(PS1-36) Bidirectionality Between Post-event Processing and Shame in the Context of Social Anxiety

Abstract Text: This study examines the bidirectional relationships between post-event processing (PEP) and shame. According to the Clark and Wells Model (1995) of social anxiety, PEP occurs when an individual engages in a detailed, repetitive, and negative review of a recent interaction with others. Researchers have argued that PEP’s role in the social anxiety cycle could explain why social anxiety is more resistant to exposure treatments than most other anxiety disorders (Brozovich & Heimberg, 2008). Shame occurs when an individual engages in aversive self-appraisals, involving a high level of negative self-reflection and self-evaluation, following an event that involved wrongful behavior, perceived transgressions, or negative consequences (Tangney, 1996). Research has demonstrated that shame and social anxiety are associated with one another, but more information about the interplay between shame and PEP is needed.

Students (N = 123) from an introductory psychology course engaged in an in-lab 10-mimunte social interaction task. They were asked to get to know another study participant who was actually a trained study confederate blind to the study’s hypotheses. Following their social interaction, participants watched a boring video while providing assessments of their state shame and state PEP levels every two minutes. A random intercept cross-lagged panel model (RI-CLPM; Hamaker et al., 2015) was used to assess bidirectionality between state shame and state PEP during the video-watching task. A five-wave RI-CLPM was used to test for bidirectionality between state shame and state PEP. Each observed score of state PEP and state shame was regressed onto its own latent factor, with two primary random intercept factors for our two variables. All factor loadings were constrained to 1 and observed-score error variances were constrained to 0. The proposed RI-CLPM was adjusted for the absence of significant change in shame at wave 5. A modified RI-CLPM demonstrated good fit; χ2 (25) = 51.91, p < .001; CFI = .984; TLI = .972; RMSEA = .095; SRMR = .043. Over the course of the video-watching task, both variables demonstrated a steady decrease. The modified RI-CLPM demonstrated that positive deviations in state shame at the 0-minute mark predicted substantially slower decreases in state post-event processing from the 0-minute to the 2-minute mark. Though some of these effects were short-lived as positive deviations in state shame at the 2- and 6-minute marks predicted slightly faster decreases in subsequent state PEP. The effect of shame on PEP was initially large, and the overall effect of initial shame on subsequent PEP appeared to progressively diminish over time. Positive deviations in state PEP at the 2-minute mark predicted slower decreases in state shame from the 2-minute to the 4-minute mark, demonstrating a modest effect of PEP on shame. Taken together, these findings indicate that in the moments after an anxiety-provoking social interaction, shame and PEP bolster one another. Afterwards, however, greater state shame led to faster-than-expected decreases in state post-event processing. The implications of these findings for theory and practice are discussed (e.g., shame’s role in social anxiety should be further addressed in social anxiety treatments).