The factor structure of static actuarial items: its relation to prediction

Area of Science:

• Forensic Psychology
• Criminology
• Risk Assessment

Background:

• Actuarial instruments are crucial for assessing sex offender recidivism risk.
• Existing tools like RASORR, Static-99, VRAG, SORAG, and MnSOST-R are widely used.
• Understanding the underlying components of these tools can refine risk prediction.

Purpose of the Study:

• To identify latent components within common actuarial risk assessment instruments for adult sex offenders.
• To examine the relationships between these components and specific recidivism types (violent vs. sexual).

Main Methods:

• Principal components analysis was applied to data from four actuarial instruments.
• The analysis utilized a dataset of 311 adult sex offenders, including child molesters and rapists.
• Statistical correlations were calculated between identified components and specific risk assessment tools.

Main Results:

• Six interpretable components emerged: Antisocial Behavior, Child Sexual Abuse, Persistence, Detached Predatory Behavior, Young and Single, and Male Victim(s).
• The Rapid Assessment of Sex Offender Risk for Recidivism (RASORR) correlated with Persistence.
• The Violence Risk Appraisal Guide (VRAG) and Sex Offender Risk Appraisal Guide (SORAG) correlated with Antisocial Behavior.

Conclusions:

• Antisocial Behavior emerged as a significant predictor of violent recidivism.
• Persistence and Child Sexual Abuse were significant predictors of sexual recidivism.
• These findings can inform more nuanced sex offender risk assessments and interventions.