ID: ACJ10-A006 01 Sep 2020
by Laurence (Larry) Motiuk, Leslie Anne Keown

Advancing Corrections Journal - Edition #10 - Understanding, Assessing, Managing and Reducing RISK (ACJ10-A006)

Article 6: Constructing Risk Assessment Algorithms For Violent, Drug And Property-Related Offenders-An Empirical Approach (ACJ10-A006)

Abstract
In the Canadian federal correctional system, the Offender Intake Assessment (OIA) and correctional planning process are primarily focused on addressing static and dynamic risk factors. These core components of OIA were examined to determine whether algorithmic equations tailored across major offence types could potentially be used for administration by means of a hand-held mobile application. In accordance with Schedules in the Corrections and Conditional Release Act, three major offence types, namely Schedule I (violent, excluding murder), Schedule II (drug), and Non-violent (property) were constructed for 6,525 male first releases over a two year period (2016-17 and 2017-18). An Offender Management System database was used to extract a set of 11 static risk indicators and 7 dynamic domain ratings for each case. Also gathered from OMS was whether or not there were any returns to federal custody. A combined static and dynamic risk index yielded impressive predictions of custodial return for violent, drug and property-related offenders with significant AUCs of .76, .70 and .71, respectively. These results suggest that combining static and dynamic factors into scoring algorithms for major offence types can be useful for moving offender risk assessment further towards streamlined applications technology.
 
Please log in to download the full article. Access is available exclusively to ICPA Full, Staff, and Professional Members.