Sadder but wiser: A failure to demonstrate that mood influences judgements of control.

A replication and extension of one of Alloy and Abramson's (1979) studies yielded no evidence that mood influences judgements of control in noncontingent button-pressing tasks. It was found that (a) overestimates of control when outcomes are frequent may be due to selective inattention to and/or poor recall of how often outcomes occurred when no response was made; (b) comparable over- estimates of control also occur when outcome frequency is low and noncontingency is detected; (c) such overestimates likely do not reflect a self-serving bias, since students in the low density outcome condition were critical of their own performance. The results suggest that the illusion demonstrated in contingency learning studies may be related to expectations of control and to a common tendency to confirm such expectations when presented with evidence of causality (i.e., co-occurrences) between at least some response-outcome pairings. The implications for future re- search in this area are discussed, and the possibility is entertained that the depressed are not wiser, but more prone to self-attribut ions of incompetence in certain contingency learning tasks.