A Principled Approach to the Construction of Risk Assessment Categories: The Council of State Governments Justice Center Five-Level System

Abstract

<jats:p> Consistent risk category placement of criminal justice clients across instruments will improve the communication of risk. Efforts coordinated by the Council of State Governments (CSG) Justice Center led to the development of a principled (i.e., a system based on a given set of procedures) method of developing risk assessment levels. An established risk assessment instrument (Level of Service Inventory–Revised [LSI-R]) was used to assess the risk-level concordance of the CSG Justice Center Five-Level system. Specifically, concordance was assessed by matching the defining characteristics of the data set with its distribution qualities and by the level/category similarity between the observed reoffending base rate and the statistical probability of reoffending. Support for the CSG Justice Center Five-Level system was found through a probation data set ( N = 24,936) having a greater proportion of offenders in the lower risk levels than a parole/community data set ( N = 36,303). The statistical probabilities of reoffending in each CSG Justice Center system risk level had greater concordance to the observed Five-Level base rates than the base rates from the LSI-R original categories. The concordance evidence for the CSG Justice Center Five-Level system demonstrates the ability of this system to place clients in appropriate risk levels. </jats:p>

Combining Static and Dynamic Recidivism Risk Information Into the Five-Level Risk and Needs System: A New Zealand Example

“…Conceptually, the process of calculating the 5-Level System's levels is not complicated (see Kroner et al, 2020, for a demonstration). In practice, however, aligning tools to the framework is not straightforward; Hogan (2020) wrote extensively about the early practical challenges encountered by researchers.…”

“…In practice, however, aligning tools to the framework is not straightforward; Hogan (2020) wrote extensively about the early practical challenges encountered by researchers. Specifically, Hogan (2020) critiqued the existing research on the 5-Level System for its focus on specific subtypes of recidivism (violent and sexual recidivism) rather than general recidivism (e.g., Davies et al, 2020;Hanson, Babchishin, et al, 2017;Hogan & Sribney, 2019;Olver et al, 2018Olver et al, , 2021 change in prediction timeframe to 5-year outcomes (e.g., all except Kroner et al, 2020), and researchers' decision to split Level IV into sublevels (Level IVa and IVb) because expected recidivism rates did not reach the lowest bound of Level V (85%; even using general recidivism outcomes, for example, Kroner et al, 2020). Such adaptations to the guidelines show that to build agreement about standardization, the field will require an encompassing and collective program of research to avoid so many individualized deviations that forensic practitioners would potentially need to offer qualifications about each tool, jurisdiction, and assessment situation.…”

Improving our Risk Communication: Standardized Risk Levels for BARR-2002R

“…Nevertheless, Coulter et al (2019) did identify 4.1% of a New Zealand community sample (N = 440) as Level V for the Roc*RoI (Bakker et al, 1999) based on expected recidivism rates of over 85% within two years. In contrast, Kroner et al (2020) did not identify any individuals as Level V in two large, US community samples (N = 24,936 and N = 36,303). The lack of a Level V may also be attributed to the current sample being exclusively individuals with a sexual offending history, who tend to have lower general (overall) recidivism rates than individuals convicted of other crimes (Stewart et al, 2019).…”

“…The New Zealand community corrections study found that most individuals fell into Level III (21%) and IV (54.5%) when using their tool, the Roc*RoI (Coulter et al, 2019). The two US community samples also found most individuals fell into Level III (40% in Sample 1, 19% in Sample 2) and Level IV (32% in Sample 1, 71% in Sample 2; Kroner et al, 2020). This is not surprising, as men with sexual offences typically score lower on general criminality than men with nonsexual offences (e.g., Craig et al, 2006).…”

“…The Five-Level system was designed for describing the risk and needs of individuals at risk for general recidivism and has been applied to instruments designed for that purpose (e.g., LSI-R; Andrews & Bonta, 2001; see Kroner et al, 2020). For example, Kroner et al (2020) applied the Five-levels to the LSI-R in a large sample of 24,936 individuals on community release. Given that this was a lower risk sample, Level V (with expected general recidivism rates over 85%) was not populated; instead, Level IV was divided into Level IVa and Level IVb (consistent with Hanson, Babchishin, et al, 2017).…”