The expectancy confirmation and egotism models of causal attributions for success and failure were investigated with 400 fifth form students taking the New Zealand School Certificate English examination. Attributions for expected exam outcome and actual exam outcome were assessed with a scale containing 33 items categorised into the attribution dimensions of stability and locus of causality. The results suggested that exam outcome attributions are a function of both expectancy confirmation and valence of actual outcome, with the effects being due primarily to subjects whose expectations for passing were disconfirmed by actual failure. Compared to pre-exam attributions based on expected outcome, these disconfirmed failure subjects reduced post-exam attributions to internal (stable and unstable) and external stable causes, but increased their attributions to external unstable causes. Whereexpectations wereconfirmed (pass or fail) attributions from pre-to post-exam were generally consistent; however, the confirmed pass subjects tended to see internal factors as more important than the other subjects, and external unstable factors as least important. The findings are discussed in terms of the implications for future achievement situations. 1971 ;Feather and Simon, 1973;McMahan, 1973;Valle and Frieze, 1976). Concomitantly, the greater the discrepancy between expected and actual outcome, the stronger the attribution to unstable factors and the smaller the discrepancy, the greater the attribution to stable factors.
Exam AttributionsDavis and Stephan (1980) take issue with this expectancy confirmation approach. They assert that the valence of the actual outcome (success or failure) is the primary determinant of post-performance attributions. Thus, attributions to internal factors (ability, effort) will increase with the magnitude of success in the outcome, whereas attributions to external factors (task difficulty, luck) will increase with the degree of failure. Davis and Stephan refer to this interpretation in terms of egotism, whereby failure is ascribed to external factors for which the individual cannot be held responsible (ego-defensive), and success is ascribed to internal factors (ego-enhancement).In support of this position, Davis and Stephan (1980) present data from a study with undergraduate university students. They found that the students generally had high expectations for success on a class exam, and that these were associated with ability and effort attributions. Actual exam outcomes consistent with those high expectations were again attributed internally to ability and effort, whereas low performance was attributed externally to test difficulty. However, in a regression analysis of the place of ability, effort, test difficulty and luck in predicting the difference score between expected result and actual result, Davis and Stephan found little or no relationship between those causal factors and degree of expectancy confirmation. Thus, it was concluded that the egotism model emphasising locus of causality wa...