Cognitive biases in social anxiety and perfectionism.

Introduction: Cognitive models of social anxiety (SA) emphasise the importance of the discrepancy between one’s standards for performance and one’s perceived ability to meet such standards in SA. This discrepancy, which directly relates to the concept of perfectionism, has been shown to be cross-sectionally related to SA. However, only few studies have attempted to determine the directionality of this relationship, with the most compelling evidence to date showing that SA influences perfectionism (Gautreau et al., 2015). Furthermore, theoretical propositions and empirical findings have demonstrated the importance of cognitive biases to SA and perfectionism. However, no studies have simultaneously investigated the impact of multiple cognitive biases on SA and perfectionism. The present study sought to address these gaps in the literature by focusing specifically on the mediating effect of interpretation bias, negative self-imagery, and post-event processing on SA and perfectionistic concerns. Method: One hundred and twenty-two participants (43 males and 2 unknown, M age = 27) completed a series of online questionnaires (i.e. the Social Phobia Inventory; The Frost Multidimensional Perfectionism Scale; the Multidimensional Perfectionism Scale; The Interpretation and Judgmental Questionnaire; the Appraisal of Social Concerns questionnaire; and the Extended Post-Event Processing Questionnaire) at three time points, placed three months apart each. Results: A measurement model including SA at Time 1, cognitive biases at Time 2, and perfectionistic concerns at Time 3 was tested using SPSS AMOS v.25. Results revealed acceptable model fit (p = .87, χ2/df < 1, CFI > .95, and RMSEA = 0). Results showed that Time 1 SA had a direct effect on Time 2 interpretation bias (β = .44, p <.001), negative self-imagery (β = .46, p <.001), and post-event processing (β = .36, p <.01). In addition, Time 2 interpretation bias had a direct effect on Time 3 perfectionistic concerns (β = .29, p <.01) and SA (β = .20, p <.05). However, Time 2 negative-self imagery and post-event processing only showed significant effects on Time 3 SA but not perfectionistic concerns. Moreover, no significant direct effect of SA on perfectionistic concerns across time was found. Discussion: Our results suggest that SA may lead to an increase in interpretation bias, which in turn, an increase in perfectionistic concerns. The nature of social interactions is inherently ambiguous and requires individuals to assess their own performance based on perceived feedback from social partners. Our data indicates that SA may affect individuals’ assessments, leading to negatively biased interpretations of performance and feedback. In turn, these negatively biased interpretations of performance and feedback may add one’s perceived inability to meet social standards, thus increasing perfectionism.