Changes in feared outcome expectancy and relief during exposure for public speaking anxiety

Learning by updating expectancies for aversive events is a proposed mechanism of exposure therapy. Reward processes and positive emotions, such as relief, may be closely tied to this learning process, though this has not been explicitly studied in an exposure therapy setting. Participants were UCLA college students who completed two sessions of exposure to public speaking (16 exposures total) and three public speaking anxiety assessments (baseline, post-assessment, one-week follow-up). Additionally, participants completed rehearsal exercises after exposures and were randomized to one of two rehearsal conditions: positive rehearsal designed to enhance reward processes and positive emotions (n = 42), or neutral rehearsal (n = 42). Self-reported expectancy for the feared outcome (e.g., social rejection) and relief that the expected feared outcome did not occur were measured across exposures using both observed and computational modeling indices. Public speaking anxiety decreased from baseline to follow-up, with no differences between groups. Observed expectancy ratings decreased across exposures at a faster rate for the neutral than the positive rehearsal group, with no group differences in computational modeling of learning rate. Observed relief ratings significantly decreased across exposures for the positive group only, with no group differences in computational modeling of relief rate. Faster expectancy-based learning rate and higher expectancy-based intercept at the first exposure of two exposure sessions predicted lower public speaking anxiety at follow-up. Higher intercept of relief at the first exposure of two exposure sessions, but not relief rate, also predicted lower public speaking anxiety at follow-up. Findings support learning via expectancy updating as an important process in exposure therapy and highlight a need for further study of relief during exposure.