Performance of recidivism risk assessment instruments in U.S. correctional settings.

Abstract

With the population of adults under correctional supervision in the United States at an all-time high, psychologists and other professionals working in U.S. correctional agencies face mounting pressures to identify offenders at greater risk of recidivism and to guide treatment and supervision recommendations. Risk assessment instruments are increasingly being used to assist with these tasks; however, relatively little is known regarding the performance of these tools in U.S. correctional settings. In this review, we synthesize the findings of studies examining the predictive validity of assessments completed using instruments designed to predict general recidivism risk, including committing a new crime and violating conditions of probation or parole, among adult offenders in the United States. We searched for studies conducted in the United States and published between January 1970 and December 2012 in peer-reviewed journals, government reports, master's theses, and doctoral dissertations using PsycINFO, the U.S. National Criminal Justice Reference Service Abstracts, and Google. We identified 53 studies (72 samples) conducted in U.S. correctional settings examining the predictive validity of 19 risk assessment instruments. The instruments varied widely in the number, type, and content of their items. For most instruments, predictive validity had been examined in 1 or 2 studies conducted in the United States that were published during the reference period. Only 2 studies reported on interrater reliability. No instrument emerged as producing the "most" reliable and valid risk assessments. Findings suggest the need for continued evaluation of the performance of instruments used to predict recidivism risk in U.S. correctional agencies. (PsycINFO Database Record

Risk, Race, and Recidivism: Predictive Bias and Disparate Impact*

“…Although this issue has been studied with juvenile offenders (e.g., Olver, Stockdale, and Wormith, 2009), forensic instruments designed to predict violence (e.g., Singh and Fazel, 2010), and indigenous/nonindigenous groups in other countries (e.g., Wilson and Gutierrez, 2014), our focus is on comparing Black and White offenders in the United States on instruments designed to predict recidivism. In a recent meta-analysis, Desmarais, Johnson, and Singh (2016) identified 53 studies of 19 risk assessment instruments used in U.S. correctional settings. Only three studies permitted comparisons of predictive accuracy by offender race-and indicated that levels of predictive utility were identical (area under the ROC curve or AUCs = .69 on the "COMPAS"; Brennan, Dieterich, and Ehret, 2009) or highly similar (odds ratio or OR = 1.03 [Black] and 1.04 [White] on the Levels of Services Inventory-Revised or LSI-R; Kim, 2010;Lowenkamp and Bechtel, 2007) across groups.…”

Risk and punishment: The recent history and uncertain future of actuarial, algorithmic, and “evidence‐based” penal techniques

“…3 Currently, there are an array of risk instruments utilized within criminal justice and penal institutions in various locations, including the United States, Canada, Argentina, the United Kingdom, and Europe. In the United States alone, 19 risk assessment instruments have been validated in correctional settings (Desmarais, Johnson, & Singh, 2016;Desmaris & Singh, 2013). Perhaps the best-known risk instrument is the Level of Service Inventory (LSI)and its various iterations, especially the Level of Service Inventory-Revised (LSI-R)-which is utilized by more than 600 correctional agencies in North America (Lowenkamp, Lattesa, & Holsinger, 2004).…”

The Relationship Between Gambling Severity and Risk of Criminal Recidivism

“…Additionally, while the LSI-R: SV indicated a significant relationship between recidivism risk and gambling severity, analyses utilizing the SFS 81 did not reach significance. Likely differences in the performances of the two measures were due to the tendency of the LSI-R: SV to assess more recent or past-year risk factors, while the content of the SFS 81 captures more lifetime factors (49). Furthermore, the sample was recruited from sites of convenience and may not be fully representative of the postrelease population.…”