Validity of Pretrial Risk Assessment Instruments in Predicting Violence: A Systematic Review of the Literature

• In Event: Poster Session 2

Abstract

Despite their increased use, pretrial risk assessment instruments remain the topic of discussion and debate. Under particular scrutiny is their validity in predicting violent crime in the pretrial period. Findings of a prior meta-analysis (Desmarais et al., 2021) showed good to excellent validity of most pretrial risk assessment instruments in predicting failure to appear in court and new criminal arrests during the pretrial period. However, only two of the 11 studies reviewed included new arrests for violent crimes as an outcome. Since then, several studies have been published that report on predictive validity in relation to violence during the pretrial period. This presentation will provide an overview of the state of science and practice as it relates to assessing risk for new violent criminal arrests in the pretrial context. Specifically, we will systematically review the literature published since December 31, 2018, adding to the data included in the Desmarais et al. (2021) meta-analysis. We will follow the PRISMA (2020) reporting guidelines to support transparency in our methodology and results. We will discuss results overall, by instrument, and when possible, across groups defined by race/ethnicity and gender. Findings will be considered in relation to recommendations for research, policy, and practice.

Authors

• Samantha Adams, Policy Research Associates

• Sarah L. Desmarais, Policy Research Associates

• Samantha Zottola, Policy Research Associates