Biased probability estimates in trait anxiety and trait depression are unrelated to biased availability.

Abstract Background and objectives People high in trait anxiety or depression overestimate the probability of negative events, and underestimate the probability of positive events, relative to people low in trait anxiety and depression. Although this probability bias may be fundamental to some emotional disorders, its causes are not well understood. The dominant explanations are based on the availability heuristic: people relatively high in anxiety or depression find it relatively easy to imagine reasons why bad things might happen to them, and this affects their probability estimates. We tested, for the first time, whether individual differences in the availability of such reasons mediate the relationships between trait anxiety or depression and probability bias, in a nonclinical sample. Methods Two hundred and seventy-eight undergraduates generated reasons why a set of positive and negative events might vs. might not happen to them, before rating those events’ probability and potential impact on their lives. Results Individual differences in the availability of reasons why good and bad events might vs. might not happen did not mediate the sizeable relationships between trait anxiety and probability bias, and between trait depression and probability bias; these relationships remained significant when availability was controlled. Results for the impact of events (‘cost bias’) were less clear. Limitations Replication with patient groups would be invaluable; different operationalisations of availability may change the results. Conclusions Availability can influence probability estimates, but it does not explain why we see probability bias in people with high trait anxiety or depression.