Structure of the MAST with an Australian sample of alcoholics

The dimensionality of the Michigan Alcoholism Screening Test (MAST) was examined utilizing an Australian sample (n = 196) of alcoholics. This study was specifically designed to overcome some of the problems with sampling and analytic techniques used in previous studies. Three moderately intercorrelated factors were revealed which covered (a) alcohol-related disabilities; (b) help-seeking behaviour and (c) recognition of a drinking problem. These factors were found to be internally consistent as shown by moderate to high alpha coefficients. Results indicate that the additive system of scoring all 24 items may result in an underestimation of the information afforded by the MAST