Multivariate approaches to the identification of delinquency proneness in adolescent males.

A sample of 337 adolescent male students were surveyed for demographic, individual, school, and familial functioning and delinquency status to investigate two questions relevant to prediction of adolescent delinquency proneness. First, three methods of scoring a delinquency self-report measure (frequency, variety, and seriousness) were compared to assess their differential relevance to the prediction of delinquency proneness. Second, a multivariate model was examined to assess its explanatory ability for identification of delinquency proneness. Findings, replicated through a series of regression analyses, demonstrate that age of onset is the best predictor. Other than family functioning, psychosocial indicators add little to the predictive model. Third, a specific factor model was preferable to a "risk count" method. Finally, the advantage of self-reports of delinquent behavior over official records is discussed as is the comparability of self-report scoring procedures. How self-reported delinquency is scored is not as critical as previously thought.