Abstract
Background: This study explores the predictive capabilities of sociodemographic characteristics and risk and protective factors for youth recidivism while comparing two analytical methodologies; logistic prediction models and qualitative comparative analysis models. Methods: information from the judicial files of 389 young offenders (14-19.03 years) were gathered from the Juvenile Court and risk and protective factors were extracted from the administration of the Youth Level Service/Case Management Inventory. Recidivism data was also obtained for a follow-up period of two years for each young person. Results: the results showed two different profiles of reoffenders. Most were young boys with high risk scores and low protective factors, but a minority were young foreign girls with crimes against persons and low protective factors. Conclusions: being able to detect the different variables that contribute to recidivism can help implement prevention programs tailored to the criminogenic needs of each specific profile.