Exploring the Relationship Between Social Anxiety and Emotional Expression Recognition Through Meta-Analytic, Psychometric, and Experimental Methods

Abstract

Existing models of social anxiety (SA) have proposed that this disorder is, in part, maintained by an attentional bias towards threat cues. Cues that may signal negative evaluation in social situations are attended to more quickly and interpreted more negatively, and therefore avoided. Among these cues are emotional facial expressions. While many studies have investigated the relationship between SA and emotional expression recognition (EER), there is a lack of consensus regarding whether global or emotion-specific EER deficits exist within this population. Few studies have explored what factors may be modulating the relationship between EER and SA.

The work presented in this thesis used meta-analytic, self-report, psychometric, and experimental methods to examine the effects of several factors on the EER/SA relationship. Across the three studies, we systematically reviewed and meta-analyzed the current literature on EER and SA while exploring the impact of comorbidity and stimulus duration (Study 1), investigated how communication medium impacts perceived social skill abilities--including emotion decoding-- in people with SA (Study 2), and sought to assess the effects of an experimentally induced state of anxiety on EER performance (Study 3). Altogether, we observed within and between-subjects emotion-specific and global EER differences. The results from Study 1 revealed particularly poor performance on EER of neutral and happy facial expressions in people with SAD, the latter of which is affected by comorbidity but not stimulus duration. Study 2 highlighted that communication medium (online versus in-person) has a prominent influence on perceived social skill abilities. In Study 3, we observed no influence of our manipulation on state anxiety, but the results nevertheless partially supported our previous findings. Our results suggested that individuals with high trait SA have significant difficulties recognizing neutral facial expressions. Together, these results uniquely broaden our understanding of the association between SA and EER, and how various relevant factors impact their relationship.