Understanding Risk and Protective Factors to Expand What We Know about Recidivism of Juvenile Offenders

Background/Purpose: Substantial research has been conducted on detained and incarcerated youth and the risk and protective factors associated with their delinquency. Less is known about risk and protective factors associated with youth on probation. This study examines gender differences in youth responses on the Youth Assessment Screening Instrument (YASI) and the association with their recidivism. The YASI assesses a youth’s risk and protection to determine appropriate case planning for them.

Methods: YASI and recidivism administrative data was collected from youth and their parents/caregivers about youth on probation in a Midwestern county between January 2009 and August 2013. These records were merged creating a unique dataset. The sample (N=5,831) was 89% male and 11% female; 6% White, 76% African American (AA), 13% Hispanic,  4% Mixed & 2% Other; and included ages 12-17. Seven percent of the sample recidivated. Chi-square, ANOVA and logistic regression analyses were conducted using SPSS 21. Logistic regression was conducted to test the effect of the predictors on youth recidivism.

Results: Gender differences were noted as girls reported greater risk for physical/sexual abuse (19%), affective D/O (16%) and suicidality (15%) than boys.  Boys reported greater protection for appropriate parental discipline (37%), closeness with teachers (28%), and extracurricular activities (26%) than girls. In this high-risk sample of girls, the results between risk and protective factors were not significant with race or recidivism. By race, White boys reported the most significant risk factors in affective D/O (22%), parent's w/mental health (10%), and no leisure activities (72%). AA boys reported the most significant protective factors in educational beliefs (80%), closeness to parents (92%) and teachers (32%). Risk factors associated with boys’ race were locked/kicked out and conduct disorder symptoms (Mixed) and problematic substance abuse (White). Risk factors predicting recidivism were not significant for boys though protective factors were significant. Logistic regression predicted the probability that boys would recidivate, N=5178. The Hosmer-Lemeshow test was not significant, indicative of good fit.  Predictor variables included age, race, locked/kicked out, substance abuse, conduct D/O symptoms, appropriate parental discipline, and closeness to prosocial peers. The full model was significant, x²(7, N=5178) = 28.67, p<.01. Boys’ reporting that their parents used appropriate discipline (OR=-.463) and were close to prosocial peers (-.273) were less likely to recidivate. The OR for race indicates that White males were -.169 less likely to recidivate than members of minority groups.

Conclusions/Implications: Study findings indicate that probation-involved youth experience risk and protective factors though only protective factors predicted recidivism in this sample. Therefore, community services that enhance parenting skills and relationships with prosocial peers are needed. The lack of significant findings for girls notes further exploration of risk, protection and recidivism in a national sample. The high percentage of females that indicated high levels of risk factors and males that indicated protective factors is rarely noted in the literature, but it is important and should be considered in the determination of service provision in the community.