Research
Prison data analysis finds choosing lowest reoffending risk scores preserves overall accuracy
Phys.org ·
The Brief
- An analysis of prison data examined how to handle equally likely reoffending risk scores when predicting recidivism.
- The study reports that choosing the lowest of several equally likely risk scores preserves overall accuracy.
- The source is a short excerpt, so further methodological detail is not available from the provided text.
Why it matters
Risk assessment tools used in criminal justice may have tie-breaking choices that affect individual outcomes, so the way scores are selected matters for science and policy readers.
Who’s affected
- researchers in criminology and machine learning
- policymakers and practitioners who use recidivism risk tools
- #recidivism
- #risk assessment
- #machine learning
- #criminal justice
- #prison data
Whether an individual reoffends after their release from prison or is successfully rehabilitated depends on several factors. When assessing the risk of recidivism, all of these are taken into account—both by human experts and by machine learning models. If several equally…
Timeline
Science News - Mathematics, Economics, Archaeology, Fossils
Prison data analysis finds choosing lowest reoffending risk scores preserves overall accuracy
Science News reported the prison data analysis on lowest reoffending risk scores