Drawing on interviews with 24 correctional practitioners who use risk assessment instruments daily, the authors examine barriers to the use and implementation of these instruments.
Findings reveal that practitioners have confidence in the state of risk assessment generally, but are skeptical about risk assessment on the job due to limited resources. They also point to barriers that inhibit their effectiveness including recognition of population heterogeneity, predictive misspecification due to data lags and overemphasis on stable predictors, and lacking guidance on appropriate use of vast available data. Instruments for measuring risk serve purposes beyond those intended by the social scientists who developed measures. The authors conclude with lessons for increasing the utility and legitimacy of risk assessments and with a call for incorporating latent uses of assessments into design. (Published abstract provided)