Abstract

The purpose of the present study was to examine the potential causal role of interpretive bias in cognitive vulnerability to generalized anxiety disorder and its primary symptom, worry. An interpretive bias induction paradigm developed by Mathews and Macintosh (2000) was used to modify participants' interpretations of ambiguous scenarios. Sixty-nine (69) individuals were randomly assigned to either the negative induction group ( n = 35) or the positive induction group ( n =34). Following training, participants completed two measures of intolerance of uncertainty (IU), a cognitive vulnerability factor implicated in worry, and an interview related to processes involved in worry. Among the two measures of IU, one was a self-report questionnaire measuring explicit beliefs about uncertainty, while the other was a computerized task designed to assess automatic threat associations related to uncertainty. It was hypothesized that compared to the positive induction group, the negative induction group would evidence: (1) more explicit negative beliefs about uncertainty, (2) stronger automatic associations related to uncertainty, and (3) higher levels of worry. Results revealed that interpretive bias was successfully induced, but did not lead to group differences on IU or worry. In contrast to previous studies (Mathews & Macintosh, 2000), no effect of the training on state anxiety was found. Potential explanations for the discrepant findings are discussed as well as treatment implications for interpretive bias modification during therapy.