/issn 1742-9552 online ª The Australian Psychological Society Ltd Published by Taylor & Francis DOI: 10.1080/13284200600939900

100 P. M. Chau & L. S. Milling striking examples. For example, the self-serving bias (Miller & Ross, 1975) refers to people s tendency to take personal credit for successes and minimise their responsibilities for failures, blaming failures instead on external factors. Other examples of common cognitive distortions include self-fulfilling prophecies (Harris & Rosenthal, 1985; Neuberg, 1989), the fundamental attribution error (Ross, 1977), the false consensus effect (Ross, Greene, & House, 1977), unrealistic optimism about future events (Weinstein, 1980), affective forecasting (Gilbert, Pinel, Wilson, Blumberg, & Wheatley, 1998), and the illusory correlation (Hamilton & Gifford, 1976). These are only a small handful of the errors or faulty attributions that normal people typically commit when attempting to make sense of themselves and their social world. Often, people are unaware of making such errors in their attributions. Theorists have presented compelling arguments that many of these biases have adaptive functions and actually help to increase positive affect and promote mental health (Baumeister, 1989; Taylor & Brown, 1988). That is, promoting and enhancing one s self-image may provide a person with the necessary confidence and sense of control or mastery in order to work through difficult goals and cope with life s inevitable stresses. Interestingly, Freud (as cited in Baumeister, Dale, & Sommer, 1998) had originally conceptualised defence mechanisms as one s distortions of reality that serve to protect oneself from anxiety. Perhaps depressive realism, then, results from the breakdown of such biases and motivations in depressed individuals, making them wiser in some instances but at the expense of psychological adjustment. If depressive realism does exist as a true phenomenon, it would have profound implications for current conceptions of psychopathology, depression, and psychological health and wellbeing. If taken at face value, findings on depressive realism would seem to indicate that in order to be psychologically healthy, one must not see the self and world as they really are. However, evidence for depressive realism has not been consistent. Some studies have provided support for depressive realism, while others have provided conflicting evidence for the hypothesis. Previous reviews of depressive realism (Ackerman & DeRubeis, 1991; Alloy & Abramson, 1988; Alloy et al., 1990; Dobson & Franche, 1989) have come to different conclusions regarding the strength and pervasiveness of the phenomenon, further complicating the picture. As Dobson and Franche (1989) note, selfperception studies provide one of the most meaningful and significant contexts in which depressive realism can be examined. The accuracy of human social judgment is a topic of particular concern and importance (Kruglanski, 1989), and it is interesting to examine the degree to which people are right or wrong about the impression they create to others. Previous investigators (Campbell & Fehr, 1990; Dobson, 1989b; Gotlib & Meltzer, 1987; Lewinsohn, Mischel, Chaplin, & Barton, 1980; Kistner, Balthazor, Risi, & David, 2001) have explored the self-perceptions of depressed and nondepressed individuals in social situations, but most of these studies were unable to directly test the depressive realism hypothesis because they did not use procedures in which accurate responses could truly be specified (Ackerman & DeRubeis, 1991). Operational or reality criteria are extremely difficult to utilise in the domain of selfperception (Funder, 1987; Robins & John, 1997), but they are necessary if one wishes to make a statement on the realism of the depressed versus nondepressed. The present study attempted to address the limitations typical of depressive realism research, by examining depressive realism in a meaningful and significant context, at the same time utilising an objective standard for accuracy in order to be able to speak directly to the depressive realism hypothesis. In addition to depressive symptomology, people s perceptions of how they convey themselves to other people may be affected by the degree to which they actively attend to themselves and the public image that they are projecting. Self-consciousness refers to an individual s tendency to devote a substantial amount of time and energy in the state of selfawareness, which can involve private aspects about the self (thoughts, feeling, attitudes, and desires), or public aspects of the self that are observable by others (one s physical appearance, speech, and actions). Sedikides and Skowronski (1995) have suggested that the accuracy of self-perceptions is influenced by how attentive people are to their inner and outer states. Self-consciousness may increase self-knowledge, which in turn increases accuracy on judgments about the self. This study examines perceptions of dysphoric and nondysphoric individuals regarding their social competence on an interpersonal task. In particular, this study evaluates whether depression plays a role in how accurate individuals are about the social impressions that they convey to other people, and if this relationship is influenced by one s level of selfconsciousness. Participants engaged in 20-min dyadic social interactions, after which they were asked to give social competence ratings for their partners and how they believed their partners would rate them. Accuracy was determined by comparing participants perceptions of their conveyed impressions with actual ratings given to them by their partners. A clarification of terminology is necessary prior to the discussion of study hypotheses. The depressed sample in this study consisted of college students who were not likely to meet DSM criteria for depression.

Depressive realism 101 Following the recommendations of Kendall, Hollon, Beck, Hammen, and Ingram (1987), these study participants will henceforth be referred to as dysphoric rather than depressed. The terms nondepressed and nondysphoric will be considered interchangeable here and used accordingly to reflect consistency throughout this article (e.g., depressed vs. nondepressed, dysphoric vs. nondysphoric). Several predictions have been generated for this study. First, the depressive realism hypothesis suggests that nondysphoric individuals will show a positivity bias and overestimate partner perceptions of their social competence. In contrast, dysphoric individuals are predicted to be either accurate or negatively biased in their perceptions. Second, based on the theoretical construct of self-consciousness, it is hypothesised that higher levels of selfconsciousness will be associated with greater accuracy, whereas lower levels of self-consciousness will be associated with less accuracy. Finally, an interaction between dysphoria and self-consciousness is predicted, such that high self-consciousness will be associated with greater accuracy in both dysphoric and nondysphoric subjects, but low self-consciousness will be associated with a pessimistic bias in dysphoric and an optimistic bias in nondysphoric subjects. Methods Participants Participants consisted of 72 female introductory psychology students who voluntarily participated in order to satisfy a course requirement. During an initial screening session, participants completed the Beck Depression Inventory (BDI; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) and the Self- Consciousness Scale (Fenigstein, Scheier, & Buss, 1975). On the basis of their initial BDI scores, participants were classified as dysphoric or nondysphoric and invited to participate in the experiment. Participants scoring 9 on the BDI (n ¼ 30) comprised the dysphoric sample, and participants scoring 8 (n ¼ 42) comprised the nondysphoric sample. This selection criterion is consistent with those utilised in previous and similar studies. Following the recommendations of Kendall et al. (1987), participants completed the BDI a second time on the day of the experiment to ensure temporal stability of BDI scores. Participants were required to meet criteria for group assignment on both occasions in order to reduce the likelihood that they were experiencing transient or uncharacteristic mood states. Individuals whose scores did not qualify them for the same group on both occasions (n ¼ 6) were excluded from the analysis. Measures Beck Depression Inventory. The BDI (Beck et al., 1961) is a 21-item self-report inventory that measures characteristic attitudes and symptoms of depression. The BDI was selected for this study because it has been used extensively by clinicians and researchers, and is demonstrated to be both a reliable and valid measure of depressive symptoms. According to Beck and Steer (as cited in Robinson & Kelley, 1996), internal consistency estimates (Cronbach s a) range from.79 to.90, and correlation coefficients range from.38 to.76 when BDI scores are correlated with other existing measures of depression. Self-Consciousness Scale. The Self-Consciousness Scale (Fenigstein et al., 1975) is a 23-item selfreport instrument measuring individual differences in private and public self-consciousness. The Self- Consciousness Scale demonstrates reasonably high internal consistency, with alpha coefficients ranging from.63 to.80 (Alanazi, 2001; Nystedt & Smari, 1989). Test retest reliability ranges from.60 and.89 (Nystedt & Smari, 1989). Social Competence Scale. The Social Competence Scale (Lewinsohn et al., 1980) involves the rating of 17 attributes related to social competence (friendly, popular, assertive, attractive, warm, socially skillful, etc.) on a scale from 1 to 7, where a rating of 1 indicates not at all characteristic and 7 indicates extremely characteristic. The psychometric properties of this instrument have not been clearly established. However, the alpha coefficients for the 17 items in this study were.92 for ratings given to participants by their partners, and.93 for ratings made by participants on how they believed their partners would rate them. The validity of the Social Competence Scale has been demonstrated in previous studies in which the attributes comprising the scale were found to differentiate between depressed and nondepressed individuals (Gotlib, 1982; Youngren & Lewinsohn, 1980). Procedure After the initial screening, students were invited to participate in the experimental session (which, for most participants, occurred 2 4 weeks after the initial screening session). In the experimental session, students were paired with another female student (dysphoric-only pairs, nondysphoric-only pairs, and dysphoric nondysphoric pairs) to participate in 20- min unstructured social interactions. Precautions were taken to insure that both individuals were strangers to each other, and no one was paired with someone she knew. The dyadic interactions took

102 P. M. Chau & L. S. Milling place in a small, comfortable room with two chairs facing each other and approximately 3 ft (1 m) away. The experimenter met with each dyad and explained that the purpose of the study was to examine how people who do not know each other relate with one another. Participants were instructed to converse and become acquainted with each other for 20 min. They were told that they were free to talk about any topic other than the current research project and their particular roles in the study. The experimenter left the room after instructions were given and understood and all questions were answered. Upon completion of the dyadic interactions, one of the participants was led to a different room and both participants were asked to rate their interaction partners and themselves on the 17 attributes from the Lewinsohn et al. (1980) Social Competence Scale. In particular, participants made ratings for (a) their partners, and (b) how they believed their partners would rate them. Participants also completed the BDI a second time to determine if they met study inclusion criteria as described earlier. Results Preliminary analyses Table I shows means and standard deviations for all relevant study variables. Anticipated partner ratings refer to ratings made by participants when asked how they believed they were assessed by their partners on the Social Competence Scale; partner ratings refer to the actual ratings participants received from their partners on the Social Competence Scale. Table II presents anticipated and actual partner ratings of social competence by participant and partner dysphoric status. As explained, participants completed the BDI twice, once during the screening session and once on the day of the experiment. Only participants whose BDI scores qualified them for the same dysphoric status (dysphoric or nondysphoric) during both assessment periods were used in the analysis. The BDI mean score for all participants was 8.08 (SD ¼ 5.76) at Time 1, and 6.56 (SD ¼ 4.77) at Time 2. BDI internal consistency estimates (Cronbach s a) for the current sample were.81 for Time 1, and.79 for Time 2. The overall mean score for the Self-Consciousness Scale was 59.99 (SD ¼ 8.10). Table III presents correlations among all relevant study variables. The Pearson correlation coefficient for the BDI scores across both assessment periods was r ¼.85, p 5.001, indicating that scores on the BDI remained relatively consistent from Time 1 to Time 2. Scores on the BDI (at Time 1) and the Self- Consciousness Scale were uncorrelated (r ¼.12, ns), suggesting that these scales measure separate constructs. Partner ratings and anticipated partner ratings were significantly correlated (r ¼.32, p 5.01). Main analysis: Evaluating the depressive realism hypothesis As mentioned, anticipated partner ratings were ratings made by participants when asked how they Table II. Anticipated and actual partner ratings of social competence Anticipated partner ratings Partner ratings Participant and partner dysphoric status M SD M SD Dysphoric participant: with dysphoric partner (n ¼ 18) 5.50 0.58 5.58 0.74 with nondysphoric partner (n ¼ 12) 5.58 0.60 5.81 0.38 Nondysphoric participant: with dysphoric partner (n ¼ 12) 5.44 0.62 5.62 0.70 with nondysphoric partner (n ¼ 30) 6.03 0.51 6.10 0.52 Table I. Descriptive data for scores on BDI, Self-Consciousness Scale, and Social Competence Scale Dysphoric participants Nondysphoric participants Scales M SD M SD BDI at Time 1 13.33 4.76 4.33 2.61 BDI at Time 2 11.20 3.20 3.24 2.25 Self-Consciousness Scale 57.63 7.62 54.81 8.32 Social Competence Scale Anticipated partner ratings 5.53 0.64 5.86 0.60 Partner ratings 5.81 0.51 5.97 0.61 Note: BDI ¼ Beck Depression Inventory. Table III. Correlation matrix Scales 1 2 3 4 5 1. BDI at Time 1.85***.09 7.29* 7.17 2. BDI at Time 2.15 7.33** 7.19 3. Self-Consciousness 7.06 7.01 Scale 4. Social Competence.32** Scale, anticipated partner ratings 5. Social Competence Scale, partner ratings Notes: BDI ¼ Beck Depression Inventory. *p 5.05; **p 5.01; ***p 5.001.

Depressive realism 103 believed they were assessed by their partners on social competence, whereas partner ratings were the actual ratings participants received from their partners on social competence. The discrepancy between anticipated and actual partner ratings of social competence was used to test the depressive realism hypothesis. According to the depressive realism hypothesis, dysphoric participants should not show a discrepancy between anticipated and actual social competence ratings, whereas nondysphoric participants would show a discrepancy between anticipated and actual social competence ratings. This discrepancy was analysed as a repeated factor (i.e., ratings of social competence) in an analysis of variance. Participants were classified as dysphoric or nondysphoric on the basis of their BDI scores and this classification (i.e., participant dysphoric status) served as a between-subjects independent variable in the analysis of variance. Accordingly, a 262 (Participant dysphoric status6ratings of social competence) mixed-model analysis of variance (ANOVA), with actual partner ratings and anticipated partner ratings of social competence as a repeated factor, yielded a significant main effect for dysphoric status, F(1,70) ¼ 4.50, p ¼.037, Z 2 ¼.06. This indicates that dysphoric participants both anticipated and received lower ratings of social competence (M ¼ 5.67, SD ¼ 0.48) from their partners than nondysphoric participants (M ¼ 5.91, SD ¼ 0.48). In addition, there was a within-subjects main effect for ratings of social competence, F(1,70) ¼ 5.22, p ¼.025, Z 2 ¼.07. Anticipated partner ratings (M ¼ 5.72, SD ¼ 0.63) were significantly lower than partner ratings (M ¼ 5.90, SD ¼ 0.57). In other words, participants (regardless of whether they were dysphoric or nondysphoric) believed that their interaction partners viewed them less positively than was actually the case. The depressive realism hypothesis would predict an interaction between participant dysphoric status and ratings of social competence, such that there would be a nonsignificant discrepancy between anticipated and actual partner ratings for dysphoric subjects, and a significant discrepancy between anticipated and actual partner ratings for nondysphoric subjects. However, the interaction between participant dysphoric status and ratings of social competence was found to be nonsignificant, F(1,70) ¼ 0.98, ns, Z 2 ¼.01. Thus, the depressive realism hypothesis was not supported. Moderator analysis: Examining the role of self-consciousness In addition to findings on depressive realism, it was further hypothesised that self-consciousness would moderate the effect of dysphoric status on ratings of social competence, such that higher levels of selfconsciousness would be associated with little discrepancy between anticipated and actual partner ratings, and lower levels of self-consciousness would be associated with greater discrepancy between anticipated and actual partner ratings. According to Baron and Kenny (1986), when the independent variable (i.e., participant dysphoric status) is a dichotomy and the hypothesised moderator (i.e., self-consciousness) is a continuous variable, moderation is tested by including the product of the moderator and the dichotomous variable in a regression equation, which in this case would be an analysis of covariance with self-consciousness as the covariate. Thus, to determine if self-consciousness moderated ratings of social competence, a 262 (Participant dysphoric status6ratings of social competence) mixed model analysis of covariance (ANCOVA) was performed, with scores on the Self- Consciousness Scale serving as the moderator variable, and with actual partner ratings and anticipated partner ratings of social competence as a repeated factor. A moderator effect would be shown by a significant three-way interaction between ratings of social competence, dysphoric status, and scores on the self-consciousness scale. However, the analysis yielded nonsignificant results for this interaction, F(2,69) ¼ 0.68, ns, Z 2 ¼.02. Supplementary analyses Next, analyses were performed to examine the influence of partner dysphoric status on ratings of the social competence of dysphoric and nondysphoric participants. These supplementary analyses were performed to provide additional information and do not relate directly to any of our study hypotheses. A 26262 (Participant dysphoric status6partner dysphoric status6ratings of social competence) mixed model ANOVA, with anticipated and actual partner ratings of social competence as a repeated factor, yielded a significant main effect for partner dysphoric status, F(1,68) ¼ 6.62, p ¼.012, Z 2 ¼.09. This indicates that dysphoric partners (M ¼ 5.59, SD ¼ 0.45 ) were given overall lower ratings of social competence than nondysphoric partners (M ¼ 5.88, SD ¼ 0.49). However, the effect of participant dysphoric status was nonsignificant, F(1,68) ¼ 1.19, ns, Z 2 ¼.02. In addition, there was a within-subjects main effect for ratings of social competence, F(1,68) ¼ 4.91, p ¼.03, Z 2 ¼.07. Participants (regardless of dysphoric status) believed that their interaction partners (M ¼ 5.72, SD ¼ 0.63) viewed them less positively than how they were actually rated by their partners (M ¼ 5.90, SD ¼ 0.57). The interaction of ratings of social competence with participant dysphoric status

104 P. M. Chau & L. S. Milling was nonsignificant, F(1,68) ¼.56, ns, Z 2 ¼.01, as was the social competence by partner dysphoric status interaction, F(1,68) ¼.285, ns, Z 2 ¼.004. There was a significant participant Dysphoric status6partner dysphoric status interaction, F(1,68) ¼ 4.76, p ¼.03, Z 2 ¼.07. Figure 1 illustrates this interaction. A Duncan s new multiple range test (p 5.05) showed that nondysphoric participants who were paired with dysphoric partners (M ¼ 5.53, SD ¼ 0.45) received significantly lower ratings of social competence than nondysphoric participants who were paired with nondysphoric partners (M ¼ 6.07, SD ¼ 0.44). All other pairwise comparisons were nonsignificant. Finally, the three-way interaction of participant dysphoric status, partner dysphoric status, and ratings of social competence was nonsignificant, F(1,68) ¼ 0.004, ns, Z 2 ¼.004. Of note, the effect of participant dysphoric status was not significant in this analysis, but was significant in the earlier main analysis. This was because adding partner dysphoric status as a second betweensubjects factor here in the 26262 ANOVA produced a different partition of the total variance than was the case for the 262 ANOVA in the main analysis. Consequently, the effect of participant dysphoric status and ratings of social competence in the current analysis would not be equivalent to their effect in the main analysis (and therefore the current analysis does not provide a test of the depressive realism hypothesis). Discussion The key question investigated in this study examined whether dysphoric and nondysphoric individuals differed in how accurate they were about the impressions they conveyed to their social partners, and if this relationship was influenced by one s level of self-consciousness. According to the depressive realism hypothesis, dysphoric individuals would be more accurate than nondysphoric individuals about the social impressions that they convey to others. Based on the theoretical construct of selfconsciousness as conceptualised by various theorists (Davis & Franzoi, 1999; Fenigstein et al., 1975; Sedikides & Skowronski, 1995), it was also speculated that higher levels of self-consciousness would result in greater accuracy and that lower levels of selfconsciousness would result in less accuracy in these judgments. However, the data did not confirm these hypotheses. As such, our results add to a substantial literature challenging the depressive realism hypothesis (Ackerman & DeRubeis, 1991). Indeed, rather than supporting the notion that dysphoric individuals have more accurate perceptions of reality, our findings were in line with cognitive theories of depression (Beck, 1967, 1991) which argue that depressed individuals experience a negative perceptual bias. For example, our results showed that dysphoric individuals anticipated and received lower ratings of social competence from their partners than nondysphoric individuals. In other Figure 1. Mean overall ratings of social competence by participant and partner dysphoric status.

Depressive realism 105 words, dysphoric individuals were seen by their partners, and expected to be seen, as less socially competent than nondysphoric individuals. That dysphoric individuals anticipated lower ratings of social competence is consistent with cognitive theories of depression (Beck, 1967, 1991), which maintain that a negative view of the self, world, and future (also known as the negative triad) lies at the core of depression. The relatively low ratings of social competence for the dysphoric sample is also consistent with multiple lines of research documenting social skills deficits in depressed individuals (Coyne, 1976a; Gotlib, 1981; Gotlib & Meltzer, 1987; Joiner, Alfano, & Metalsky, 1992; Segrin, 2000; Strack & Coyne, 1983; Tse & Bond, 2004). These findings have important implications for the treatment of depression. First, our results suggest that depressed individuals are helped not by training them to engage in positive biases and unwarranted optimism (e.g., Taylor & Brown, 1988), but rather by assisting them to think in a more realistic and rational manner (Beck, 1967, 1991). Second, social cognitive and interpersonal models of depression (Coyne, 1976a; Sacco, 1999) suggest a destructive circular pattern of interpersonal behaviour commonly found in depression. Individuals who are depressed are more likely to consider themselves to be socially inept and incapable of having satisfying experiences with other people (Strack & Coyne, 1983). As data from this study suggest, depressed individuals anticipate that others view them negatively. Consequently, depressed individuals may withdraw from social situations or enter social encounters with trepidation or disinterest. As a result, other people are likely to respond to them negatively, which in turn perpetuates their depression and exacerbates their view of themselves, as these negative appraisals become directly or indirectly communicated. Thus, the treatment of depression should always incorporate an assessment of the depressed individual s interpersonal relationships, characteristic social functioning, and beliefs about their social skills. Inaccurate perceptions about the impression that one conveys to others is also of central importance in disorders aside from depression, such as social anxiety (Norton & Hope, 2001). An interesting (and unexpected) finding from this study was that partner ratings of social competence were significantly more favorable than participants beliefs about how their partners would rate them. That is, individuals were negatively skewed in their assessments of the social impressions they conveyed to their partners. When compared to their partner s actual impressions, participants (regardless of their dysphoric status) underestimated the positivity of their partners evaluations. Although the perceptions of dysphoric participants were less accurate than that of nondysphoric participants (i.e., dysphoric participants were more negatively distorted than nondysphoric participants in estimating the impressions they conveyed to their interaction partners), this difference was not found to be statistically significant. The underestimation of partner ratings by depressives is contrary to the depressive realism hypothesis, which suggests that they would be more accurate in their perceptions of the world. Yet, this finding is consistent with the Beck (1967, 1991) theory of depression: that depressed individuals have a negative view of the self and the world (and therefore a negative view of how they might be perceived by others). Although this investigation was not able to make a statement whether depressives are more or less accurate than nondepressives in these perceptions, the research literature and results from this study do support an association between depression and a negative view of the self and world. It may not be entirely surprising that depressives might perceive their world negatively, but we would not expect this to be the case for nondepressives. Indeed, this finding runs counter to previous investigations illustrating a consistent tendency for nondepressed individuals to demonstrate unwarranted and overly optimistic biases in their selfappraisals (Baumeister, 1989; Taylor & Brown, 1988). For example, in Lewinsohn et al. (1980), nondepressive subjects gave themselves significantly higher ratings of social competence than ratings given to them by noninteractive observers. Why then, did the results of this study not corroborate previous findings of nondepressive optimism? This may partly result from the design of the present study. Rather than using self-ratings of social competence as a criterion to be judged against partner ratings of social competence (which was the criterion used in many previous studies of this nature, including the study by Lewinsohn et al.), this experiment used a reflected appraisals method (Robins & John, 1997) in order to study accuracy of social perceptions. Therefore, partner ratings of social competence were compared not to self-ratings, but rather, to how individuals believed their partners would rate them on a measure of social competence. When asked to predict how others might see them, individuals may be more modest in their selfappraisals than when asked how they would rate their own social competence as they, not others, see it. A social desirability response bias may be present when people are asked to assess how they are viewed by others, such that they want to appear modest in their predictions. Our unexpected finding is also consistent with what has been called the spotlight effect : the tendency to believe that one s social shortcomings are more noticeable to other people than they actually are (Gilovich & Savitsky, 1999;

106 P. M. Chau & L. S. Milling Savitsky, Epley, & Gilovich, 2001). Additionally, perhaps another reason for this presumed modesty effect relates to the somewhat contrived and artificial nature of the experimental task, and furthermore with strangers with whom participants were not previously acquainted. People may be less modest and more accurate in their beliefs about how they are socially perceived by others with whom they are better acquainted and have more meaningful significance in their lives, such as family, close friends, or work colleagues. It would be interesting to examine whether these differences, if such differences truly exist, qualify as one of the boundary conditions on depressive realism and nondepressive optimism. Another finding from this study worthy of discussion involves the interaction between participant dysphoric status and partner dysphoric status. Whereas dysphoric individuals anticipated and received similar ratings from their dysphoric and nondepressed partners, ratings given to nondysphoric individuals appeared to depend on the depressive status of their partners. When nondysphoric participants had nondysphoric social partners, they believed they demonstrated relatively high levels of social competence to their partners (which was confirmed by partner ratings). However, nondysphoric participants who were paired with a dysphoric partner received lower overall ratings than nondysphoric participants who had nondysphoric partners. In other words, nondysphoric individuals believed that ratings given to them by dysphoric partners would be less favorable than ratings given to them by nondysphoric partners, and they also received lower ratings from their dysphoric partners than their nondysphoric partners. These results are complex, but fairly intuitive and generally consistent with previous investigations (Coyne, 1976b; Gotlib & Meltzer, 1987, Joiner et al., 1992; Joiner & Metalsky, 1995; Strack & Coyne, 1983) examining the responses that depressed people engender in those with whom they interact. For example, Gotlib and Meltzer found that individuals who interact with depressed people report feeling less socially competent than individuals who interact with nondepressed people. Similarly, Coyne (1976b) found that college students who participated in a 20-min telephone conversation with either depressed or nondepressed psychiatric patients or normal controls responded negatively to the depressed patients, in that the students reported a negative mood following the telephone conversation in addition to an unwillingness to continue the social interaction. According to the Coyne (1976a) interpersonal model of depression, a key and perpetuating aspect of depression is the negative responses that depressed individuals elicit from other people. Perhaps nondepressives, in their interactions with depressed individuals, have to exert more effort than is typically the case or perceive some other negative message(s) from the social interaction, and thus are led to believe that they did not leave a favorable impression to their depressed partners. Yet, it remains unclear why this would not be the case for depressive peoples as well. Dysphoric individuals appeared to anticipate and receive similarly low ratings from their partners, regardless of whether their partners were dysphoric or nondepressed. It is possible that depressive individuals are accustomed to negative social experiences and thus are less sensitive than nondepressive individuals in gauging the reactions of others during social interactions. That is, depressive people generally elicit negative reactions from people in their life, and may interpret the situation negatively regardless of whether they are interacting with a depressed or nondepressed individual. According to cognitive theories of depression, depressed individuals have acquired a negative schema that is, a tendency to interpret the world negatively through one s previous life experiences, such as past criticism from others or loss of a loved one. These negative schemata (such as the beliefs that I am unlikable or People will criticise me if I let them see my weaknesses ) are activated when depressed people enter new situations, and may function as their default social position even though the social encounter is a neutral or even a positive event. Again, although these speculations are contrary to the depressive realism hypothesis, they would account for some of the findings of this study. Several limitations of the present study should be noted. First, participants in this study were university students who were classified as dysphoric or nondysphoric solely on their scores on the BDI. Depressive realism studies have often been criticised for using subclinically depressed (i.e., dysphoric) college students rather than clinically depressed samples (Haaga & Beck, 1995). Although having participants complete the BDI twice rather than once reduced the possibility of misclassifying nondysphoric participants as dysphoric or vice versa, the range of dysphoria in this sample was highly restricted and thus may not generalise to a clinically depressed population. However, these drawbacks are common in investigations such as the present study, which take place in academic settings with little or no access to clinically depressed populations. Furthermore, as Pacini, Muir, and Epstein (1998) have noted, the study of subclinical depression may be a worthy endeavour in its own right. It is likely that the number of people who are subclinically depressed far exceed those who are clinically depressed. Despite external appearances of the subclinically depressed and their deemed adequate levels of functioning, it is clear from their self-descriptions that they are quite

Depressive realism 107 distressed and unhappy. Thus, conducting research to further understand the cognitions and behavioural patterns of the subclinically depressed may provide an insight into their problems and help discover ways to alleviate their difficulties. Additionally, participants in this study consisted of female students only in order to avoid the possibly confounding factor of gender. The results of previous investigations (Beyer, 2002; Kistner et al., 2001; Martin, Abramson, & Alloy, 1984) suggest that there may be an interaction between depression and gender. For instance, Martin et al. suggest that traditional sex role stereotypes may apply only between certain social interactions between female and nondepressed male subjects, but not between female and depressed male subjects. It would be interesting for future investigations to examine these possible differences. Last, all of our statistically significant findings yielded relatively small effect sizes. This may be due to the small sample size in the current study (n ¼ 72), which may be improved with larger samples in similar future investigations. In sum, the results of this investigation do not support the notion that depressive people are sadder yet wiser in perceiving interpersonal phenomena. Research on depressive realism has yielded highly inconsistent results, which suggests that there are instances in which depressive realism occurs and instances in which it does not occur. Clearly, further empirical research is necessary to elucidate these boundary conditions on depressive realism and additionally, the motivational and cognitive processes underlying the phenomenon. Future research on depressive realism should (a) use clinically depressed samples in order to be able to generalise to such populations, (b) use experimental methods that allow for accuracy to be objectively determined, and (c) study depressive realism in meaningful contexts, such as in the area of interpersonal phenomena, that allow for maximum ecological validity (Ackerman & DeRubeis, 1991; Alloy & Abramson, 1988; Alloy et al., 1990; Dobson & Franche, 1989; Haaga & Beck, 1995). Clinicians who work with depressed individuals may wish to help them think more realistically and rationally as illustrated, for example, by Beck s cognitive therapy for depression (1967, 1991), rather than teaching such patients to engage in positive perceptual biases. Research on depressive realism has both basic and applied value. In addition to increasing our understanding of depression, the comparison of depressive and nondepressive cognition may illuminate the functions of nondepressive and depressive biases in human cognition and examine the psychological consequences associated with realism or distortion. These endeavours are themselves worthwhile goals, but they may also aid us in the process of developing treatment strategies for subclinical and clinical depression. Acknowledgment This article is based on a master s thesis prepared by Phuong M. Chau under the direction of Leonard S. Milling. References Ackerman, R., & DeRubeis, R. J. (1991). Is depressive realism real? Clinical Psychology Review, 11, 565 584. Alanazi, F. M. (2001). The revised Self Consciousness Scale: An assessment of factor structure, reliability, and gender differences in Saudi Arabia. Social Behavior and Personality, 29, 763 776. Alloy, L. B., & Abramson, L. Y. (1979). Judgment of contingency in depressed and nondepressed students: Sadder but wiser? Journal of Experimental Psychology, 108, 441 485. Alloy, L. B., & Abramson, L. Y. (1988). Depressive realism: Four theoretical perspectives. In L. B. Alloy (Ed.), Cognitive processes in depression (pp. 223 265). New York: Guilford Press. Alloy, L. B., Albright, J. S., Abramson, L. Y., & Dykman, B. M. (1990). Depressive realism and nondepressive optimistic illusions: The role of the self. In: R. I. Ingram (Ed.), Contemporary psychological approaches to depression (pp. 71 86). New York: Plenum Press. Baron, R. M., & Kenny, D. A. (1986). The moderator mediator distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173 1182. Baumeister, R. F. (1989). The optimal margin of illusion. Journal of Social and Clinical Psychology, 8, 176 189. Baumeister, R. F., Dale, K., & Sommer, K. L. (1998). Freudian defense mechanisms and empirical findings in modern social psychology: Reaction formation, projection, displacement, undoing, isolation, sublimation, and denial. Journal of Personality, 66, 1081 1116. Beck, A. T. (1967). Depression. Philadelphia: University of Pennsylvania Press. Beck, A. T. (1991). Cognitive therapy: A 30-year retrospective. American Psychologist, 46, 368 375. Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). A scale for measuring depression. Archives of General Psychiatry, 4, 561 571. Beyer, S. (2002). The effects of gender, dysphoria, and performance feedback on the accuracy of self-evaluations. Sex Roles, 47, 453 464. Campbell, J. D., & Fehr, B. (1990). Self-esteem and perceptions of conveyed impressions: Is negativity associated with greater realism? Journal of Personality and Social Psychology, 58, 122 133. Colvin, C. B., & Block, J. (1994). Do positive illusions foster mental health? An examination of the Taylor and Brown formulation. Psychological Bulletin, 116, 3 20. Coyne, J. C. (1976a). Toward an interactional description of depression. Psychiatry, 39, 28 40. Coyne, J. C. (1976b). Depression and the response of others. Journal of Abnormal Psychology, 85, 186 193. Davis, M. H., & Franzoi, S. L. (1999). Self-awareness and self consciousness. In V. Derlega, B. Winstead, & W. Jones (Eds.), Personality: Contemporary theory and research (pp. 307 338). Chicago: Nelson-Hall Publishers.

108 P. M. Chau & L. S. Milling Dobson, K. S. (1989a). A meta-analysis of the efficacy of cognitive therapy for depression. Journal of Consulting and Clinical Psychology, 64, 391 399. Dobson, K. S. (1989b). Real and perceived interpersonal responses to subclinically anxious and depressed targets. Cognitive Therapy and Research, 13, 37 47. Dobson, K. S., & Franche, R. (1989). A conceptual and empirical review of the depressive realism hypothesis. Canadian Journal of Behavioral Science, 21, 419 433. Fenigstein, A., Scheier, M., & Buss, A. (1975). Public and private self-consciousness: Assessment and theory. Journal of Consulting and Clinical Psychology, 43, 522 527. Funder, D. C. (1987). Errors and mistakes: Evaluating the accuracy of social judgment. Psychological Bulletin, 101, 75 90. Gilbert, D. T., Pinel, E. C., Wilson, T. D., Blumberg, S. J., & Wheatley, T. (1998). Immune neglect: A source of durability bias in affective forecasting. Journal of Personality and Social Psychology, 75, 617 638. Gilovich, T., & Savitsky, K. (1999). The spotlight effect and the illusion of transparency: Egocentric assessments of how we are seen by others. Current Directions in Psychological Science, 8, 165 168. Gotlib, I. H. (1981). Self-reinforcement and recall: Differential deficits in depressed and non-depressed psychiatric patients. Journal of Abnormal Psychology, 90, 521 530. Gotlib, I. H. (1982). Self-reinforcement and depression in interpersonal interaction: The role of performance level. Journal of Abnormal Psychology, 84, 346 357. Gotlib, I. H., & Meltzer, S. J. (1987). Depression and the perception of social skill in dyadic interaction. Cognitive Therapy and Research, 11, 41 54. Haaga, D. A. F., & Beck, A. (1995). Perspectives on depressive realism: Implications for cognitive theory of depression. Behavioral Research and Therapy, 33, 41 48. Hamilton, D. L., & Gifford, R. K. (1976). Illusory correlation in interpersonal perception: A cognitive basis of stereotypic judgments. Journal of Experimental Social Psychology, 12, 392 407. Harris, M. J., & Rosenthal, R. (1985). Mediation of interpersonal expectancy effects: 31 meta analyses. Psychological Bulletin, 97, 363 386. Joiner, T. E., & Metalsky, G. I. (1995). A prospective test of an integrative interpersonal theory of depression: A naturalistic study of college roommates. Journal of Personality and Social Psychology, 69, 778 789. Joiner, T. E., Alfano M. S., & Metalsky, G. I. (1992). When depression breeds contempt: Reassurance seeking, self-esteem, and rejection of depressed college students by their roommates. Journal of Abnormal Psychology, 101, 165 173. Kendall, P. C., Hollon, S. D., Beck, A. T., Hammen, C. L., & Ingram, R. E. (1987). Issues and recommendations regarding use of the Beck Depression Inventory. Cognitive Therapy and Research, 11, 289 299. Kistner, J., Balthazor, M., Risi, S., & David, C. (2001). Adolescents perceptions of peer acceptance: Is dysphoria associated with greater realism? Journal of Social and Clinical Psychology, 20, 66 81. Kruglanski, A. W. (1989). The psychology of being right : The problem of accuracy in social perception and cognition. Psychological Bulletin, 106, 395 409. Lewinsohn, R. M., Mischel, W., Chaplin, W., & Barton, R. (1980). Social competence and depression: The role of illusory self-perceptions. Journal of Abnormal Psychology, 89, 203 212. Martin, D. J., Abramson, L. Y., & Alloy, L. B. (1984). Illusion of control for self and others in depressed and nondepressed college students. Journal of Personality and Social Psychology, 46, 125 136. Miller, D. T., & Ross, M. (1975). Self-serving biases in attribution of causality: Fact or fiction? Psychological Bulletin, 82, 213 255. Neuberg, S. L. (1989). The goal of forming accurate impressions during social interactions: Attenuating the impact of negative experiences. Journal of Personality and Social Psychology, 56, 374 386. Nystedt, L., & Smari, J. (1989). Assessment of the Fenigstein, Scheier, and Buss Self Consciousness Scale: A Swedish Translation. Journal of Personality Assessment, 5, 342 352. Norton, P. J., & Hope, D. A. (2001). Kernels of truth or distorted perceptions: Self and observer ratings of social anxiety and performance. Behavior Therapy, 32, 765 786. Pacini, R., Muir, F., & Epstein, S. (1998). Depressive realism from the perspective of cognitive experiential self-theory. Journal of Personality and Social Psychology, 74, 1056 1068. Robins, R. W., & John, O. P. (1997). The quest for self-insight: Theory and research on accuracy and bias in self-perception. In R. Hogan, J. Johnson, & S. Briggs (Eds.), Handbook of Personality Psychology (pp. 649 679). San Diego: Academic Press. Robinson, B. E., & Kelley, L. (1996). Concurrent validity of the Beck Depression Inventory as a measure of depression. Psychological Reports, 79, 929 930. Ross, L. (1977). The intuitive psychologist and his shortcomings: Distortions in the attribution process. In L. Berkowitz (Ed.), Advances in experimental social psychology. 10, 174 221. New York: Academic Press. Ross, L., Greene, D., & House, P. (1977). The false consensus effect : An egocentric bias in social perception and attribution process. Journal of Experimental Social Psychology, 13, 279 301. Sacco, W. P. (1999). A social cognitive model of interpersonal processes in depression. In T. Joiner, & J. C. Coyne (Eds.), Interactional nature of depression: Advances in interpersonal approaches (pp. 329 362). Washington, DC: American Psychological Association. Savitsky, K., Epley, N., & Gilovich, T. (2001). Do others judge us as harshly as we think? Overestimating the impact of our failures, shortcomings, and mishaps. Journal of Personality and Social Psychology, 81, 44 56. Sedikides, C., & Skowronski, J. J. (1995). On the sources of selfknowledge: The perceived primacy of self-reflection. Journal of Social and Clinical Psychology, 14, 244 270. Segrin, C. (2000). Social skills deficits associated with depression. Clinical Psychology Review, 20, 379 403. Strack, S., & Coyne, J. C. (1983). Social confirmation of dysphoria: Shared and private reactions to depression. Journal of Personality and Social Psychology, 44, 198 806. Taylor, S. E., & Brown, J. B. (1988). Illusion and well-being: A social psychological perspective on mental health. Psychological Bulletin, 103, 193 210. Tse, W. S., & Bond, A. J. (2004). The impact of depression on social skills: A review. Journal of Nervous & Mental Disease, 192, 260 268. Weinstein, N. D. (1980). Unrealistic optimism about future life events. Journal of Personality and Social Psychology, 39, 806 820. Youngren, M. A., & Lewinsohn, P. M. (1980). The functional relation between depression and problematic interpersonal behavior. Journal of Abnormal Psychology, 89, 333 341.