Dynamic risk factors reassessed regularly after release from incarceration predict imminent violent recidivism

Area of Science:

• Criminology
• Forensic Psychology
• Corrections

Background:

• Reassessing dynamic risk factors enhances recidivism prediction in community corrections.
• Evidence for predicting violent recidivism specifically through reassessments is less established.
• This study investigates the impact of reassessment proximity and aggregation on predicting imminent violence.

Purpose of the Study:

• To determine if reassessment of dynamic risk factors improves the prediction of violent recidivism compared to initial assessments.
• To explore whether the aggregation or proximity of reassessments enhances the prediction of imminent violence.
• To identify specific risk factor domains that best predict violent recidivism.

Main Methods:

• A prospective study was conducted on individuals under community supervision in New Zealand.
• Data included 75,917 assessments from 3,421 participants, primarily men.
• Dynamic Risk Assessment for Offender Re-entry (DRAOR) ratings and static risk scores were used to predict violence within two weeks.

Main Results:

• Violent recidivism was associated with higher initial risk ratings and greater week-to-week fluctuations in DRAOR scores.
• Aggregated scores of acute, faster-changing risk factors demonstrated the best predictive accuracy (c-index = 0.68).
• Changes in these acute risk factors incrementally improved predictions beyond criminal history and initial scores.

Conclusions:

• Consistent patterns of interpersonal hostility and reactivity, as aggregated over time, are more predictive of imminent violence than current presentations.
• Supervision officers should focus on overall risk factor trends rather than assuming isolated incidents signal immediate danger.
• Findings align with social cognition theories explaining violent behavior.