Factors contributing to juvenile recidivism in a predominately hispanic population
MetadataShow full item record
Identifying adolescents at risk for re-offense, specifically among a predominately Hispanic population, is rarely empirically examined (Cintrón, 2006). The Hispanic population comprises the largest minority population in the United States, and accounts for over 60% of the population in Nueces County (US Census Bureau, 2011). The purpose of this study was to explore the boundaries and utility of the Massachusetts Youth Screening Instrument Version 2 (MAYSI-2) through its subscales, specifically in a predominately Hispanic population. A binary logistic regression was conducted to explore the extent to which MAYSI-2 subscale scores, demographic variables, and offense information predict re-offense. A sample of 884 adolescent offenders from Nueces County was utilized to examine factors that are predictive of recidivism. The regression analysis included all offenders from the year 2010. A listwise deletion was conducted to remove offenders with multiple offenses, and missing MAYSI-2 scores, or missing data. Data were collected ex-post facto and analyzed using a hierarchical logistic regression. The results of the logistic regression indicated all three domains (MAYSI-2 subscale scores, demographic variables, and offense information) were statistically significant predictors of juvenile re-offense. Three of the MAYSI-2 subscales, Alcohol/Drug Use, Angry-Irritable, and Somatic Complaints, were significant predictors of re-offense. Additionally, six of the demographic variables were found to be predictive of recidivism: juvenile age, number of siblings, Hispanic adolescents, gang affiliation, type of school, and legal guardian. Significant variables identified as predictive of re-offense from the offense information included misdemeanors and prior violent offenses. Overall, each of the three domains provides statistically significant contributions to the prediction of the dichotomous dependent variable, re-offense. The most powerful contribution for predicting juvenile recidivism is from the demographic variables, age and legal guardian (i.e., coming from a home with an absent biological parent). The results of this study imply that personal interactions with offender may be more helpful than psychometric measures at identifying adolescents at risk to reoffend. Although attempting to categorize adolescents merely based on demographic information can result in biases, and stereotyping, the information can be used to identify risk factors that may impede an adolescent's success. Identifying adolescents that present with characteristics indicating higher risk for re-offense, can assist clinicians in developing treatments.
Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy.