Correctional Service Canada
Symbol of the Government of Canada

Common menu bar links

Compendium 2000 on Effective Correctional Programming

Warning This Web page has been archived on the Web.


Effective Correctional Practice with Women Offenders


Until recently, there was a dearth of knowledge regarding what constitutes effective corrections for women. Indeed, women offenders were commonly considered “correctional afterthoughts” due to the lack of research to guide strategies for intervention with this group. As a result, correctional services for women were markedly poorer in quality, variety, and availability than those for their male counterparts (Ross & Fabiano, 1985). Currently, some authors continue to note that facilities and services offered to women inmates are derived from men models of corrections (Chesney-Lind, 1997; Garcia-Coll, Miller, Fields, & Mathews, 1997).

The Canadian federal correctional system has begun to address this problem; the past decade has ushered in a new philosophy of women's corrections in Canada. In brief, the new philosophy sets out standards of practice that are based on research that is sensitive to the unique situation of women offenders. Moreover, virtually every aspect of the post-sentence correctional process has been amended to reflect the distinctive needs and abilities of women. Particular modifications range from a physical decentralization and restructuring of the correctional environment (Construction Policy and Services, 1992) to creating a separate program strategy for women (Federally Sentenced Women Program, 1994; Office of the Deputy Commissioner for Women, 2000).

The particular results of the “new philosophy” are discussed in more detail later in this chapter. First, a brief history of women's corrections in the Canadian federal system will be presented. This will be followed by a discussion of issues germane to women offender assessment as well as qualities of effective programs for women, focusing explicitly on the principles of risk, need, and responsivity. Finally, the chapter will include a very brief discussion of factors that are particularly salient to the evaluation of correctional programs for women.

While the knowledge base regarding effective corrections for women offenders still lags behind that for their male counter-parts, researchers and practitioners have been working diligently to fill the information gap. In short, the icon of the woman offender as “correctional afterthought” is slowly dissipating. Research evidence regarding what constitutes effective correctional practice with women offenders is beginning to accumulate; and forms the basis for this chapter.


The first Canadian federal correctional facility for women, the Prison for Women, opened in Kingston, Ontario in 1934. Within four years of its opening, the Archambault Commission became the first of many commissions to recommend its closure (Arbour, 1996; Vachon, 1994). The institution was repetitively criticized on numerous grounds, including: overly austere security measures, poor programming, and inability to adequately address the needs of Aboriginal and Francophone women. Indeed, between 1938 and 1990, at least fifteen government reports had identified serious deficiencies in the services provided to women inmates (Arbour, 1996) The Prison for Women was the only federal prison for women offenders.2 This was the subject of fundamental and widespread concern; many federally sentenced women were isolated from their families and social support networks and had greater difficulty preparing for release and reintegration into the community. Despite these concerns, the Prison for Women remained the only Canadian women's federal correctional facility for well over half a century. The last inmate was transferred out of the Prison for Women in May; it was officially closed on July 6, 2000.

Fortunately, correctional practice with women offenders has changed dramatically over the past decade. Many of the progressive developments can be attributed to recommendations put forth by the Task Force on Federally Sentenced Women (1990). In brief, the Task Force was established in the late 1 980s to address long-standing concerns with the inequitable treatment of women offenders. Its principle mandate was to develop a comprehensive strategy for the management of federally sentenced women.

The research and consultation conducted by the Task Force was largely qualitative and included surveys of both staff and women offenders, as well as comprehensive literature reviews. It was the first time in the Correctional Service of Canada's (CSC) history that the voices of women offenders were afforded such serious consideration in the development of strategic policy direction (Stableforth, 1999). In 1990, the Task Force published its report: Creating Choices. It represented a new definition of effective corrections for women offenders, reached through consensus by a broad range of correctional practitioners and government/ non-government agencies. Creating Choices was, and continues to be, considered exceptional in its advent of a woman-centered approach to corrections.

The Task Force developed a holistic approach to corrections for women using five guiding principles:

  • Empowerment,
  • meaningful and responsible choices,
  • respect and dignity,
  • supportive environment, and
  • shared responsibility.

These principles drove specific recommendations to replace the Prison for Women with four regional facilities and an Aboriginal healing lodge. Importantly, it was recommended that these facilities be constructed and operated using a “community-living” model, where the women offenders would reside in houses and be responsible for their daily meals, laundry, cleaning, and leisure time. The Task Force further called for the development of women-centered interventions, including survivors of abuse therapy and mother-child programming. Finally, it was strongly suggested that an effective community strategy, enhancing resources and sup-port networks, be established for women offenders.

While some dissidents maintain that CSC has improperly interpreted the recommendations of the Task Force (see, for example, Hannah-Moffat, 1995), others argue that the Service has now operationalized a fundamentally distinct concept of effective corrections for women offenders (Stableforth, 1999). In accordance with the Task Force proposals, CSC opened five new facilities for federally sentenced women, as previously described.

Pursuant to Task Force recommendations, operations and programming both within the institution and post-release have been amended. In particular, the implementation of a Women Offender Program Strategy (Federally Sentenced Women Program, 1994; Office of the Deputy Commissioner for Women, 2000) has provided an opportunity for participants to benefit from programs that were developed specifically to meet women's needs and styles of learning. For instance, a mother-child program is operational at all regional facilities, allowing young children to reside with their mothers on a full-time basis, while older children are permitted part-time residency. In recognition of the pervasive and disparate mental health needs of women offenders, a separate, gender-specific Mental Health Strategy has also been developed and implemented (Laishes, 1997). Additionally, in late 1999, an Intensive Intervention Strategy (IIS) was introduced for women offenders. The strategy provides a protocol for safe and secure accommodation for maximum-security women and women with special needs, while emphasizing intensive staff intervention, programming, and treatment.

A comprehensive Community Strategy for federally sentenced women has been established (Women Offender Sector, 1998). It includes a variety of residential alternatives for post-incarceration community living, such as: community correctional centers and community residential centers, satellite apartments, private home placements, and day reporting centers. To date, the Strategy has been implemented successfully at the regional level.

Research is the cornerstone for the innovations that we have witnessed in the Canadian women's federal correctional system. Fortunately, the importance of a gender-sensitive research paradigm is being increasingly recognized. Investigators have reached consensus on the importance of using multi-method approaches in data collection and analysis. While the value of empirical research will always be recognized, it is equally necessary within the context of women's corrections, to conduct qualitative studies. This ensures that the data include the views of individual women and staff. In addition, the Service regularly consults with community partners, advocacy groups, and external experts in the “Women Offender” domain. Consequently, studies with women offenders are becoming both more abundant and more feminist-oriented.

Correctional program evaluation research is increasingly responsive to methodological limitations common to most studies on women offenders. Problems include the small and dispersed population of federally sentenced women and lack of a sound, gender-specific, program theory. Proper evaluation research will mitigate these effects through consideration of the contextual framework and views of the participants, usage of a multi-method approach, and attending to structural or environmental issues such as management support for the program.

Research in the area of actuarial tool development is becoming more gender-specific. In particular, the importance of recruiting samples comprised solely of women offenders to develop separate and unique classification instruments for women is now recognized. Although these efforts require considerable resources, they are crucial to the attainment of equity for women in conflict with the law.

Over the past decade, the Service has evolved considerably in its treatment of federally sentenced women. The correctional environment has changed in terms of both structure and philosophy. Progressive, woman-centered programs have been implemented and innovative research methodologies are being used to evaluate those programs. Similarly, new actuarial tools are being designed to reflect the unique characteristics of the women offender population.

It is acknowledged, however, that the parameters of effective correctional practice for women offenders are still not extremely well defined. Sustained research efforts will continue to address this issue. Prospective investigations will more solidly demonstrate what is “effective correctional practice” for women offenders, while recognizing the heterogeneity of this group.


Although there has recently been increased attention to women offender issues, some argue that still “little is known about the program elements [for women] that promote successful outcomes such as economic and social independence, family reunification, and reduced involvement in the criminal justice system. ” (Koons, Burrow, Morash, & Bynum, 1997, p. 513). Perhaps because the overwhelming majority of offenders are men, the services offered to women inmates have traditionally been based on models derived from their male counterparts. Accordingly, past research which examined the ability of programming to meet the needs of women offenders suggested that treatment for women was both inappropriate and unavailable (Dauvergne-Latimer, 1995; Gray, Mays, & Stohr, 1995; Task Force on Federally Sentenced Women, 1990). It is important to note, however, that studies of programs conducted after full implementation of CSC's woman-centered model of corrections have shown more promise (Blanchette & Eldjupovic-Guzina, 1998; Dowden & Blanchette, 1999).

Studies of gender specific correctional interventions are essential because the law in Canada mandates distinctive programming for women offenders. Section 77 of the Corrections and Conditional Release Act (CCRA; 1992) directs that the Correctional Service of Canada:

  1. Provide programs designed particularly to address the needs of female offenders.
  2. Consult regularly about programs for female offenders with i) appropriate women's groups, and
    ii) other appropriate persons and groups with expertise on, and experience in working with, female offenders.

The disparate treatment needs of women, and the Service's obligation to properly address those, was emphatically restated in Justice Arbour's Commission of Inquiry into Certain Events at Prison for Women (1996). The publication of the Commission of Inquiry rekindled efforts to more strictly adhere to the ideology espoused in Creating Choices.

Correctional programming for women has thus evolved from “gender sensitive” to the contemporary “gender specific”. While the current approach is more palatable for feminists, Bloom (1998) aptly noted that “it is often difficult to understand how effective women-specific services differ from effective services in general” (p.32). There is mounting evidence to suggest that the basic principles of “what works” for men offenders are also applicable to women offenders (Dowden & Andrews, 1999). However, results of some studies suggest that additional or unique parameters should be applied to optimize correctional treatment for women (Austin, Bloom, & Donahue, 1992; Bloom, 1998; Covington, 1998a, 1998b; Doherty, 1998). As such, in discussing effective practice for women offenders, it is necessary to consider elements common to treatment for men, as well as deviations from, and supplements to the standard male model.


Individualized assessment is necessary to match women offenders' needs to treatment resources. In Canada this is accomplished at intake, through the Offender Intake Assessment (OIA) process (Motiuk, 1997). The OIA is an integrated process that incorporates a variety of methodologies. It was designed with the intention of providing a pragmatic, consumer friendly evaluation measure, with good predictive validity. In addition, the assessment process incorporates multiple methods to yield both qualitative and quantitative data (Motiuk & Blanchette, 1998). Therefore, the OIA process is appropriate for use with a variety of populations.3 Data derived from the case-specific OIA process is input into an automated system (Offender Management System; OMS) to contribute to an electronic database for Canada's entire federal offender population. As such, OIA information provides both individual and aggregate (e.g., institutional population profiles) assessment information.

Group and individual assessment is necessary for effective correctional programming. Accordingly, most correctional researchers concur that effective correctional treatment addresses the principles of risk, need, and responsivity. While there is good empirical evidence to support these principles (Andrews, Zinger et al., 1990; Andrews & Bonta, 1998; see also Andrews, Chapter 2 in this Compendium), the substantiating data are based largely on samples of male offenders. As a result, some authors question their applicability to women (Hannah-Moffat, 1999; Hannah-Moffat & Shaw, 2000; Kendall, 1998).

Risk principle

The risk principle posits that level of treatment should be matched to the risk level of the offender. More specifically, intensive services should be provided to higher risk offenders, while lower risk offenders fare better with minimal or no intervention. As mentioned, while there is ample empirical support for the risk principle (Andrews, 1989; Andrews & Bonta, 1998; Andrews, Bonta, & Hoge, 1990; Andrews, Zinger et al., 1990), the research is based almost entirely on samples of men offenders. Accordingly, some authors argue that the concept of risk is “gendered” and “racialized” and should not be applied to minority groups (Hannah-Moffat, 1999).

Recent meta-analytic4 research by Dowden and Andrews (1999) examined the validity of the risk principle for women offenders. The authors included treatment studies that met the following criteria:

  • the study was composed predominantly or entirely of women offenders;
  • the study included a follow-up period;
  • the study compared offenders who had received some form of intervention to a control group who did not receive the primary intervention; and
  • the study included a measure of recidivism (reconviction, rearrest, parole failures).

Dowden and Andrews' meta-analysis tested the risk principle by coding studies as treating “high risk” or “low risk” women. Specifically, treatment groups were categorized as high risk if “the majority of those [participants] in the study had penetrated the justice system at the time of the study and/or had a previous criminal offence” (p. 442). Alternatively, treatment groups that had no criminal offences or had been diverted from the justice system were coded as low risk.

Results revealed that treatment services were more effective with the higher-risk offenders. Specifically, the data (45 effect sizes) generated a 19% reduction in recidivism for high-risk groups, and no treatment effect for low-risk groups. Moreover, when the authors narrowed the focus to include exclusively women treatment studies (rather than predominantly women treatment studies), this effect was even more pronounced, and a 24% reduction in recidivism was observed for the high-risk group. The authors concluded that these data support the risk principle for effective intervention with women offenders.

While the study by Dowden and Andrews provides valuable preliminary insight into the applicability of the risk principle for women, some limitations to their research should be acknowledged. Recall that the basic tenet of the risk principle matches level of service to level of risk. However, Dowden and Andrews' meta-analysis does not fully address this issue, as treatment “dosage”, was not reported. Rather, the authors described reductions in recidivism for treated (versus untreated) groups. Also, the method of partitioning treatment studies into “high” and “low” risk groups was questionable. Specifically, those with a current involvement in the criminal justice system (the high-risk groups) are much more likely to enjoy reductions in recidivism than their low-risk counterparts because they have higher base rates of offending at the outset.

It is important to note, however, that classification of women offenders into “high” and “low” risk groups will continue to present a challenge in prospective studies. In particular, available risk classification schemas, developed on samples of men, lose validity and reliability when applied to women (Blanchette, 1996; Bonta, Pang, & Wallace-Capretta, 1995; Hann & Harman, 1989; Salekin, Rogers, & Sewell, 1997).

The principle pre-release risk assessment instrument used by CSC is the Statistical Information on Recidivism (SIR) scale (Nuffield, 1982). Used primarily for parole decision-making in Canada, the Statistical Information on Recidivism score provides an estimate of the probability that an individual will reoffend within three years after release. Each offender's total score on the SIR scale is a simple summation of (15) item scores, with total scores ranging from -30 (very high risk) to +27 (very low risk). SIR scores have been shown to accurately predict release outcome for non-Aboriginal men offenders (Hann & Harman, 1988; Motiuk & Porporino, 1989). While research results have suggested that the SIR scale is somewhat predictive of release risk for women, its power is considerably less than that for men (Blanchette, 1996; Bonta et al., 1995; Hann & Harman, 1989). Given these results, the SIR scale is not currently used for the evaluation of risk for women offenders.

Assessment of “psychopathy” is also routinely completed for the evaluation of risk in offender populations. In brief, psychopathy refers to a constellation of affective, interpersonal and behavioural traits associated with a marked absence of compassion and a lack of personal integrity. The Psychopathy Checklist- Revised (PCL-R; Hare, 1991) is currently the most widely accepted measure of psychopathy. The PCL-R consists of 20 items, scored on a 3-point scale, on the basis of a semi-structured interview with the offender and institutional file information. Items on the checklist are summed to provide a total score and two sub-scale scores. The first sub-scale (Factor 1) is defined by interpersonal and affective characteristics and is labelled “callous, selfish, and remorseless use of others”. Factor 2 is defined by behavioural traits indicative of a “chronically unstable and anti-social lifestyle”.

Average PCL-R scores derived from women samples are generally lower than those obtained with men samples. The prevalence of psychopathy amongst women offenders, based on PCL-R diagnosis across 5 studies, ranges from 11-31% (Mailloux, 1999); the median rate is 16% (Salekin et al., 1997). Typically, the prevalence rate in men offender samples ranges between 25% and 30% (Hare, 1991); although percentages as low as 11% have been reported (Simourd & Hoge, 2000).

The construct of psychopathy has been developed, and largely defined according to the characteristics of men forensic samples. Research on the PCL-R with women offenders has shown some gender differences in the factor structure; the discrepancy more apparent within factor 2 (behavioural) items (Salekin et al., 1997). Moreover, the applicability of certain PCL-R items to women has been questioned, specifically: grandiose sense of self worth, failure to accept responsibility, revocation of conditional release, and juvenile delinquency (Salekin, Rogers, Ustad, & Sewell, 1998).

To date, only two published studies have examined the predictive utility of the PCL-R with women offenders. The results of one study indicated that the measure is a relatively weak predictor of recidivism for women offenders (Salekin et al., 1998). Conversely, based on a sample of 80 released federal women offenders, Loucks and Zamble (1999) argue that “psychopathy is as important in predicting general offending in female serious offenders as it is in serious male offenders” (p.28). Collectively, the research suggests that there is not currently enough evidence to support the use of the PCL-R in clinical risk assessment of women offenders.

Another standard risk assessment instrument for correctional populations is the Level of Service Inventory- Revised (LSI-R; Andrews & Bonta, 1995). Based on social learning theory, the LSI-R consists of 54 individual items that measure the following risk/need domains: criminal history, education/employment, financial, family/marital, accommodation, leisure/recreation, companions, alcohol/drug problems, personal/emotional, and attitudes. Scoring for the LSI-R is based on a semi-structured interview with the offender and institutional file information. Each item is scored in a dichotomous fashion (0 or 1), where 1 indicates the presence of a risk or need factor. Individual items are then added to provide a composite score; higher overall scores suggest higher risk of recidivism and need for correctional intervention.

The LSI-R is probably the most extensively researched risk classification instrument in North America. The original LSI (Andrews, 1982) was constructed with a development sample of mostly men offenders. Importantly, however, norms were established based on a large sample of both men and women (n = 1,414 women) (Andrews & Bonta, 1995). Moreover, there is a growing body of empirical evidence demonstrating the predictive accuracy of the LSI-R for women offenders in particular (Coulson, Ilacqua, Nutbrown, Giulekas, & Cudjoe, 1996; Gendreau, Goggin, & Smith, 1999; Rettinger, 1998; McConnell, 1996).

The research literature supporting the predictive utility of the LSI-R (and its predecessor, the LSI) has been based predominately on samples of provincial incarcerates and probationers. There is one published study reporting good psychometric properties for the LSI with a sample of federally sentenced offenders (Loza & Simourd, 1994). Based on a sample of 161 men federal offenders, the authors reported that the LSI possesses acceptable psychometric properties and demonstrates convergent validity with measures of relevant criminogenic constructs (SIR scale, PCL-R). Also, while the authors provided good preliminary evidence to support the use of the LSI with federal offenders in terms of its psychometric properties, its utility as a risk prediction measure for this particular population was not examined.

To date, there are no published reports exploring the efficacy of the LSI/LSI-R with federally sentenced women. In an unpublished Honour's thesis, McConnell (1996) tested the predictive ability of the LSI with a sample (n = 50) of federal women offenders. The LSI was scored retrospectively based on file information, and recidivism was defined as conviction for a new offence within three years of release. While the LSI total score accounted for an impressive proportion (36%) of the variance in outcome, subsequent analyses revealed that only two (criminal history, companions) of the ten LSI subscales contributed significantly to the prediction of recidivism. More research using the LSI with samples of federally sentenced women is necessary before firm conclusions regarding its psychometric properties and utility as a risk prediction measure can be drawn.

The Historical Clinical Risk Scheme (HCR-20; Webster, Douglas, Eaves & Hart, 1997; Webster, Eaves, Douglas, & Wintrup, 1995) is a 20-item violence risk assessment instrument that conceptually aligns risk markers into past, present, and future. The ten historic (H) variables consider past behaviour and functioning; they are static or unchangeable. The five clinical (C) items reflect current, dynamic correlates of violence. Finally, the five risk management (R) items concern the future, focusing attention on situational post-assessment factors that may either aggravate or mitigate risk (Douglas, 1999).

The HCR-20 has demonstrated robust psychometric properties and research results have been favourable in terms of its utility as a risk prediction measure for both men and women. However, the vast majority of these studies have drawn samples from civil psychiatric settings (Douglas, Ogloff, & Nicholls, 1997; Douglas, Ogloff, Nicholls, & Grant, 1999; Klassen, 1996; Nicholls, Ogloff, & Douglas, 1997; Ross, Hart, & Webster, cited in Douglas, 1999) or forensic psychiatric (Belfrage, 1998; Dernevik, 1998; Douglas et al., 1998; Grann, Belfrage, & Tengström, 2000).

To date, there have been only two published studies using the HCR-20 risk assessment scheme with regular offender samples (Belfrage, Fransson, & Strand, 2000; Douglas & Webster, 1999). Recently, Belfrage et al. (2000) documented the HCR-20's ability to predict institutional violence in a sample of 41 men in Swedish maximum-security prisons. Douglas and Webster (1999) coded the HCR-20 for a sample of 75 Canadian male maximum-security inmates. Data from their postdictive research offered strong support for the use of the HCR-20 in assessing/classifying risk for violence. Collectively, the results of these studies offer good preliminary support for the use of the HCR-20 for predicting risk (especially violent) in men maximum-security inmates. There is currently no evidence to suggest that the HCR-20 would be as valuable in measuring risk in women offender populations.

The Violence Risk Appraisal Guide (VRAG; Harris, Rice, & Quinsey, 1993) was designed to predict risk of violent recidivism among mentally disordered offenders. The construction sample consisted of 618 men admitted to a maximum-security psychiatric facility between 1965 and 1980. The 12-items VRAG measures a variety of static risk factors including demographic, childhood, criminal history, and victim information data. Items are weighted and added together to derive a composite score, ranging from -27 to +35; higher scores reflect a greater probability of violent recidivism.

Similar to the HCR-20, research support for the VRAG is largely derived from samples of male offenders in psychiatric settings (Grann et al., 2000; Rice & Harris, 1997). There are studies demonstrating sound psychometric properties (Loza & Dhaliwal, 1997) and predictive efficacy in sex offender (Bélanger & Earls, cited in Quinsey, Harris, Rice, & Cormier, 1998) and violent offender (Kroner & Mills, 1997) samples. Again, however, all substantiating data are based entirely on samples of men. As such, the VRAG is currently not appropriate as a risk classification/prediction instrument for women offenders.

In summary, the most well researched and commonly used risk assessment instruments for offender populations in Canada include: the SIR scale (Nuffield, 1982), PCL-R (Hare, 1991), LSI-R (Andrews & Bonta, 1995), HCR-20 (Webster et al., 1995) and the VRAG (Harris et al., 1993). Unfortunately, there is currently not sufficient evidence to support the use of any of these measures with federal women offenders. Thus, concerns with the method by which Dowden and Andrews (1999) classified their sample according to risk cannot be easily addressed. This limitation is a consequence of the scarcity of relevant empirical data on women offenders.

While Dowden and Andrews provided a preliminary contribution to research on the applicability of the risk principle for women offenders, more empirical evidence is needed for confirmation. As published studies on women offenders continue to accumulate, it is hoped that two critical elements will be addressed in future meta-analyses. First, prospective studies should include detailed descriptions of their treatment groups. This will allow meta-analytic researchers to more accurately code treatment intensity (also called “dosage” or “level”). Second, a risk assessment instrument with strong substantiating data is required to accurately classify federally sentenced women into “low”, “moderate”, and “high” risk groups.

This review has highlighted the need to test the reliability and validity of existing risk assessment instruments for women. However, it would be even more judicious to develop and validate a gender-specific model of risk assessment for women. A recently published study indicated that risk factors for women differ substantially from those for men. As well, separate (gender-specific) risk assessment instruments were noted to improve the prediction of reoffending, especially for women (Funk, 1999).

Need principle

The need principle distinguishes between criminogenic and noncriminogenic needs. Andrews and Bonta (1998) offer a clear definition: “Criminogenic needs are a subset of an offender's risk level. They are dynamic attributes of the offender that, when changed, are associated with changes in the probability of recidivism. Noncriminogenic needs are also dynamic and changeable, but these changes are not necessarily associated with changes in recidivism” (p. 243). Clearly, criminogenic needs are a subset of risk: noncriminogenic needs are not. Still, some authors claim that “there is slippage between the terms risk and need” (Hannah-Moffat & Shaw, 2000, p. 58; see also Canadian Elizabeth Fry Societies, 1998). More precisely, there is overlap between the concepts of “risk” and “need”, as explained above.

Fundamentally, the need principle asserts that in order to reduce recidivism, treatment services should target criminogenic needs. Promising targets for intervention include: anti-social attitudes and feelings, anti-social associates, poor self-control, self-management, and/or problem-solving skills, substance abuse problems, lack of education and/or vocation, lack of familial ties or dysfunctional family relationships, and poor use of recreational/ leisure time (Andrews & Bonta, 1998; Motiuk, 1997). The general acceptance of these dynamic factors as criminogenic is based on a considerable body of research (Andrews et al., 1990; Andrews & Bonta, 1998; Andrews, Dowden, & Gendreau, 1999; Lösel, 1995). However, the need principle's applicability to women has been disputed in the correctional literature. Again, the scepticism derives from the fact that the supporting research is based on samples of men offenders.

Regarding women offenders, it is not the need principle per se that has been subject to scrutiny. Rather, the debate is focused on the specific nature of which needs are criminogenic for this particular group. Based on a review of the current literature, Howden-Windell and Clark (1999) concluded “it is evident from the empirical evidence currently available that the criminogenic factors associated with male offenders are relevant to female offenders but their level of importance and the nature of association may differ”. There is some research to support this suggestion (Dowden & Andrews, 1999; Research Branch, 2000; Simourd & Andrews, 1994), though the strength of the evidence varies, dependent on the dynamic factor in question.

A recent study conducted by CSC's Research Branch (2000) demonstrated that a variety of dynamic factors, assessed at intake, could reliably predict recidivism in released women offenders. All women evaluated through the OIA process and released before December 1997 were included in the representative sample (n = 420). The average follow-up period was approximately one year, with a range from one day to 3.5 years. The dynamic factors assessed by the OIA include: employment/education, marital/family, associates, substance abuse, attitudes, community functioning, personal/emotional orientation. Analyses revealed a consistent relationship between need level rating and recidivism rate. As expected, women with higher need level ratings were more likely to recidivate than those with lower need level ratings: this was true for all seven need domains. These findings suggest that the seven dynamic factors assessed at intake can be classified as criminogenic needs for women offenders. However, corroborative research in most of these need domains is either conflicting or non existent.

While there are a few studies examining the relationship between employment/education needs and women offender recidivism, results of these are inconsistent. With a large sample (n = 441) of women offenders, Rettinger (1998) identified education/ employment as an important contributor to the prediction of recidivism. Similarly, in their meta-analysis on the correlates of female delinquency (n = 34 effect sizes), Simourd and Andrews (1994) found that ‘educational difficulties' were moderately to strongly related to delinquent behaviour in girls. In contrast, Dowden and Andrews' (1999) recent meta-analysis revealed that programs targeting school/work (n = 7 effect sizes) for women offenders showed a non-significant negative correlation with reductions in recidivism. Finally, using a large representative sample (n = 136) of released federal women offenders, Bonta et al. (1995) showed that employment was not significantly related to recidivism. Thus, while there is evidence to suggest that education/employment variables predict recidivism in samples of men offenders (Gendreau, Goggin, & Gray, 2000), the results are still equivocal in regards to whether this domain is criminogenic for women.

Even considering the more abundant literature based on men offenders, researchers are “far from elucidating the causal relationship between family life and adult criminality” (Oddone-Paolucci, Violato, & Schofield, 1998; p. 20). However, some authors have suggested that family issues are important treatment targets for women offenders in particular (Austin et al., 1992; Bloom, 1998; Federally Sentenced Women Program; cited in Hannah-Moffat, 1997). Recent research supports this contention. Based on a review of the literature, Leschied, Cummings, Van Brunschot, Cunningham, and Saunders (2000) reported that dysfunctional family processes and family dynamics are instrumental in promoting and maintaining aggressive behaviour in adolescent girls. In their meta-analysis, Dowden and Andrews (1999) found that programs treating family process issues (n = 9 effect sizes) yielded the strongest reductions in reoffending for women samples. Simourd and Andrews (1994) reported that a poor parent-child relationship (attachment and supervision) was a moderate correlate of offending behaviour among female youths. However, the same authors noted no significant association between family structure or parental problems and criminality.

In one of the few studies looking at family variables as criminogenic needs for adult women, Bonta et al. (1995) reported that, while having dependants was not associated with post-release outcome, single-parents showed significantly higher recidivism rates than those with partners. Rettinger's (1998) research results confirmed that parenthood was not predictive of recidivism, though contrary to findings by Bonta et al., the data also showed no association between single-parenthood and outcome. Results of Rettinger's study did suggest, however, family or marital conflict was predictive of violent recidivism. These results were not supported by data presented by Loucks and Zamble (1999). Based on a sample of 80 released federally sentenced women, these authors reported that family cohesiveness did not contribute to the prediction of recidivism.

There is a lack of consensus regarding whether family variables constitute important criminogenic needs for women. It is suggested that conflicting research results might be derived from the use of diverse definitions and measures for the “marital/family” construct. Still, while there is some disagreement between research findings, the greater evidence suggests that family variables warrant further investigation as a potential criminogenic need area for women.

The dynamic factor of anti-social associates is routinely hailed as among the most potent predictors of recidivism, and there-fore a priority treatment target (Andrews & Bonta, 1998; Andrews et al., 1999; Andrews, Zinger et al., 1990). Although the majority of the evidence is based on samples of men offenders, research on women offenders can be considered conclusive: anti-social associates represents an important criminogenic need domain for women. Dowden and Andrews (1999) reported a strong positive association between correctional programming in this area and reduced reoffending for women. Similarly, Simourd and Andrews (1994) noted that anti-social peers or attitudes comprised the greatest risk factor for female youths. Based on a sample of 81 released federally sentenced women, Blanchette and Motiuk (1995) demonstrated that “criminal associates” was a powerful predictor of violent recidivism. Rettinger (1998) replicated those results with a larger sample (n = 441) of provincially sentenced women.

Antisocial attitudes are also considered amongst the most valuable treatment targets to reduce recidivism in offender populations (Andrews & Bonta, 1998; Andrews et al., 1999; Andrews, Zinger et al., 1990). Compared to their women counterparts, anti-social attitudes are more prevalent among men offenders (Motiuk, 1997). Notwithstanding that, research evidence suggests that this dynamic factor discriminates between women inmates by security level (Blanchette, 1997a) and is criminogenic in nature, regardless of gender. Simourd and Andrews' (1994) meta-analytic results suggested that anti-social attitudes (and peers) are the most important risk factors for female youths. Dowden and Andrews (1999) demonstrated that targeting anti-social attitudes in treatment renders significant reductions in reoffending for women samples. Finally, Rettinger's (1998) study showed that anti-social attitudes/peers predicted violent recidivism for women offenders.

The relationship between substance abuse and criminal activity is well documented: about two-thirds of offenders experience some degree of substance abuse problems (Boland, Henderson, & Baker, 1998). A recent review by Weekes, Moser, and Langevin (1998; cited in Dowden & Brown, under review) concluded that there is a consistent positive association between substance abuse and various forms of general and violent criminal activity. This conclusion supports results of other studies and theoretical arguments which suggest that substance abuse is a criminogenic need (Andrews & Bonta, 1998; Andrews, Zinger et al., 1990). However, there is no consensus in the literature regarding whether substance abuse constitutes a criminogenic need for women offenders in particular.

Based on results of a large sample study with adult inmates (1,030 men and 500 women), McClellan, Farabee, and Crouch (1997) reported that substance abuse problems were stronger predictors of criminal activity for women than men. Rettinger (1998) found that substance abuse was predictive of both general and violent recidivism for women offenders. Similarly, results of a study by Dowden and Blanchette (1999) revealed that treated women substance abusers were significantly less likely to reoffend than their untreated counterparts.

It appears however, that for every study that identifies sub-stance abuse as a criminogenic need for women, there is another to negate those findings. For instance, results of Dowden and Andrews' (1999) meta-analysis suggested that substance abuse is not a valuable treatment target for women offenders (n = 5 effect sizes). Similarly, Bonta et al. (1995) found that substance abuse was not predictive of post-release outcome for their sample of federal women offenders. In a recent meta-analysis on the role of substance abuse in predicting recidivism, Dowden and Brown (under review) reported that alcohol abuse was a weak predictor, while drug abuse was a moderately strong predictor for women offenders (n = 7 effect sizes). However, Loucks and Zamble (1999) reported that drug abuse was not a significant predictor of recidivism in their sample (n = 80) of released women offenders.

Blanchette (1996) found that varying the definition/method of measurement for “substance abuse problem” affected the data accordingly. Specifically, with a sample of 76 federally sentenced women offenders, analyses revealed that meeting diagnostic criteria for substance abuse disorder was not predictive of recidivism. However, when the variable was re-defined according to whether or not the offender had consumed alcohol/drugs prior to the commission of her original offence, “substance abuse” was predictive of recidivism. This suggests that, of women offenders who meet diagnostic criteria for substance abuse or dependence disorders, there is a subset for which the need is criminogenic. For other women offenders, their substance abuse problems do not represent a criminogenic need.

Measures of community functioning include leisure, accommodation, finance, support, deportment, and health. According to the results of a recent meta-analysis, the research support for “community functioning” as a criminogenic need is moderate, at best (Gates, Dowden, & Brown, 1998). The authors identified 20 studies that yielded 79 effect sizes pertaining to community functioning variables. An overall weighted mean effect size of 0.10 was obtained.5 While the majority of the effect sizes were based on studies of men, the second author desegregated the data by gender and found 12 effect sizes in studies of exclusively women offenders. The weighted mean effect size of “community functioning” variables for women only was 0.09 (Dowden, August 15, 2000, personal communication). While there is no clear evidence that the broad area of “community functioning” is criminogenic for women, there is a strong possibility that particular subcomponents of this domain might be appropriate treatment targets. For instance, Gates, Dowden, and Brown (1998) showed that “leisure” produced a very strong effect size (0.24). Rettinger (1998) reported that “accommodation” was a strong predictor of both general and violent recidivism in her sample of 441 provincial women offenders. Unfortunately, there is currently not enough empirical data to reinforce these findings or to examine the predictive utility of other subcomponents of this domain for women.

With a sample of 420 released federal women offenders, CSC's Research Branch (2000) demonstrated a strong correlation between a global measure of personal/emotional orientation (assessed at intake) and recidivism. Specifically, within the aver-age one-year post-release follow-up, 11% of those assessed as having “no difficulty” were returned to custody. Twenty-two percent of those assessed as having “some difficulty” recidivated, compared to 34% of those assessed as having “considerable difficulty”. These findings were partially supported6 by Robinson, Porporino, and Beal (1998), though the authors stressed that “the state of the literature on personal/emotional needs factors remains under-developed particularly with respect to the predictors of recidivism” (p. 77). The problem with the lack of literature is greatly increased when one considers women-specific research. As such, there is little information with respect to which (if any) specific components of the personal/emotional domain are criminogenic in nature. Notwithstanding that, Dowden and Andrews (1999) showed that programs targeting anti-social cognition and self-control deficits in women (n = 8 effect sizes) reduced recidivism, on average, 32%. Similarly, a measure of anti-social thinking -- the Psychological Inventory of Criminal Thinking Styles (PICTS) -- was found to be moderately successful in predicting institutional adjustment and post-release outcome for women offenders (Walters & Elliott, 1999).

It is apparent from the above review of seven specific dynamic factors that there is a lack of predictive research with samples of women offenders. Moreover, what little is available is often inconsistent and therefore collectively provides only modest insight into which particular needs are criminogenic for women offenders. While there is compelling evidence to suggest that “anti-social associates” and “anti-social attitudes” are valuable treatment targets for women, the status of other dynamic factors including education/employment, marital/family, substance abuse, community functioning, and personal/emotional orientation remains equivocal.

Several authors have suggested that women offenders have additional criminogenic needs, though more research is required to confirm the relationship of these variables to recidivism (Federally Sentenced Women Program, 1994; cited in Hannah-Moffat, 1997; Jackson & Stearns, 1995; Koons et al., 1997; Leschied, Cummings, VanBrunschot, Cunningham, & Saunders, 2000). Dynamic factors that are commonly cited as women-specific criminogenic needs can be generally subsumed in the “personal/emotional” domain, and include low self-esteem, past and current victimization, and self-injury/attempted suicide.

Based on research evidence with men samples, most empiricists believe that self-esteem is not criminogenic (Andrews & Bonta, 1998; Andrews et al., 1990; Gendreau, Little, & Goggin, 1996). However, qualitative research by others has suggested that self-esteem is a promising treatment target for women offenders (Koons et al., 1997). Although relevant gender-specific empirical data are scarce, there is some evidence to support results by Koons and her colleagues. Simourd and Andrews' (1994) meta-analysis of the correlates of delinquency found an effect size of 0.10 for women and 0.09 for men (n = 14 studies each) for a predictor domain labelled “personal distress”. The applicability of these findings to self-esteem research, however, is limited; “personal distress” also included effect sizes relating to anxiety and psychopathology. Larivière (1999) referenced several studies correlating low self-esteem to acts of violence against weaker, vulnerable victims (as in spousal abuse and child abuse). Moreover, he cited six studies linking low self-esteem in women to acts of child abuse (5 studies) and neglect (1 study).

Larivière's own research also supports this position. Using meta-analytic techniques, he reviewed 39 studies containing 80 effect sizes pertinent to self-esteem. Results revealed a significant overall effect (r = -0.17), suggesting a strong association between self-esteem and anti-social behaviour (general delinquency, aggression, and violence). Notably, the magnitude of the effect more than doubled (r = -0.38) when the focus was narrowed to women offenders (n = 13 studies). Larivière argues that this finding is not surprising, since women tend to express more guilt about criminal and aggressive behaviours, experience more anxiety about the harm they have caused, and demonstrate less support for the use of violence (Campbell, 1995; cited in Larivière, 1999). The author cautioned, however, that women samples included in the meta-analysis were over-represented by subjects who had engaged in child abuse, possibly resulting in an increased effect size. Notwithstanding that, these results have important implications for treating women offenders; particularly those convicted of child neglect or abuse.

Compared to men inmates, women inmates report significantly more victimization experiences (McClellan et al., 1997). It is now incontestable that there is a strong correlation between experiences of abuse and criminal behaviour (Howden-Windell & Clark, 1999), with the vast majority of women offenders having been victimized at some point in their lives (Blanchette, 1996; Owen & Bloom, 1995; Shaw, 1991a; 1991b). One study revealed self-reported victimization rates as high as 82% among Canadian women offenders (Task Force on Federally Sentenced Women, 1990). Moreover, in comparison to both the general population of women, and to men offenders, women offenders are more likely to have experienced victimization that is violent, sexual, incestuous, committed by numerous perpetrators, and extended over a long period of time (Task Force on Federally Sentenced Women, 1990).

The appallingly high incidence of abuse reported by the Task Force has been supported by independent Canadian research. Tien, Lamb, Bond, Gillstrom, and Paris (1993) noted that 81% of their sample of women offenders reported experiencing some form of abuse (sexual, physical, or psychological) in their current relationship. More recently, these findings were sup-ported by Bonta et al. (1995), where 61% of their federal women offender sample reported past physical abuse, and 54% reported past sexual abuse. Similarly, in Blanchette's (1996) sample of federal women offenders, 61% were identified as victims of child-hood abuse, and 59% were identified as victims of abuse in adult-hood. Data from U.S. women inmate samples is comparable, with about 60% reporting childhood victimization, and about 75% reporting experiencing abuse as an adult (McClellan et al., 1997). Thus, there is a well-documented link between victimization experiences, both in childhood and adulthood, and criminal behaviour in women. The exact nature of this relationship, however, remains nebulous.

Although research by Koons et al. (1997) suggests that past victimization represents an important treatment target for women offenders, the authors do not necessarily suggest that “victimization” represents a need that is criminogenic. Loucks (1995) examined the nature of the association between victimization experiences and anti-social behaviour in a sample of federally sentenced women inmates (n = 100). Results of his study revealed that pre-adolescent sexual abuse correlated positively with both institutional convictions and violence (criminal and institutional), while post-adolescent physical abuse correlated positively with institutional convictions and criminal violence. However, when the victimization variables were entered into a prediction equation with other variables, their value was negligible.

In an investigation into the predictors of recidivism among Canadian federally sentenced women, Bonta et al. (1995) reported that victimization experiences were not statistically predictive, with the exception of physical abuse as an adult. Importantly, those who had experienced physical abuse as an adult were actually less likely to reoffend than their counter-parts. These findings were supported by Blanchette (1996), who, controlling for time at risk in the community, noted no relationship between victimization experiences and recidivism. Moreover, these results were sustained regardless of how ‘recidivism' was defined.7 Similar to findings reported by Bonta et al., a negative association (r = -0.24) was reported between abuse in adulthood and criminal (new conviction) recidivism; the correlation approached statistical significance. Finally, results presented by Rettinger (1998) also suggest that abuse experiences are not statistically predictive of recidivism or of violent recidivism in women offenders.

Collectively, the research to date suggests that victimization, although very common amongst women offenders, is not a criminogenic need. Despite this, the astonishingly high prevalence of survivors in the correctional system signals a requirement for service providers to address this issue. Many women suffer from post-traumatic symptoms that can impede progress in addressing criminogenic need areas. In Canada, this has been recognized, and “Survivors of Abuse and Trauma” has been incorporated into core8 programming for federal women offenders.

Women in prison show much more frequent mental health problems than women in general, men in general and incarcerated men, including higher levels of depression, suicidal and self-injurious behaviour (Blanchette, 1996; Blanchette, 1 997b; Loucks & Zamble, 1994). Studies of Canadian federal women offenders report that about 54 to 59% have engaged in some form of self-injurious behaviour such as head banging, cutting, burning, or slashing (Heney, 1990; Loucks, 1995). Rates of attempted suicide among federal women offenders are reported at 48% (Loucks & Zamble, 1994), with a range of 20 to 71%, depending on security level (Blanchette, 1997b).

With a sample of 100 federal women offenders, Loucks (1995) examined the relationship between self-harm and criminal behaviour. The researcher used a broad definition of self-harm, including any intentional action that resulted in physical harm to the self; he did not distinguish between actions that were intended to commit suicide and those that were for other reasons (e.g., attention seeking). Results revealed that 54% of the sample reported engaging in at least one incident of self-harm at some point in their life. Moreover, self-harm was found to be positively correlated with both criminal convictions (r = 0.25) and criminal violence (r = 0.24).

While the prediction research in this area is not copious, those studies that do exist suggest that self-injury/attempted suicide is criminogenic in nature. Bonta et al. (1995) found that self-injury was predictive of general recidivism (new convictions or parole revocations) in a sample of federal women offenders; 78% of those with a history of self-injurious behaviour recidivated, versus 25% of those with no such history. Blanchette and Motiuk (1995), reported that a history of attempted suicide was a potent predictor of violent recidivism (r = 0.47; p<0.001) in a sample of 81 federally sentenced women. Statistical analyses further revealed that, together with two other variables (expectations about incarceration, associates), a history of attempted suicide accounted for 45% of the variance in violent recidivism. These findings were replicated with a larger sample of provincially sentenced women offenders (n = 441), where a history of self-injury was predictive of violent recidivism (Rettinger, 1998).

It is apparent from this review of the literature that, as suggested by Dowden and Andrews (1999), the need principle is applicable to women offenders. What is more contentious is whether the traditional9 criminogenic needs are also applicable to women, or whether women have unique criminogenic needs. The research suggests that both are true: while some of the traditional dynamic factors, such as anti-social attitudes and associates are criminogenic for women, there is also some evidence that they have unique criminogenic needs, such as propensity to self-injure or attempt suicide. Prospective prediction studies on self-esteem will more firmly determine whether it is a gender-specific criminogenic need for women offenders.

Responsivity principle

The responsivity principle suggests delivering treatment services in a style and mode that is conducive to the ability and learning style of the offender (Andrews & Bonta, 1998). Responsivity considerations are both general and specific.

The general responsivity principle refers to broad treatment strategies (and their theoretical foundations): it suggests that the most powerful approaches to produce change are behavioural/social learning/cognitive behavioural styles of service. Within the general correctional literature, there is very strong evidence to sup-port the general responsivity principle (Andrews, Zinger et al., 1990; Izzo & Ross, 1990; Lösel, 1995). While most of the relevant research has used men samples, there is also preliminary evidence to suggest that behavioural/social learning/cognitive behavioural approaches are most effective for women's correctional interventions (Austin et al., 1992; Dowden & Andrews, 1999).

Specific responsivity factors consider individual offender characteristics such as education level/literacy, mental health, inter-personal anxiety, race, and gender. By identifying specific responsivity issues, treatment services can be better matched to the offender. Prendergast, Wellisch, & Falkin (1995) highlighted the importance of specific responsivity factors: “whether women offenders obtain the services they need depends not only on the avail-ability of services but also on the ability of programs to identify their clients' needs and match them to appropriate services” (p. 252).

Consideration of the demographics and history of the women offender population is vital in the development of gender-responsive programming. Some authors have suggested that an understanding of the unique life experiences of women, the context in which they live, and their pathways to crime, are essential in gender-specific program planning and delivery (Bloom, 1998; Chesney-Lind, 1998, Covington, 1998b). Having said that, however, it has been noted that “there is no uniform procedure or instrument for identifying women's needs, nor is there a commonly accepted theory-based method for matching clients to programs and services” (Prendergast, Wellisch, & Falkin, 1995, p. 252). Thus, while few would argue that gender specific issues should be reflected in the development and delivery of women-centred programs, there are still no quantitative empirical studies examining gender as a specific responsivity consideration.

An increasing number of theoretical reports and qualitative evidence is suggesting that adherence to particular programming principles increases treatment efficacy for women offenders. The research conducted by the Task Force on Federally Sentenced Women (1990) suggested five basic principles should drive correctional programming for women. These include: empowerment, meaningful and responsible choices, respect and dignity, supportive environment, shared responsibility.

  • Empowerment is the process through which women gain insight into their personal situation, identify their strengths, and are supported and challenged to take positive action to gain greater control of their lives. There is independent research evidence to suggest that the empowerment model of skill building helps develop competencies and enables women to achieve independence (Austin et al., 1992; Blanchette & Eldjupovic-Guzina, 1998).
  • Meaningful and responsible choices provide women with options that allow them to make responsible choices that relate to their needs, past experiences, culture, values, spirituality, abilities and skills. A history of dependence (e.g., alcohol/drugs, men, financial assistance) has denied many federally sentenced women the opportunity and/or ability to make meaningful and responsible choices in their lives (Blanchette & Eldjupovic-Guzina, 1998; Office of the Deputy Commissioner for Women, 2000).
  • Respect and dignity refers to the reciprocal respect that is needed among offenders, staff, and between the two. It is expected that the new philosophy of women's corrections will engender more respect and dignity with the dynamic security, community-living model. For instance, Primary Workers, who work with the women to establish and address treatment goals, have replaced “correctional officers”. These initiatives are important progressions because it has been suggested that women are more relationship-oriented and that treatment should be provided accordingly (Covington, 1998a; Garcia et al., 1997).
  • Supportive environment is required to promote physical and psychological health and personal development for women offenders. The importance of a supportive environment in programming for all offenders is highlighted in CSC's Correctional Strategy.
  • Shared responsibility refers to the suggestion that all levels of government, corrections, volunteer organizations, businesses, private sector services and the community have a role to play in the development of support systems and continuity of service for federally sentenced women.

As noted earlier, the five basic principles of appropriate treatment for women were developed based primarily on qualitative research findings. The principles have been accepted as gender-specific responsivity elements, and are incorporated into current practice with federal women offenders. It is hoped that prospective quantitative data will provide corroborative evidence for their effectiveness in treatment for women offenders.

Other suggested responsivity considerations in programming for women include using ethnically diverse staff (Austin et al., 1992) including a balance of professionals and ex-offenders for role modeling. In Canada, it is particularly important to include Aboriginal staff and program facilitators in programming for women. While Aboriginal people comprise less than three per-cent of the Canadian population, they represent over 20% of women inmates. Compared to non-Aboriginal inmates, Aboriginal inmates are younger, less educated, and assessed as higher risk and having more intensive treatment needs (Finn, Trevethan, Carrière, & Kowalski, 1999). These issues should also be attended to for optimal treatment effectiveness.

Possibly the most important feature of gender-specific responsivity is recognition of the fact that equality does not mean sameness. To provide the most effective interventions, services within the correctional system must be responsive to the unique needs and learning styles of women offenders. Treatment should be comprehensive, woman-centred, and holistic in nature. Intervention should be coordinated with an appropriate continuum of care and resources to bridge between institutional and community settings.


It is well documented that programs for women offenders are sorely lacking in solid outcome evaluation data. This is true in both Canada (Kendall, 1998) and the United States (Austin et al., 1992; Koons et al., 1997). Current program development strategies should include “built-in” assessment into their treatment paradigm. This means that the collection of data related to treatment effectiveness would be an integrated component of the program. Moreover, ongoing process and impact evaluation data will provide an opportunity for continuous improvement of program content and service delivery methods.

Evaluation is particularly challenging for women's programs, as a number of unique challenges are presented. Women offenders issues are situated in a highly visible context. Sound, gender-specific program theory is non-existent, and researchers continue to debate the most appropriate methodology to use in evaluative studies. The small, heterogeneous, and geographically dispersed population of federally sentenced women further impedes the collection of both qualitative and quantitative data.

Creative approaches are required to mitigate disadvantages in evaluating services for women offenders. As mentioned at the beginning of this chapter, theses include consideration of the contextual framework and views of the participants, usage of a multi-method approach, and attending to structural or environmental issues such as management support for the program. In quantitative research using standardized measures, it is paramount to ensure that there are normative and validity data available for women offenders.

Evaluation of new approaches and programs for women offenders is especially important to gauge their success against the more traditional programs, which have been developed primarily for men. Despite additional obstacles in assessing services for women offenders, researchers and practitioners are continuing to identify and refine the most effective strategies for service development and delivery.


Canada is leading a new era of correctional services for women offenders. CSC now has good descriptive data profiling the women offender population; this is maintained and updated on an ongoing basis. In recognition of the lower risk levels and unique needs and responsivity issues of women offenders, gender-specific policies and strategic planning tactics have been implemented. The Service has begun to introduce women-centred classification and treatment paradigms; the practice of offering traditional “gender-neutral” services is dissipating. As reflected in Section 77 of the Corrections and Conditional Release Act, it is no longer acceptable to “add women and stir”.

At the federal level, a new philosophy of corrections for women is operational. The mother-child and peer support10 programs are just two examples of how the system is adapting to respond to the distinctive needs of women. The closure of Prison for Women and construction of the regional facilities has begun to address needs related to linguistic and cultural diversity in the small federal women offender population. The community-living style accommodations at the regional facilities is the first step in a reconciliation of the incongruity between a justice system that inevitably fosters dependence and a women population that needs to become more independent.

While progress has been both swift and substantial, there is still room for improvement in delivering effective correctional services to women offenders. Amelioration will result from rigorous evaluation (including cost-benefit analysis) of new women-centred programs. Austin and colleagues (1992) noted the same deficiency in the United States, where “information on the long-term effectiveness of ...gender-specific correctional treatment strategies for women is non existent” (p. 21).

There is some preliminary evidence regarding the criminogenic needs of women offenders. However, much more research is required before confident conclusions can be drawn. It is important to note, as well, that many women have non-criminogenic needs that may prevent them from successfully participating in correctional programs. As discussed earlier, victimization issues are important treatment targets in this context. In community treatment programs, addressing specific non-criminogenic needs is particularly important. For instance, lack of transportation or lack of appropriate childcare arrangements may prevent women from attending programs addressing true criminogenic needs. As such, these are important considerations in program development for women offenders in the community (Bloom, 1998).

In conclusion, a review of the research suggests that the qualities of effective programs in general are also qualities of effective programs for women. More specifically, programs should attend to the principles of risk, need, responsivity. There are several caveats, however. First, there is still no firm support for the ability of any actuarial tool to predict risk for both federal and provincial women offenders. Also, the criminogenic nature of many need areas for women remains unclear. Finally, more research is needed to examine gender as a specific responsivity factor.

The importance of recognizing the distinctive qualities of women offenders is inestimable. While there are many similarities in the social characteristics of men and women offenders, there are also considerable gender differences that must continue to be reflected in the development of women-centered correctional interventions.

1 Correctional Service of Canada

2 Beginning in 1972, the federal government entered into Exchange of Services Agreements (ESAs) with the provinces. ESAs provide an opportunity, under certain conditions, for federally sentenced offenders to serve their sentences in provincial institutions.

3 While the OIA is used at intake, it is important to note that a successful multi-evaluation process, the Reintegration Potential Reassessment, co-exists for women offenders under community supervision.

4 Meta-analysis is a statistical method to aggregate data across numerous studies, providing an “average” result.

5 An overall effect size of 0.10 translates into an approximate 10% reduction in recidivism for treated groups.

6 Some components of the Personal/Emotional domain were supported (e.g., self-control, problem-solving deficits, lack of interpersonal skills) as criminogenic needs, while others (e.g., mental ability, mental health problems) were not.

7 Various definitions of “recidivism” were used, including: return to custody for any reason, revocation for technical violation, new criminal conviction, and new violent criminal conviction.

8 Core programs are priority interventions that must be widely available to offenders in institutions and the community. CSC's Correctional Strategy supports four core programs for women: living skills (e.g., cognitive skills, anger management, parenting), literacy and continuous learning, substance abuse, and survivors of abuse and trauma.

9 Traditional criminogenic needs are those based largely on research with men offenders, including: employment/education, marital/family, associates, substance abuse, attitudes, community functioning, personal/ emotional orientation

10 For more information on the Peer Support program, see Blanchette, K., & Eldjupovic-Guzina, G. (1998).


Andrews, D. A. (1982). The Level of Supervision Inventory (LSI): The first follow-up. Toronto, ON: Ontario Ministry of Correctional Services.

Andrews, D. A. (1989). Recidivism is predictable and can be influenced: Using risk assessments to reduce recidivism. Forum on Corrections Research, 1(2), 1 1-18.

Andrews, D. A., & Bonta, J. (1995). LSI-R: The Level of Service Inventory -- Revised. Toronto, ON: Multi-Health Systems.

Andrews, D. A., & Bonta, J. (1998). The psychology of criminal conduct. Cincinnati, OH: Anderson Publishing Company.

Andrews, D. A., Bonta, J., & Hoge, R. D. (1990). Classification for effective rehabilitation: Rediscovering psychology. Criminal Justice and Behavior, 17, 19-52.

Andrews, D. A., Dowden, C., & Gendreau, P. (1999). Clinically relevant and psychologically informed approaches to reduced reoffending: A meta-analytic study of human service, risk, need, responsivity and other concerns in justice contexts. Manuscript submitted for publication.

Andrews, D. A., Zinger, I., Hoge, R. D., Bonta, J., Gendreau, P., & Cullen, F. T. (1990). Does correctional treatment work? A clinically relevant and psycho-logically informed meta-analysis. Criminology, 28, 369-404.

Arbour, L. (1996). Commission of inquiry into certain events at the Prison for Women in Kingston. Ottawa, ON: Public Work and Government Services Canada.

Austin, J., Bloom, B., & Donahue, T. (1992). Female offenders in the community: An analysis of innovative strategies and programs. Washington, DC: National Institute of Corrections.

Belfrage, H. (1998). Implementing the HCR-20 scheme for risk assessment in a forensic psychiatric hospital: Integrating research and clinical practice. Journal of Forensic Psychiatry, 9, 328-338.

Belfrage, H., Fransson, G., & Strand, S. (2000). Prediction of violence within the correctional system using the HCR-20 Risk Assessment Scheme. Journal of Forensic Psychiatry, 11(1), 167-175.

Blanchette, K. (1996). The relationship between criminal history, mental disorder, and recidivism among federally sentenced female offenders. Unpublished master thesis, Ottawa, ON: Carleton University.

Blanchette, K., (1997a). Risk and need among federally sentenced female offenders: A comparison of minimum-, medium-, and maximum-security inmates. Research Report R-58. Ottawa, ON: Correctional Service of Canada.

Blanchette, K. (1997b). Maximum-security female and male federal offenders: A comparison. Research Report R-53. Ottawa, ON: Correctional Service of Canada.

Blanchette, K., & Eldjupovic-Guzina, G. (1998). Results of a pilot study of the peer support program for women offenders. Research Report R-73. Ottawa, ON: Correctional Service of Canada.

Blanchette, K. & Motiuk, L. L. (1995). Female offender risk assessment: The case management strategies approach. Poster presented at the Annual Convention of the Canadian Psychological Association, Charlottetown, Prince Edward Island.

Bloom, B. (1998). Beyond recidivism: Perspectives on evaluation of programs for female offenders in community corrections. The ICCA Journal on Community Corrections, p. 32-35.

Boland, F., Henderson, K., & Baker, J. (1998). Case need domain: Substance abuse assessment review. Forum on Corrections Research, 10(3), 32-34.

Bonta, J., Pang, B., & Wallace-Capretta, S. (1995). Predictors of recidivism among incarcerated female offenders. The Prison Journal, 75, 277-294.

Canadian Elizabeth Fry Societies (1998). Position paper regarding the classification and carceral placement of women classified as maximum security prisoners. Ottawa, ON: unpublished report.

Chesney-Lind, M. (1997). The female offender: Girls, women, and crime. Thousand Oaks, CA: Sage Publications.

Chesney-Lind, M. (1998). What to do about girls? Promising perspectives and effective strategies. Paper presented at the 6th Annual International Community Corrections Association (ICCA) Research Conference in Arlington, Virginia.

Construction Policy and Services/National Implementation Committee (1992). Regional facilities for federally sentenced women: Operational plan. Ottawa, ON: Correctional Service of Canada.

Corrections and Conditional Release Act (1992). C-20.

Coulson, G., Ilacqua, G., Nutbrown, V., Giulekas, D., & Cudjoe, F. (1996). Predictive utility of the LSI for incarcerated female offenders. Criminal Justice and Behavior, 23(3), 427-439.

Covington, S. (1998a). The relational theory of women's psychological development: Implications for the criminal justice system. In R.T. Zaplin (Ed.), Female offenders: Critical perspectives and effective interventions (pp. 1 13-132). Gaithersburg, MD: Aspen Publishers.

Covington, S. (1998b). Creating gender-specific treatment for substance-abusing women and girls in community correctional settings. Paper presented at the 6th Annual International Community Corrections Association (ICCA) Research Conference in Arlington, Virginia.

Dauvergne-Latimer, M. (1995). Exemplary community programs for federally sentenced women: A literature review. Report submitted to Federally Sentenced Women Program. Ottawa, ON: Correctional Service of Canada.

Dernevik, M. (1998). Preliminary findings on reliability and validity of the Historical-Clinical-Risk Assessment in a forensic psychiatric setting. Psychology, Crime, and the Law, 4, 127-137.

Doherty, J. (1998). Power-belief theory: Female criminality and the dynamics of oppression. In R.T. Zaplin (Ed.), Female offenders: Critical perspectives and effective interventions (pp.1 33-160). Gaithersburg, MD: Aspen Publishers.

Douglas, K. S. (1999). HCR-20 Violence risk assessment scheme: Overview and annotated bibliography. Unpublished manuscript.

Douglas, K. S., Klassen, C., Ross, D., Hart, S. D., Webster, C. D., & Eaves, D. (1998, August). Psychometric properties of HCR-20 Violence Risk Assessment Scheme in insanity acquitees. Poster presented at the Annual meeting of the American Psychological Association, San Francisco, CA.

Douglas, K. S., Ogloff, J. R., & Nicholls, T. L. (1997, June). Assessing the risk for inpatient psychiatric violence. Paper presented at the Annual Convention of the Canadian Psychological Association, Toronto, ON.

Douglas, K. S., Ogloff, J. R., Nicholls, T. L., & Grant, I. (1999). Assessing risk for violence among psychiatric patients: The HCR-20 risk assessment scheme and the Psychopathy Checklist: Screening version. Journal of Consulting and Clinical Psychology, 67(6), 917-930.

Douglas, K. S., & Webster, C. D. (1999). The HCR-20 Violence Risk Assessment Scheme: Concurrent validity in a sample of incarcerated offenders. Criminal Justice and Behavior, 26(1), 3-19.

Dowden, C., & Andrews, D. A. (1999). What works for female offenders: A meta-analytic review. Crime and Delinquency. 45, 438-452.

Dowden, C., & Blanchette, K. (1999). An investigation into the characteristics of substance-abusing women offenders: Risk, need, and post-release outcome. Research Report R-8 1. Ottawa, ON: Correctional Service of Canada.

Dowden, C., & Brown, S. L. The role of substance abuse factors in predicting recidivism: A meta-analysis (manuscript under review).

Federally Sentenced Women Program (1994). Correctional program strategy for federally sentenced women. Ottawa, ON: Correctional Service of Canada.

Finn, A., Trevethan, S., Carrière, G., & Kowalski, M. (1999). Female inmates, Aboriginal inmates, and inmates serving life sentences: A one-day snapshot. Catalogue 85-002-XIE Vol. 19, no. 5. Ottawa, ON: Ministry of Industry.

Funk, S. J. (1999). Risk assessment for juveniles on probation: A focus on gender. Criminal Justice and Behavior, 26(1), 44-68.

Garcia-Coll, C., Miller, J. B., Fields, J. P., & Mathews, B. (1997). The experiences of women in prison: Implications for services and prevention. Women and Therapy, 20(4), 11-28.

Gates, M., Dowden, C., & Brown, S. L. (1998). Case need domain: Community functioning. Forum on Corrections Research, 10(3), 35-37.

Gendreau, P., Goggin, C., & Gray, G. (2000). Case needs review: Employment domain. Research Report R-90. Ottawa, ON: Correctional Service of Canada.

Gendreau, P., Goggin, C., & Smith, M. (1999, May). Is there an actuarial measure that is demonstrably superior to all others? Paper presented at the 60th Annual Convention of the Canadian Psychological Association. Halifax, NS.

Gendreau, P., Little, T., & Goggin, C. (1996). A meta-analysis of the predictors of adult offender recidivism: What works! Criminology, 34(4), 575-607.

Grann, M., Belfrage, H., &Tengström, A. (2000). Actuarial assessment of risk for violence: Predictive validity of the VRAG and the Historical part of the HCR-20. Criminal Justice and Behavior, 27, 97-114.

Gray, T., Mays, G. L., & Stohr, M. K. (1995). Inmate needs and programming in exclusively women's jails. The Prison Journal, 75(2), 186-202.

Grove, W. M., & Meehl, P. E. (1996). Comparative efficiency of informal (subjective, impressionistic) and formal (mechanical, algorithmic) prediction procedures: The clinical-statistical controversy. Psychology, Public Policy, and Law, 2(2), 293-323.

Hann, R., & Harman, W. (1988). Release risk prediction: A test of the Nuffield Scoring System. Ottawa, ON: Ministry of the Solicitor General of Canada.

Hann, R., & Harman, W. (1989). Release risk prediction: A test of the Nuffield Scoring System for native and female inmates. Ottawa, ON: Ministry of the Solicitor General of Canada.

Hannah-Moffat, K. (1995). Feminine fortresses: Women-centered prisons? The Prison Journal, 75(2), 135-164.

Hannah-Moffat, K. (1997). From christian maternalism to risk technologies: Penal powers and women's knowledges in the governance of female prisons. Unpublished Doctoral Thesis, Toronto, ON: Centre of Criminology, University of Toronto.

Hannah-Moffat, K. (1999). Moral agent or actuarial subject: Risk and Canadian women's imprisonment. Theoretical Criminology, 3(1), 71-94.

Hannah-Moffat, K., & Shaw, M. (2000). ‘Taking risks' -- Gender, diversity, security classification and risk/needs assessment with Canadian federally sentenced women (Draft Final Report). Ottawa, ON: Status of Women Canada.

Hare, R. D. (1991). Manual for the Revised Psychopathy Checklist. Toronto, ON: Multi-Health Systems.

Harris, G. T., Rice, M. E., & Quinsey, V. L. (1993). Violent recidivism of men-tally disordered offenders: The development of a statistical prediction instrument. Criminal Justice and Behavior, 20, 315-335.

Heney, J. (1990). Report on self-injurious behaviour in the Kingston Prison for Women. Ottawa, ON: Correctional Service of Canada.

Howden-Windell, J., & Clark, D. (1999). Criminogenic needs of female offenders. HM Prison Service. London, UK: Unpublished report.

Izzo, R. L., & Ross, R. R. (1990). Meta-analysis of rehabilitation programs for juvenile delinquents: A brief report. Criminal Justice and Behavior, 17(1), 134-142.

Jackson, P. G., & Stearns, C. A. (1995). Gender issues in the new generation jail, The Prison Journal, 75(2) 203-221.

Kendall, K. (1998). Evaluation of programs for female offenders. In R.T. Zaplin (Ed.), Female offenders: Critical perspectives and effective interventions (pp. 361-378). Gaithersburg, MD: Aspen Publishers.

Klassen, C. (1996). Predicting agression in psychiatric inpatients using 10 historical risk factors: Validating the “H” of the HCR-20. Unpublished Honour's Thesis, Burnaby, BC: Simon Fraser University.

Koons, B. A., Burrow, J. D., Morash, M., & Bynum, T. (1997). Expert and offender perceptions of program elements linked to successful outcomes for incarcerated women. Crime and Delinquency, 43(4), 5 12-532.

Kroner, D. G., & Mills, J. F. (1997, February). The VRAG: Predicting institutional misconduct in violent offenders. Paper presented at the Annual Convention of the Ontario Psychological Association, Toronto, ON.

Laishes, J. (1997). Correctional Service of Canada Mental Health Strategy for Women Offenders. Ottawa, ON: Correctional Service Canada.

Larivière, M. (1999). The relationship between self-esteem, criminality, aggression, and violence: A meta-analysis. Unpublished comprehensive examination, Ottawa, ON: Carleton University.

Leischied, A. W., Cummings, A., Van Brunschot, M., Cunningham, A., & Saunders, A. (2000). Female adolescent aggression: A review of the literature and the correlates of aggression. Ottawa, ON: Public Works and Government Services Canada.

Lösel, F. (1995). The efficacy of correctional treatment: A review and synthesis of meta-evaluations. In J. McGuire (Ed.), What works: Reducing reoffending (pp. 79-111). Chichester, UK: John Wiley & Sons.

Loucks, A. D. (1995). Criminal behaviour, violent behaviour, and prison maladjustment in federal female offenders. Unpublished Doctoral Thesis. Kingston, ON: Queen's University.

Loucks, A. D., & Zamble, E. (1994). Some comparisons of female and male serious offenders. Forum on Corrections Research, 6(1), 22-24.

Loucks, A. D., & Zamble, E. (1999). Predictors of recidivism in serious female offenders. Corrections Today, 61(1), 26-32.

Loza, W., & Dhaliwal, G. K. (1997). Psychometric evaluation of the Risk Appraisal Guide (RAG): A tool for assessing violent recidivism. Journal of Interpersonal Violence, 12, 779-793.

Loza, W., & Simourd, D. (1994). Psychometric evaluation of the Level of Supervision Inventory (LSI) among male Canadian federal offenders. Criminal Justice and Behavior, 21(4), 468-480.

Mailloux, D. L. (1999). Victimization, coping and psychopathy as plausible explanations for violent behavior among female offenders. Unpublished Master Thesis. Ottawa, ON: Carleton University.

McClellan, D. S., Farabee, D., & Crouch, B. M. (1997). Early victimization, drug use, and criminality: A comparison of male and female prisoners. Criminal Justice and Behavior, 24(4), 45 5-476.

McConnell, B. (1996). The Prediction of female federal offender recidivism with the Level of Supervision Inventory. Unpublished Honours Thesis. Kingston, ON: Queen's University.

Motiuk, L. L. (1997). Classification for correctional programming: The Offender intake assessment (OIA) process. Forum on Corrections Research, 9(1), 18-22.

Motiuk, L. L., & Blanchette, K. (1998). Assessing female offenders: What works. Paper presented at the 6th Annual International Community Corrections Association (ICCA) Research Conference in Arlington, Virginia.

Motiuk, L. L., & Porporino, F. (1989). Offender risk/needs assessment: A study of conditional releases, Report R-01. Ottawa, ON: Correctional Service of Canada.

Nicholls, T. L., Ogloff, J. R., & Douglas, K. S. (1997, June). Comparing risk assessments with female and male psychiatric inpatients: Utility of the HCR-20 and Psychopathy Checklist: Screening version. Paper presented at the 57th Annual Convention of the Canadian Psychological Association, Toronto, ON.

Nuffield, J. (1982). Parole decision-making in Canada: Research to ward decision guidelines. Ottawa, ON: Communications Division.

Oddone Paolucci, E., Violato, C., & Schofield, M. (1998). Case need domain: Marital and family. Forum on Corrections Research, 10(3), 20-23.

Office of the Deputy Commissioner for Women (2000). Correctional program strategy for women offenders. Unpublished draft. Ottawa, ON: Correctional Service of Canada.

Owen, B., & Bloom, B. (1995). Profiling women prisoners: Findings from national surveys and a California sample. The Prison Journal, 75(2), 165-185.

Pendergast, M. L., Wellisch, J., & Falkin, G. P. (1995). Assessment of and services for substance-abusing women offenders in community and correctional settings. The Prison Journal, 75(2), 240-256.

Quinsey, V. L., Harris, G. T., Rice, M. E., & Cormier, C. A. (1998). Violent offenders: Appraising and managing risk. Washington, D.C: American Psychological Association.

Research Branch (2000). The Dynamic Factor Identification and Analysis Instrument (DFIA) and women offenders: What does the research say? Field Consultation Document: Unpublished draft. Ottawa, ON: Correctional Service of Canada.

Rettinger, J. L. (1998). A recidivism follow-up study investigating risk and need within a sample of provincially sentenced women. Unpublished doctoral thesis. Ottawa, ON: Carleton University.

Robinson, D., Porporino, F., & Beal, C. (1998). A Review of the literature on personal/emotional need factors. Research Report R-76. Ottawa, ON: Correctional Service Canada.

Rice, M. E., & Harris, G. T. (1997). Cross validation and extension of the Violence Risk Appraisal Guide for child molesters and rapists. Law and Human Behaviour, 21, 231-241.

Ross, R. R., & Fabiano, E. A. (1985). Correctional afterthoughts: Programs for female offenders. (User Report 1985-18). Ottawa, ON: Ministry of the Solicitor General of Canada.

Salekin, R. T., Rogers, R., & Sewell, K. W. (1997). Construct validity of psychopathy in a female offender sample: A multitrait-multimethod evaluation. Journal of Abnormal Psychology, 106, 576-585.

Salekin, R. T., Rogers, R., Ustad, K. L., & Sewell, K. W. (1998) Psychopathy and recidivism among female inmates. Law and Human Behavior, 22, 109-127.

Schram, P. J. (1998). Stereotypes about vocational programming for female inmates. The Prison Journal, 78(3), 244-270.

Shaw, M. (1991a). Survey of federally sentenced women: Report to the Task Force. (User Report 1991-4). Ottawa, ON: Ministry Secretariat, Solicitor General Canada.

Shaw, M. (1991b). The release study: Survey of federally sentenced women in the community. (User Report 1991-5). Ottawa, ON: Ministry Secretariat, Solicitor General Canada.

Simourd, L., & Andrews, D.A. (1994). Correlates of delinquency: A look at gender differences. Forum on Corrections Research, 6(1), 26-31.

Simourd, D., & Hoge, R. (2000) Psychopathy -- A risk need perspective. Criminal Justice and Behavior, 27(2), 256-272.

Stableforth, N. L. (1999). Effective corrections for women offenders -- What is it and how has research assisted in its development? Forum on Corrections Research, 11(3), 3-5.

Task Force on Federally Sentenced Women (1990). Report of the Task force on federally sentenced women: Creating Choices. Ottawa, ON: Ministry of the Solicitor General.

Tien, G., Lamb, D., Bond, L., Gillstom, B., & Paris, F. (1993). Report on the needs assessment of women at the Burnaby Correctional Centre for Women. Burnaby, BC: BC Institute on Family Violence.

Vachon, M. M. (1994). It's about time: The legal context of policy changes for female offenders. Forum on Corrections Research, 6(1), 3-6.

Walters, G. D., & Elliott, W.N. (1999). Predicting release and disciplinary outcome with the psychological inventory of criminal thinking styles: Female data. Legal and Criminological Psychology, 4, 15-21.

Webster, C. D., Eaves, D., Douglas, K. S., & Wintrup, A. (1995). The HCR-20 Scheme: The Assessment of dangerousness and risk. Burnaby, BC: Simon Fraser University.

Webster, C. D., Douglas, K. S., Eaves, D., & Hart, S. D. (1997). HCR-20: Assessing risk for violence (Version 2). Burnaby, BC: Simon Fraser University.

Women Offender Sector (1998). Community strategy for women on conditional release. Draft discussion paper. Ottawa, ON: Correctional Service of Canada.


Previous PageTop Of Page Table Of ContentsNext Page