U.S. flag

An official website of the United States government, Department of Justice.

Prediction is Local: The Benefits of Risk Assessment Optimization

NCJ Number
300947
Journal
Justice Quarterly Dated: 2021
Date Published
2021
Length
23 pages
Author(s)
Zachary Hamilton; Alex Kigerl ; Melissa Kowalski
Agencies
NIJ-Sponsored
Annotation

This study used a large, 10-state sample of assessed youth (N=494,050) in isolating, testing, and evaluating the relative impact of notable risk assessment variations, including item selection, response weighting, outcome definition/duration, and jurisdiction, and it further combined approaches to evaluate an ‘optimized’ development approach.

Abstract

In most states and jurisdictions, risk assessments are incorporated into justice system practice. Despite decades of use, the methods of tool development are rarely translated to the field. Many agencies implement ‘off-the-shelf’ versions, where a tool developed with a unique set of methods and subjects demonstrates prediction shrinkage when applied to a new jurisdiction. The current study found substantial gains with each variation tested, where optimized models provided a full effect size predictive improvement. Based on these findings, this article discusses best practices for the future of risk assessment, noting the predictive accuracy lost when implementing tools off-the-shelf; and it notes how optimization techniques substantially improve risk prediction, specifying a given tool to an agency’s needs. (publisher abstract modified)

Date Created: May 21, 2021