Cognitive models of social anxiety (SA) have posited that individuals elevated in SA exhibit a bias to interpret neutral social situations as threatening. While social-evaluative interpretation bias (IB) has been extensively studied, little research has examined the relationship between SA and hostile IB, a specific form of IB characterized by the interpretation of ambiguous stimuli as indicating hostile intent, meanness, or aggression. In Study 1 (N = 120), we found in a sample of college students that while SA was associated with hostile IB cross-sectionally, it did not predict hostile IB one month later, while controlling for baseline hostile IB. In Study 2 (N = 69), we examined the effect of a one-month safety behavior fading intervention for SA, wherein participants were asked to systematically reduce their use of SA safety behaviors. Relative to a waitlist control, this one-month safety behavior fading intervention led to lower hostile IB at posttreatment among individuals elevated in baseline SA. We also found that changes in SA mediated the effect of treatment on hostile IB. Overall, the present study provides preliminary evidence suggesting that SA maintains hostile IB, and that hostile IB may be an outcome worth investigating in future intervention studies pertaining to clinical SA.
Keywords: Social anxiety; anger; hostile interpretation bias; safety behavior.