Summary

This study examined the predictive validity of the Positive Achievement Change Tool-Prescreen (PACT-PS) risk assessment for predicting staff assault among 787 male juvenile offenders in correctional facilities. It found the tool had poor predictive validity overall and for minority youth.

Key Takeaways

  • The PACT-PS had poor predictive validity for staff assault, with AUC values of 0.59 overall.
  • The PACT-PS performed better for white youth (AUC 0.68) than Hispanic (0.57) and black (0.56) youth. 
  • Criminal history factors like prior felony referrals were most predictive of staff assault.
  • Social history factors did not improve predictions for minority youth. 
  • The tool needs modification with factors tailored to institutional misconduct to improve predictive validity.