LAUSR

Abstract It has been suggested that relapse after exposure therapy for social anxiety disorder is related to the renewal effect. This study aimed to investigate whether the renewal effect occurs in fear conditioning utilizing social stimuli and the quantitative predictions obtained by the Rescorla-Wagner and Bouton’s model using Bayesian modeling fit their data. A total of 63 students participated in this study and received differential fear conditioning, extinction, and testing. For participants in Group AAA, a single context (context A) was used during all the phases. For those in Group ABA, the acquisition and test phases were conducted in the same context (context A), whereas the extinction phase was completed differently (context B). This experiment revealed that the renewal effect was observed in the expectancy rating, but not the valence rating. Bayesian modeling demonstrated that both predictions were generally consistent with the data. These results suggest that relapse of social anxiety is related with the renewal effect in Pavlovian fear conditioning. These findings might provide some implications for the improvement of exposure therapy.