Trained interpretive bias: validity and effects on anxiety

Mathews and Mackintosh [(2000). Induced emotional interpretation bias and anxiety. Journal of Abnormal Psychology, 109, 602-615] developed a clever training procedure that enables the investigation of a causal relationship between interpretive bias and anxiety. The present study examined the validity of this paradigm by testing (1) the effects of interpretation training on two other tasks (homograph EAST and open-ended questionnaire) that are less closely related to the interpretation training itself as in previous studies and (2) the robustness of the training effects on state and trait anxiety. Results indicated that while the two original dependent measures (i.e., a reaction time and recognition measure) showed that the training procedure was successful in changing interpretations, the two additional measures (i.e., EAST and questionnaire) did not. This might reflect a measurement artefact, but other explanations for the findings are also possible. Moreover, evidence was found for effects of biased interpretations on anxiety. This demonstrates the viability of the present paradigm, which has implications for clinical practice.