An Examination of a Reweighted Custody Rating Scale for Women

To obtain a PDF version of the full report, contact the following address:

Key Words

custody rating scale, validation, reweighting, women offenders, security level, custody classification

Why we did this study

The placement of offenders at the correct security levels is a critical incarceration decision that influences inmates’ living conditions, access to the community, programming options as well as public safety. To date, Correctional Service of Canada (CSC) has used the Custody Rating Scale (CRS), an actuarial tool, along with additional information to classify and place women into a security level at admission. However, this scale was originally developed with a sample of men and, because of this, critics argue against its use with women. Moreover, previous research demonstrates that a high number of Aboriginal women are recommended for placement in medium and maximum security levels. The goal of the current research is to examine the possibility of reweighting the CRS’s items to improve its validity and relative efficacy for Aboriginal and non-Aboriginal women.

What we did

Women who had received an initial custody classification between June 2007 and April 2011 were randomly assigned to one of two groups, a construction sample and a validation sample. In total, 541 women (143 Aboriginal women and 398 non-Aboriginal women) were included in the construction sample and 542 women (143 Aboriginal women and 399 non-Aboriginal women) were included in the validation sample.

Reweighting of the CRS, using the Burgess method, was undertaken with the construction sample in order to give greater weight to the scale items that were more predictive of offenders’ subsequent institutional adjustment. Following the reweighting, the validity of the scale was analyzed for both the samples.

What we found

The results demonstrated that reweighting increased the extent to which the CRS predicted institutional incidents and charges for the full sample of women and for the non-Aboriginal sub-sample. These improvements were not consistent for the Aboriginal sample. For example, reweighting slightly reduced the number of Aboriginal women at higher security levels. But, over-representation of Aboriginal women at medium and high levels of security persisted.

Overall, reweighting slightly improved the scale for non-Aboriginal women, but had less of an effect for Aboriginal women. Given these differing patterns and the limited magnitude of improvements, implementation of the reweighted scale is not recommended.

What it means

Ultimately, it was determined that although reweighting improves the scale for non-Aboriginal women, the relatively small sample size of Aboriginal women limited the extent to which their data influenced the reweighting and, therefore, the validity of the reweighted scale for Aboriginal women. Other approaches could improve initial custody classification for women such as separately reweighting the scale for Aboriginal and non-Aboriginal samples.

For more information

Rubenfeld S. (2014). An Examination of a Reweighted Custody Rating Scale for Women. Research Report R289. 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 website for a full list of research publications.

Prepared by: Sara Rubenfeld