ADOLESCENT SOCIALIZATION AND DIFFERENTIAL RESPONSES OF HIGH RISK ADOLESCENTS TO A SCHOOL-BASED, SUBSTANCE ABUSE PREVENTION PROGRAM (DSM-III, DRUG, GROUP COUNSELING)

Two hundred forty-six high risk young adolescents referred to a substance abuse prevention program completed the "A.B.U.S.E. Scale," a self-report instrument devised by the investigator to assess socialization, peer attachment, and past and anticipated future pro-social, substance abuse, environmental risk, and problem-behavior involvement. Discriminant analysis as a classification technique documented that socialization, as operationalized in the DSM-III, is a dichotomous, psychological variable with definable demographic, attitudinal, and behavioral components: that socialized and undersocialized young adolescents are distinct subgroups within a population at high risk for both problem-behavior and substance abuse involvement. Stepwise discriminant analysis (Wilks method) yielded a 10-variable, statistically significant discriminant function, with a hit rate of 76%. In a pre-test/post-test, experimental/non-equivalent control group design, a school-based prevention intervention program, involving social and refusal skill training, had an overall positive effect on young adolescents' intentions to get into serious trouble in school, intentions to use drugs in the future, and tendency to utilize peers as a support system. Undersocialized young adolescents, as well as those with an acknowledged drug history, appeared to derive the most benefit from group counseling. Stepwise discriminant analysis also accurately predicted direction or lack of change for 80% of the experimental group using 15 variables, including socialization, on the "school trouble" criterion measure, and for 89% of this group, using 14 variables, on a combined criterion measure. The control group hit rate on "school trouble" was 95%, utilizing 14 variables, including peer attachment, but not socialization: 19 variables predicted 100% of this group's combined criterion outcome. The implications of the existence of the subgroups socialized and undersocialized were discussed, as were the ramifications of being able to use the derived discriminant functions in tandem to predict the direction of change in a young adolescent's intentions vis-a-vis problem-behavior and substance abuse involvement, in the presence or absence of a counseling intervention. The potential utility of discriminant function in both research and clinical practice was also discussed and suggestions for future research were presented.