Screening tests and self-identification in the detection of alcoholism.

Abstract

Since screening tests for the detection of pathological conditions such as alcoholism frequently are developed and validated on known groups in which the subjects are identified, and identify themselves, with the condition, the question is raised as to whether or not the screening device in such cases reflects the subjects' identification as alcoholics rather than, or as well as, the alcoholic condition. This question is considered with reference to a particular screening device -the Michigan Alcoholism Screening Test (MAST). The test was administered to two groupings of alcoholics from alcoholism wards (the self-identified alcoholics) and a grouping of matched alcoholics from general psychiatric wards (non-self-identified alcoholics). It was hypothesized that if the MAST truly reflected self-identification as alcoholics then the selfidentified alcoholics, relative to the other alcoholics, would (1) score significantly and appreciably higher on the MAST and (2) respond positively with a significantly greater frequency to the specific MAST items that were independently judged (by two of three raters) to reflect alcoholic self-identification. Both hypotheses were supported. The theoretical, methodological, and policy implications of these findings are considered with reference to the development and use of screening devices for the detection of alcoholism and other pathological conditions. ANOTHER aspect of the continuing discussion of the role of the labeling