This project will provide a meta-analytic review of alcoholism treatment research published in English between 1970 and 1989. It will have three interrelated foci: examining treatment effects, identifying patient prognostic indicators, and examining patient-treatment matching effects. Analyses of studies in each of these three areas will proceed in six stages: (a) describing the characteristics of studies, (b) aggregating significance levels across studies; (c) computing individual indices of effect magnitude, (d) testing the homogeneity of effect magnitudes, (e) computing and testing the significance of average effect magnitudes, and (f) accounting for variation in effect magnitudes across studies. The review will reveal the topography of alcoholism treatment research over the past 20 years, including the extent to which such research incorporates the features of state-of-the-art treatment research and whether the state-of-the-art is being more closely approximated over time. In addition, the review will determine (a) whether treatment effects estimates vary as a function of research design, outcome dimension tapped, measures used, etc.; (b) the extent to which research has been underpowered and thus has yielded misleading no-difference findings in comparisons of different treatment modalities; and (c) the effect sizes and levels of during-treatment and during-study patient attrition that are likely with different treatment approaches. Substantively, the review will determine (a) the quantitative effects of treatment versus no minimal treatment; (b) the relative effects of different treatment approaches; (c) the extent to which treatment effects decay over time; (d) the extent to which uncontrolled patient population differences across studies may have led to misleading conclusions in prior reviews regarding the superiority of specific treatment modalities; (e) the patient characteristics that are predictive of positive outcome in general, as well as within specific treatment modalities; (f) an equation a la Costello (1980) for predicting treatment outcome given basic information on patient and treatment characteristics; and (g) the effects of patient-treatment matching efforts.
Showing the most recent 10 out of 11 publications
