Women Offenders: Major Findings from the DFIA-R Research Studies

Research Highlights: The DFIA-R is a valid tool for women offenders and useful in pointing to areas of need associated with revocations.


No R-395_W

August 2017

Research at a glance - PDF

Women Offenders: Major Findings from the DFIA-R Research Studies

Why we did this study

The Dynamic Factors Identification and Analysis-Revised (DFIA-R) tool is a key component of the offender intake assessment used by the Correctional Service of Canada (CSC). This research brief summarizes key findings of two large-scale research projects validating the revised tool and exploring alternative calculation methods of scoring offender need as they pertain to women offenders.

What we did

Offenders with at least one DFIA-R assessment were obtained through the Offender Management System resulting in 24,798 men (24% Indigenous) and 1,368 women (37% Indigenous) with data available. Of this group, 16,743 men and 992 women were released and had follow-up data allowing the examination of the relationship of ratings with community outcomes. Analyses focused on the prevalence of the ratings and indicators, which domain ratings influenced the overall need rating, and which indicators and domains were related to community outcomes.

As well, calculated domain and overall need rating methods were explored to determine whether they improved the tool. All analyses were disaggregated by gender and Indigenous identity when possible.

What we found

Need level on nearly all domains was highest for women offenders, particularly Indigenous women. The two highest need domains for women offenders were the Personal/Emotional and Substance Abuse domains (> 75% rated moderate or high need, for each). Parole officer ratings on the Attitude, Substance Abuse, and Employment/Education domains had the greatest influence in the determination of high overall need for women Footnote 1.

DFIA-R ratings were more dynamic for women offenders than men, that is, more women were likely to see a reduction in need over time. For example, 26.6% of all women who had a second assessment were rated lower need on the Substance Abuse domain and 34% of Indigenous women had a reduction in need rating.

Higher needs on all domains were significantly related to community outcomes for women offenders.1 For example, women who were rated as high need in the Substance Abuse domain were nearly 5 times more likely to return to custody compared to those rated as low or no need (see Table 1 comparing the hazard of revocation for men and women with high ratings on each domain to those with asset, or no need). Results were similar though somewhat weaker for women with moderate need.

Table 1
Association between DFIA-R Domain Ratings of High Need and Any Return to Custody
Domain All Men
All Women
Employment/Education 3.77*** 4.66***
Marital/Family 1.36*** 1.42*
Associates 2.18*** 2.94***
Substance Abuse 3.53*** 4.73**
Community Functioning 2.75*** 1.86***
Personal/Emotional 2.04*** 1.94*
Attitudes 2.19*** 1.59**
Note HR = Hazard ratio; * <.05;** <.01; ***<.001

The majority of DFIA-R indicators were related to community outcomes for women offenders. 1 Those with the strongest association with returns to custody were related to substance misuse, antisocial associates, financial instability, education concerns, and problematic intimate relationships. Some particularly strong indicators included:

  • Associates with substance abusers (those with this indicator endorsed were about 4 times more likely to return to custody than those who did not have it endorsed
  • Alcohol or drug use has resulted in law violations (about 4 times more likely)
  • Alcohol and/or drug use is part of the offence cycle (> 3 times more likely)
  • Has gone on drug-taking bouts or binges (> 3 times more likely)
  • Job history has been unstable (> 3 times more likely)
  • Has used social assistance (2.5 times more likely)
  • Financial instability (> 2 times more likely)
  • Has less than high school diploma or equivalent (> 2 times more likely)
  • Intimate relationships have been problematic (> 2 times more likely)

Examining alternative methods of scoring the tool showed that calculated ratings based on the percentage of indicators endorsed could improve the predictive validity of the tool over that of the structured professional judgment of parole officers.1 However, as women offenders more frequently had more indicators endorsed, this calculated method of rating need produced a higher prevalence of high need ratings. We therefore altered the classification rules for women which resulted in fewer women in the high and more in the low and medium classifications. Applying this amended formula improved the prediction of the tool for women.

What it means

These results support continued use of the DFIA-R tool for women offenders. Results suggest that interventions could be effective for women if they focused on factors related to revocations including: reducing substance misuse, reducing contact with antisocial peers, increasing financial stability/employment, addressing education concerns, and supporting women in the development of healthy and prosocial intimate relationships.

While parole officer ratings of women’s need were found to be reliable and valid, calculated ratings improved the value of the tool in predicting outcomes for women offenders. Almost all indicators in the tool were related to women’s outcomes but further research would help determine whether unidentified indicators that could be specific to women could further improve the tool.

For more information

Stewart, L. A., Wardrop, K., Wilton, G., Thompson, J., Derkzen, D., & Motiuk, L. (2017). Reliability and validity of the Dynamic Factors Identification and Analysis – Revised (Research Report R-395). Ottawa, Ontario: Correctional Service of Canada.

Wilton, G., Stewart, L. A., & Motiuk, L. L. (R-400). Can the predictive validity of the Dynamic Factors Identification and Analysis – Revised be improved by calculated ratings? 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.

Date modified: