Facial crowds of emotion connoting approval or criticism are linked to the fears of socially anxious individuals. We examined evaluation ratings and decision latencies of mixed facial displays by individuals with generalized social phobia (GSPs, n=18n=18), individuals with comorbid depression and GSP (COMs, n=18n=18), and normal controls (CONs, n=18n=18). First, we postulated that GSPs will assign more negative ratings to predominantly disapproving audiences as compared to CONs, and that GSPs will be faster in their evaluation of these audiences (negative bias hypothesis). Second, we expected depression, but not social anxiety, to be associated with diminished positive evaluation of audiences containing predominantly happy expressions and with a slower processing of such positive cues (the impaired positivity hypothesis). Results supported the negative bias hypothesis, and provided partial support for the impaired positivity hypothesis. The importance of examining the processing of complex non-verbal cues in social anxiety and in depression is discussed.
Cognitive theories have emphasized the role that cognitive biases in general, and evaluation biases in particular play in the maintenance of anxiety (e.g., Beck & Emery, 1985; Williams, Watts, MacLeod, & Mathews, 1997). It is hypothesized that evaluation and interpretation biases are central to the maintenance of anxiety in general, and of social anxiety in particular (Foa & Kozak, 1985). Indeed, multiple studies have examined the role of evaluation biases in social anxiety and social phobia (see Heindrich & Hofmann, 2001). Most of this literature has focused on the evaluation of the probability and cost of social events for socially anxious individuals (e.g., Amir, Foa, & Coles, 1998; Gilboa-Schechtman, Franklin, & Foa, 2000). Results have indicated that highly socially anxious individuals (HSAs) tend to interpret mildly negative social scenarios as more likely and more “costly” than low socially anxious individuals (LSAs).
Only a few studies investigated whether social anxiety is associated with a negative bias in the evaluation of non-verbal information. So far, research has focused on the examination of individually presented emotional expressions (e.g., Lundh & Ost, 1996; Winton, Clark, & Edelmann, 1995). In fact, we are not aware of any studies to date examining the evaluation of multi-facial displays (i.e., facial crowds) by clinical populations. Like individual facial expression, multi-facial displays are ubiquitous in social interaction and thus represent ecologically valid stimuli. In fact, performing in front of an audience is frequently rated as the most challenging, most anxiety provoking, and most avoided of all social interactions. Thus, multi-facial displays of emotion, connoting approval or criticism are likely to be linked to the fears and anxieties of most socially anxious individuals. Second, multi-facial displays are likely to contain mixed messages—some members of the audience may seem pleased, others—bored, and still others may display signs of contempt or disapproval. Mixed displays require individuals to integrate information of conflicting valence, whereas the evaluation of facial expressions used in previous research did not require such integration, since they were constructed to be restricted to connote a single affective signal. The examination of the interpretation of mixed multi-facial displays was the first goal of the present study.
Another unresolved issue stemming from the existing literature on the evaluation biases in SPs is the evaluation of positive information. Several studies have found that in addition to biased interpretation of negative events, SAs are also characterized by a failure to positively interpret ambiguous events or by a tendency to see the negative implications of positive events (e.g., Gilboa-Schechtman et al., 2000; Wallace & Alden, 1997). Of relevance to the question of positive bias in interpretation is the comorbidity between social phobia and depression (e.g., Lecrubier & Weiller, 1997). Indeed, diminished emotional responding to pleasant stimuli is frequently found in depression (e.g., Clark & Watson, 1991; Sloan, Strauss, & Wisner, 2001). However, the examination of the contributions of social anxiety and depression to the evaluation of positive non-verbal displays has not yet been undertaken. Such an examination was the second goal of the present study.
Two main hypotheses were examined. First, consistent with previous findings, we postulated that all SPs would be negatively biased in their evaluation of ambiguous multi-facial displays (the negative bias hypothesis). Moreover, following Mogg and Bradley’s (1998) cognitive-motivational account of anxiety, we expected the relationship between SA and ratings of disapproval of the crowd to follow a quadratic function. The cognitive-motivational account posits that anxiety is associated with a lower threshold of threat appraisal. Following this logic, fewer threat elements need to be identified before a “threat” decision is reached, thus, speeding the decision process. Consequently, we expected the decision latencies of GSPs regarding facial displays to be inversely related to the crowd’s level of disapproval.
Second, based on the hypothesis of diminished sensitivity to pleasant stimuli in depression, we expected depression, but not SA, to be associated with diminished positive evaluation of audiences containing predominantly happy expressions (the impaired positivity hypothesis). We also anticipated that depression, but not SA, will be related to a higher threshold for positivity judgments, and thus to a slower processing of positive cues.
