Diagnosis of alcoholism: the Munich Alcoholism Test (MALT).

The purpose of our project was to develop and empirically test a diagnostic instrument which would permit sufficiently reliable differentiation between alcoholics and healthy as well as sick persons and which would also be easy and economical to administer. For this purpose, roughly 250 diagnostically relevant items were selected from the extensive literature on alcoholism and, with a total of 1335 patients (661 of these were alcoholics), were evaluated in three separate phases for their ability to differentiate. The best items were selected on the basis of various statistical criteria and then cross-validated. The result is the Munich Alcoholism Test (MAT) which consists of two complementary parts: a 7-item physician's assessment part and a 24-item self-assessment part. In a cross-validation study on 675 unselected in- and outpatients a validity of r = 0.94 was obtained. The reliability of the self-assessment part was r = 0.84.