Abstract
Two experiments were conducted in order to investigate the relationship between causal attribution and expectancy in failure condition. In Exp. I, subjects were college students. On three achievement-related tasks, the expectancy was lower in the subject who ascribed their failure to stable factors than those who ascribed to unstable factors. After three experiences of failure, most of the subjects came to ascribe them more to stable causal factors. In Exp. II, subjects were junior high school pupils. In the midterm examination, subjects who ascribed failures to stable causal factors expected to get less marks in the next examination than those who ascribed failures to unstable factors. The effects of the attribution to stability dimension on expectancy were confirmed.