The persistent use of negative affect by anxious individuals to estimate risk.

Three experiments investigated how trait anxiety would influence individuals' assumptions about the relevance of their experiences of state anxiety for judgments of risk. Experiment 1 found that attributions of state anxiety to a judgment-irrelevant source reduced the risk estimates of low, but not of high, trait-anxious individuals. The results of Experiment 2 suggest that attribution manipulations reduce the influence of state affect on judgment only when the state affect is inconsistent with participants' trait affect. Experiment 3 revealed that these effects can be controlled by explicitly manipulating participants' assumptions about the relevance of their feelings. Regardless of the level of trait anxiety, attributions were effective at reducing mood effects when facts, but not feelings, were assumed to be the relevant basis for judgment. Overall, the results suggest that trait-consistent affect is more readily assumed to be informative and hence is more likely to be relied on than trait-inconsistent affect.