Perceived self-efficacy and beliefs about the locus of outcome causality must be distinguished

The difficulty in judging what type of behavior works well arises not only because a given course of action does not always produce the outcomes. Similar outcomes can occur for reasons other than the person's actions, which further complicates inferential judgment. Effects that arise independently of one's actions distort the influence of similar effects produced by the actions, but only on some occasions. Given a strong cognitive set to perceive regularities, even chance joint occurrences of events can be easily misjudged as genuine relationships of low contingent probability

Moreover, joint occurrences tend to be better recalled than instances when the effect does not occur. The proneness to remember confirming instances, but to overlook disconfirming ones, further serves to convert, in thought, coincidences into causalities.

If you look at our theories of social pathology and then at the dismal conditions in which children grow up in our ghettos, you would predict that all of them would be on drugs or psychological basket cases. Yet if you use criteria like gainful employment, forming partnerships and life without crime, you will find that most of those kids make it.

People who are burdened by acute misgivings about their coping capabilities suffer much distress and expend much effort in defensive action... they cannot get themselves to do things they find subjectively threatening even though they are objectively safe. They may even shun easily manageable activities because they see them as leading to more threatening events over which they will be unable to exercise adequate control.

For many activities, people cannot rely solely on themselves in evaluating their ability level because such judgments require inferences from probabilistic indicants of talent about which they may have limited knowledge. Self-appraisals are, therefore, partly based on the opinions of others who presumably possess evaluative competence

Regression analyses show that self-efficacy contributes to achievement behavior beyond the effects of cognitive skills