Adjustment of the security reclassification scale: Elimination of the administrative segregation item

Research Highlights: After removing the administrative segregation item, the SRS continues to be a valid and reliable assessment tool.

The Security Reclassification Scale (SRS) is a research-based tool designed by the Correctional Service of Canada (CSC) to help caseworkers determine the most appropriate level of security for an offender. More specifically, it includes factors related to institutional adjustment, escape risk, and public safety concerns. As administrative segregation policy has changed over the last few years, and with the elimination of administrative segregation with the impending passage of Bill C-83, this study examined the impact of dropping the segregation item on the validity of the SRS.

This study used the dataset from the recent revalidation of the SRS (Farrell MacDonald, Beauchamp, Conley, Cociu, & Scott, 2018). Overall, 6,281 SRS assessments were examined for the study period from April 1, 2014 to March 31, 2016. These assessments represented 5,433 federal men offenders.

For the purpose of this study, the dataset were randomly divided into construction (N = 3,141) and validation (N = 3,140) samples. Indigenous offenders accounted for 28% of each of the two samples. Revalidation analyses examined the modified scale's reliability, convergent validity, and the prediction of relevant outcomes (e.g., offender security level, disciplinary charges, and post-release outcomes).

To adjust the SRS score, the weighted value of segregation was removed from the assessed SRS score (0.5 for those without a segregation period and 3.0 for those that had one or more). Very few offenders assessed as minimum had segregation periods during the review period compared to those assessed as medium or maximum (2% versus 50% and 98%, respectively). The original SRS has a total range of score of 25 (10 to 35) based on scoring weights. Removal of the segregation period item reduced the overall scoring range to 22.5. Adjusted SRS score cut-offs were minimum (9.5 to 15.5), medium (16.0 to 24.0), and maximum (24.5 to 32).

The results demonstrated that the adjusted SRS is a reliable and valid tool for the security reclassification process, for both Indigenous and non-Indigenous offenders. The adjusted SRS scores and cut-offs ensured that a comparable proportion of men offenders were identified as minimum, medium, and maximum security compared to the original SRS. The overall concordance for Indigenous and non-Indigenous offenders between the original and adjusted SRS was 88.6% and 88.9%, respectively, indicating the revised scores and cut-offs are suitable for both groups.

Convergent validity analyses indicated that the adjusted SRS continues to have moderate associations with other offender risk and need measures, which were comparable for Indigenous and non-Indigenous offenders. In terms of predictive validity, the results demonstrated that the adjusted SRS assessed level was predictive of offender security level, disciplinary charges, the rates of discretionary release, and returns to custody, with and without an offence. Finally, all of the findings in this study were replicated with the validation sample, thereby reinforcing the consistency of the findings.

It is recommended that the adjusted SRS cut-offs identified in this study be integrated into the OMS algorithm for security reclassification. Additional factors such as Aboriginal Social History, institutional adjustment, escape risk, and risk to public safety will still be taken into consideration by CSC staff prior to the final security placement decision being made. Overall, on-going use of this tool is supported by the findings of this study.

Farrell MacDonald, S., Smeth, A., Sullivan, R., & Derkzen, D. Adjustment of the Security Reclassification Scale (SRS): Elimination of the administrative segregation item (Research Report R-432).Ottawa, Ontario: Correctional Service of Canada.

To obtain a PDF version of the full report, or for other inquiries, please e-mail the Research Branch or contact us by phone at (613) 995-3975.

You can also visit the Research Publications section for a full list of reports and one-page summaries.