Hull J G, Mendolia M
Department of Psychology, Dartmouth College, Hanover, New Hampshire 03755.
J Pers Soc Psychol. 1991 Jul;61(1):85-97. doi: 10.1037//0022-3514.61.1.85.
Structural modeling techniques were used to assess relations of attributional style, expectancies, and depression. According to an initial theoretical model, attributions are directly related to expectancies, and expectancies are directly related to depression, but attributions are only indirectly related to depression by means of their relation to expectancies. The results of Study 1 indicated that this model was flawed in 2 respects: (a) Attributions for positive and negative events did not form a single latent variable, and (b) attributions for negative events both were indirectly related to depression by means of expectancies and were directly related to depression. Attributions for positive events only were indirectly related to depression by means of expectancies. The model derived in Study 1 was replicated in Study 2. Discussion centers on the interpretation of this modified model and on issues in the measurement of attributional style.
