Predictive Performance of a Self-Report Adaption of a Criminogenic Risk Assessment Too
Presentation Type
Poster
Faculty Advisor
Chrisopher King
Access Type
Event
Start Date
26-4-2024 2:15 PM
End Date
26-4-2024 3:15 PM
Description
Using structured tools to assess an individual’s risk for repeated crime has become standard practice throughout the justice system to assist with making informed decisions about rehabilitation and management. Popular validated risk assessment tools, such as the Level of Service/Case Management Inventory (LS/CMI), are designed to be utilized by trained evaluators, and thus entail appreciable, practical demands. The current study examined the convergent and predictive validity of a novel self-report adaptation of the LS/CMI, as a potential option when the administration of the evaluator-rated tool is not feasible. The sample consisted of 338 individuals from a correctional facility who were administered both versions of the LS/CMI and were then tracked for rearrest during a follow-up period of 3.5 to 5 years (21% of the sample was subsequently rearrested). Convergent validity between evaluator-rated and self-report risk scores were examined using correlations, and the predictive utility of both scores using correlations and stepwise logistic regression. The two risk scores correlated significantly, positively, and modestly, and both scores also correlated significantly, positively, and modestly with rearrest. However, in a multivariable model, only the evaluator-rated risk score significantly incrementally predicted rearrest. Despite finding modest evidence for convergent validity between the evaluator-rated and self-report risk scores, the self-report version of the LS/CMI did not demonstrate incremental predictive utility. Limitations of the study call for replication of the latter null result. Additionally, both versions of the LS/CMI measure rehabilitation-relevant factors beyond those that relate to future crime.
Predictive Performance of a Self-Report Adaption of a Criminogenic Risk Assessment Too
Using structured tools to assess an individual’s risk for repeated crime has become standard practice throughout the justice system to assist with making informed decisions about rehabilitation and management. Popular validated risk assessment tools, such as the Level of Service/Case Management Inventory (LS/CMI), are designed to be utilized by trained evaluators, and thus entail appreciable, practical demands. The current study examined the convergent and predictive validity of a novel self-report adaptation of the LS/CMI, as a potential option when the administration of the evaluator-rated tool is not feasible. The sample consisted of 338 individuals from a correctional facility who were administered both versions of the LS/CMI and were then tracked for rearrest during a follow-up period of 3.5 to 5 years (21% of the sample was subsequently rearrested). Convergent validity between evaluator-rated and self-report risk scores were examined using correlations, and the predictive utility of both scores using correlations and stepwise logistic regression. The two risk scores correlated significantly, positively, and modestly, and both scores also correlated significantly, positively, and modestly with rearrest. However, in a multivariable model, only the evaluator-rated risk score significantly incrementally predicted rearrest. Despite finding modest evidence for convergent validity between the evaluator-rated and self-report risk scores, the self-report version of the LS/CMI did not demonstrate incremental predictive utility. Limitations of the study call for replication of the latter null result. Additionally, both versions of the LS/CMI measure rehabilitation-relevant factors beyond those that relate to future crime.