This report was submitted by Team Aurors as a participant in the U.S. Justice Department’s National Institute of Justice’s (NIJ’s) Recidivism Forecasting Challenge, which aimed to improve the ability to forecast recidivism.
The work reported by the Aurors Team addresses forecasting models that include regression analysis methods (binary logit and LASSO regressions) and machine learning techniques (random forest), combined through a model averaging procedure. The output consists of the percent likelihood of individuals recidivating within one, two, or three years from release. Although this report explains the database construction process and the modeling approach, the results focus on the section of female parolees recidivating within 3 years, since that is the category for which the Aurors team came in second place.
Downloads
Similar Publications
- Longitudinal Cohort Study: Predictive Validity of the Structured Assessment of Violence Risk in Youth Individual/Clinical Risk Factor on Recidivism Among Mississippi Justice-Involved Youth
- Just Science Podcast: Just Resolving a Capital Murder Case in Denton, Texas
- Neighborhood Disadvantage, Social Groups, and Adolescent Violence: Assessing Mechanisms in Structural-Cultural Theories