THE ROLE OF GENERALISED EXPECTANCIES IN DETERMINING CAUSAL ATTRIBUTIONS FOR SUCCESS AND FAILURE IN TWO SOCIAL CLASSES

Summary . The sources of discrepancies in causal attributions for success and failure and for self and other attributions of outcome may become clearer when specific research attention is paid to the role of generalised expectancies in the attributional process. The present study hypothesised that when information about task and outcome is standardised, the attributions of both actors and observers will tend to reflect generalised expectancies, evoked in this study by two levels of SES in both actor and observer conditions, more than they will self-serving biases. 230 sixth grade Jewish Israeli pupils of two SES levels were assigned to one of three experimental conditions. All pupils received eight anagrams, four soluble by all and four insoluble. Those in the self-attribution condition attributed their own outcomes to various causes, while those in the similar and different other conditions attributed the same outcomes for pupils of apparently similar or different social class, after having first completed the anagrams themselves. The results indicated that while pupils of high SES tended to attribute both their own and others' outcomes in ways consistent with high generalised expectancy for success, pupils of low SES attributed their own outcomes more to external, unstable factors, and differentiated consistently in their attributions for the self, for similar and for different others. It was argued that these differences reflect undifferentiated, global perceptions of causality among high SES pupils, whose patterns of attribution are consistent both with teacher values and their own experience. Low SES pupils have more differentiated perceptions of causality since uncertainty as to the real causes of their learning outcomes motivates them to greater, but not always adaptive, attributional activity.