Abstract
Prediction of dropout from therapeutic community treatment for heroin addiction is of interest, given that premature termination is the norm. Previous literature indicated a negative association between elevation on various MMPI scales and dichotomized measures of retention, as well as other outcome indices. The present study examined the predictive relationship between pretreatment MMPI scores and a continuous measure of retention. Results indicated that the MMPI could significantly differentiate groups staying less than and more than 20 d in treatment, but had no power to predict to more precise ranges of retention.