The Attribution Cube and Reactions to Educational Outcomes

Hypotheses derived from Kelley's cube model of causal inferences were tested in an educational setting by relating high and low students' attributions to reported affect and expectations. In general, low and high scoring students clustered in different cells of the cube, with high scoring students claiming their scores were nondistinctive, but low scorers emphasizing (a) distinctiveness, (b) high consensus and consistency, or (c) low consensus and consistency. Predictions concerning estimates of distinctiveness, consistency, and consensus and their relation to affect and expectations were also supported, since the cube model appeared to adequately summarize sources of causal information found in the classroom.