Perceptions of stressful life events and depression: a test of attributional models.

A brief review of tests of the attributional model of depression suggests that there is only weak or inconsistent support for the predicted causal ascriptions by depressed persons for negative events. Moreover, in very few studies have researchers actually tested the causal predictions of the model. Finally, because many of the studies were conducted with normal college students, the generality of the model as applied to different populations needs to be investigated. We addressed three questions: (a) the validity of the hypothesized independent and direction relation between each of the dimensions of internality, stability, controllability, intentionality, and globality and depression; (b) the causal relation between attributions and depression in a 2-month prospective study; and (c) evaluation of the model on two disparate samples: normal college students and elderly depressed outpatients ascribing causes for their personally stressful events. Causal modeling statistical procedures were applied to both the question of concurrent relations and causal relations between cognitions and depression. The results suggest minimal support for the attributional model: The dimensions were not each independently and directly associated with depression in the manner predicted by Abramson, Seligman, and Teasdale (1978), and the model that best fit the data was generally congruent for both the normal and clinical samples. In terms of direction of causality, the data were more consistent in indicating that depression causes cognitions than in indicating the reverse.