Abstract
In a wide variety of cognitive tasks, people's expectations of what their own performance is, was, and will be are an overestimate of reality. After documenting a number of situations in which people exhibit such overconfidence, several reasons for it are discussed. It is suggested that in these situations, the optimistic bias is not attributable exclusively to self-deception or wishful thinking. Rather, the information yielded up by the cognitive system, in combination with the heuristics used for making judgments of future and past performance tailored to the specific questions asked of participants, produces a bias toward believing that one knows, knew, and will perform better than actual performance substantiates. Consequently, in the cognitive domain, the inflated beliefs that result in overconfidence also result in cessation of efforts before the correct solution of problems is ascertained, before accurate retrieval of memorial information is attained, or before adequate learning of new material has been accomplished. This effect seems in contrast to findings on people's moods and self-esteem in real-world or threatening situations that suggest that an optimistic bias in these domains be person protective and adaptive.

This publication has 71 references indexed in Scilit: