Validating the Predictive Accuracy of the Static Factors Assessment (SFA) Risk Scale for Federally Sentenced Offenders in Canada
The Static Factors Assessment (SFA) is used by the Correctional Service of Canada to assess criminal risk. It includes 137 items in three sub-components: the Criminal History Record (CHR), Offence Severity Record (OSR), and Sex Offence History Checklist; the first two sub-components are examined in...
1. VerfasserIn: | |
---|---|
Beteiligte: | |
Medienart: | Elektronisch Aufsatz |
Sprache: | Englisch |
Veröffentlicht: |
2017
|
In: |
Canadian journal of criminology and criminal justice
Jahr: 2017, Band: 59, Heft: 1, Seiten: 1-25 |
Online-Zugang: |
Volltext (Resolving-System) |
Journals Online & Print: | |
Verfügbarkeit prüfen: | HBZ Gateway |
Schlagwörter: |
Zusammenfassung: | The Static Factors Assessment (SFA) is used by the Correctional Service of Canada to assess criminal risk. It includes 137 items in three sub-components: the Criminal History Record (CHR), Offence Severity Record (OSR), and Sex Offence History Checklist; the first two sub-components are examined in this study (109 items). Although the SFA has been used for all federal offenders for nearly 20 years, there are no studies examining its ability to predict community outcomes. This study included 8,767 federal offenders within a five-year follow-up period, and it examined revocations without an offence, readmissions for any offence, and readmissions for a violent offence. The overall SFA, CHR, and OSR were related to recidivism outcomes, although the sum of the items in the CHR significantly out-predicted the overall SFA rating. Most items in the CHR had significant predictive accuracy, whereas roughly half the OSR items were predictive; nonetheless, the OSR added positive incremental validity to the CHR. The SFA overall rating and the CHR and OSR sub-components are valid for offender risk assessment with Canadian federal offenders, although the current results suggest that improvements to the SFA should be undertaken. |
---|---|
ISSN: | 1911-0219 |
DOI: | 10.3138/cjccj.2015.E04 |