Abstract
Using a series of multiple regression procedures this study explores the contribution toward explained variance of a substance usage index for several operational variables that have been considered in recent substance intervention and prevention program strategies. Essentially, the study seeks to determine whether or not there is a basis for treatment program stratification according to various adolescent subpopulation categories (age, sex and socioeconomic status). Findings from this study of 3759 adolescents reveal that while demographic features are generally important in predicting substance involvement among these youngsters, there is little to argue for development of separate programs designed to affect them. Except for younger adolescents who display a pattern different from all others, subgroup categories appear to discriminate little among the relative contributions of the operational variables (self image, parental relations, group membership and religiosity) toward explaining substance usage.