비행 예방 프로그램의 개발을 위한 청소년비행 예측 요인의 발굴

Identifying the predictor variables of juvenile delinquency for the development of prevention programs

This study conducted a step-wise discriminant analysis to determine which variables play key roles in distinguishing among 3 mutually exclusive delinquent groups - disclosed delinquency group, closed delinquency group, and no delinquency group - and to 'develop a procedure for predicting group membership for new cases whose group membership is undetermined. Variables included in this study were collected by the review of previous empirical studies, which were identified as related to juvenile delinquency. The variables can be classified as character-related variables, family-related variables, school-related variables, peer-related variables, and socioeconomic variables. A sample of 693 youths were employed to conduct the statistical analysis. Out of 35 possible predictors of juvenile delinquency 14 variables were included in the pool of predicting variables. This study used a hold-out sample (n=300) to test if the linear discriminant function classify cases correctly into one of 3 groups. The percentage of cases classified correctly by the discriminant function was turned out to be acceptable in the area of social sciences. Discussions and suggestions were provided in terms of prevention and intervention for juvenile delinquency with focus on the 14 predictor variables.