Controllability Attributions and Learned Helplessness: Some Methodological and Conceptual Problems

The attribution literature related to the reformulated model of learned helplessness theory was examined, with particular attention paid to the methodological/procedural aspects of such research. The extent to which controllability attributions were examined within the literature and whether appropriate statistical and procedural methods was employed to control for confounded variance among attributional dimensions were of central interest. Literature published between January 1978 and June 1990 was examined. Of 134 studies examined, only 12% measured controllability attributions. In contrast, the majority of studies (over 50%) focused exclusively on attributional dimensions important to the reformulated model. Methodological/procedural and data-reporting concerns that were identified included inattention to statistical controls to prevent confounding of variance among dimensions and incomplete reporting of correlational matrices necessary for meta-analyses. Studies that examined controllability are reviewed, and implications of the findings are discussed.