EXPECTANCIES AND ATTRIBUTIONS FOR JUNIOR HIGH SCHOOL EXAM GRADES

Abstract

The prediction and causal postdiction of midterm exam grades were investigated in a controlled field study. It was hypothesized that: (a) relatively expected outcomes would be ascribed more stably than relatively unexpected outcomes, (b) a positive relationship would obtain between pretest judgments of ability, task, and studying and pretest expectancy, (c) greater causal attributions to perceived stable attributes would result in a positive relationship between exam grades and posttest expectancies, and (d) the relatively expected outcome would be ascribed more to teacher power and teacher sentiment than the relatively unexpected outcome.;Three days before taking a junior high school midterm science exam, subjects completed a questionnaire designed to measure pretest judgments of ability, task, studying, and expectancy. After receiving false exam grade feedback, subjects completed a questionnaire designed to measure the perceived value of the exam grade, attributions to ability, task, effort, luck, studying, teacher power, and teacher sentiment; and expectancy for the final exam grade.;The data supported the first two hypotheses (a & b). The relatively expected outcome was ascribed more stably (and more internally) than the relatively unexpected outcome. Pretest judgments of ability accounted for 27% of the variance in pretest expectancies. Hypothesis c received suggestive support. The data indicated that the postulated mediational effects of the stability dimension of perceived causality upon the relationship between exam grades and posttest expectancies may be moderated by pretest expectancies. Hypothesis d received no support. These findings were discussed in relation to the information processing model and the balance model of causal attribution. It was concluded that both models must be integrated in order to explain the attributional process. Examination of sex differences in expectancies revealed no differences; but males judged themselves to be higher in ability, and females compensated for their perceived lower ability by planning to study more. The data were also investigated for the existence of motivational influences and sex differences in attributions. None were found. Additionally, the perceived value of outcome was examined in relation to the theoretical viewpoints of Helson and Atkinson.;The findings were applied to Seligman's model of learned helplessness and the social learning construct of I-E. An attributional alternative stressing contextual variables was suggested for the model of learned helplessness and the construct of I-E.