Predicting Recidivism in Juvenile Offenders: Comparison of Risk Factors for Adolescent Male versus Female Offenders

The purpose of the proposed study was to examine if differences existed in the riskfactors predictive of recidivism in adolescent male offenders versus adolescent femaleoffenders. Specific independent variables examined included special educationdiagnosis, diagnosis of an emotional disability, diagnosis of a learning disability, age attime of the most serious offense, socioeconomic status, ethnicity, grade point average,high school credits, grade level, standardized achievement scores in reading, writing, and math, adjudication status, offense history, and offense severity. It was hypothesized that these variables would not predict sex membership or recidivism in female delinquents,but would significantly predict recidivism in male juvenile delinquents. It was also hypothesized that no significant differences would exist between factors predictive of recidivism in male versus those for female delinquents.In regard to factors predicting group membership, analysis indicated that offenseseverity could significantly differentiate between male and female delinquents. Forrecidivism in female delinquents, analyses found that 7 of 21 factors were significantlypredictive of recidivism in females, including: total number of parole violations,diagnosis of an emotional disability, total number of status offenses, adjudication status,the total number of personal and property misdemeanor offenses, and grade pointaverage. For males, 9 of 21 variables significantly predicted recidivism, including: totalnumber of parole violations, diagnosis of an emotional disability, standardized readingperformance scores, total number of status offenses, total number of drug offenses,adjudication status, total personal and property misdemeanor offenses, and grade level.Thus, six predictor variables for recidivism were shared by both female and maledelinquents, namely, total parole violations, diagnosis of an emotional disability, totalstatus offenses, adjudication status, total misdemeanor personal offenses, and totalmisdemeanor property offenses. A hierarchical regression was also computed with sex as an independent variable, resulting in 12 variables being predictive of recidivism. Theresults revealed that differences existed between males and females in terms ofrecidivism, and highlight that risk assessment instruments should be individualized based on sex of the juvenile. The implications of these findings, areas for future research, and limitations to the study are discussed.