Research Reports

The Application of the Community Intervention Scale to Women Offenders:Preliminary Findings

Craig Dowden, Ralph Serin, Kelley Blanchette

Research Branch
Correctional Service of Canada


January, 2001

EXECUTIVE SUMMARY

The appropriate management of offenders on conditional release is one of the keys to their successful community reintegration. Consequently, one of the primary tasks of probation and parole officers is to continuously evaluate and assess offender needs. The Community Intervention Scale (CIS) and its predecessor, the Community Risk/Needs Management Scale (CRNMS) is the risk/needs assessment instrument used by the Correctional Service of Canada (CSC) to allocate resources in terms of frequency of contact for offenders who are under community supervision. Previous research, however, has not demonstrated the utility of the CIS with female offenders.

Therefore, the purpose of the present investigation was to explore the utility of the CIS with a large sample of federally sentenced women offenders (N = 725) who were under some form of conditional release. The sample selected was predominantly a Caucasian, low-risk group of offenders. However, since the primary focus of this report was to explore women offender needs, the majority of the analyses examined this area.

The data demonstrated that the vast majority of offenders experienced problems in each of the need domains with the exception of attitudes. In addition, the proportion of offenders who experienced problems in these areas generally declined the longer they were in the community. Nonetheless, the proportion of women who experienced problems in the attitude and personal/emotional domains remained relatively stable throughout the follow-up period.

The employment domain appeared to be particularly problematic for this sample of women offenders. Notably, the proportion that experienced substance abuse problems was quite small when compared to the other need domains. This was surprising, as many researchers have suggested that substance abuse is one of the primary need areas for women offenders.

The final set of analyses focused on whether offenders who experienced problems in a particular need domain were significantly more likely to recidivate than those who did not. Interestingly, each of the domains of the CIS with the exception of marital/family was significantly related to post-release outcome. Further, four of the domains (associates, community functioning, personal/emotional, and substance abuse) were positively correlated with violent recidivism. Also, offenders whose needs increased in severity while under community supervision were significantly more likely to recidivate than those whose problem areas stayed constant or decreased. This latter finding highlights the important contribution of dynamic and repeated need assessment to the management of offenders on conditional release.

The results from the present investigation provide a relatively comprehensive overview of women offender needs assessment in the community and its relation to recidivism.

TABLE OF CONTENTS

• EXECUTIVE SUMMARY
• LIST OF TABLES
• INTRODUCTION
• METHODOLOGY
  • Sample
  • Procedure
  • Analyses
• RESULTS
  • Demographic information
  • Marital Status
  • Community Intervention Scale Need Areas
  • Changes in Criminogenic Need Across Assessments
  • Need Ratings and Recidivism
  • Maximum Need Change and Recidivism
• DISCUSSION
• DIRECTIONS FOR FUTURE RESEARCH
• CONCLUSION
• REFERENCES

LIST OF TABLES

• Table 1 Demographic Characteristics of the Sample
• Table 2 Distribution of Risk Level for the Sample
• Table 3 Proportion of Women Offenders who Experienced Problems in EachNeed Domain Across Different Time Periods
• Table 4 Rates of Recidivism Across Each of the CIS Need Domains.

INTRODUCTION

The assessment and evaluation of an offender's risk and need level is of paramount importance in the development of an appropriate community reintegration plan. An equally important consideration is to detect changes that take place within these need areas such that appropriate interventions can be ensured to be responsive.

The Community Intervention Scale (CIS) and its predecessor, the Community Risk/Needs Management Scale (CRNMS) is the risk/need assessment instrument used by the Correctional Service of Canada (CSC) to achieve this goal. It provides a comprehensive and systematic assessment methodology to continuously evaluate the changing need levels of offenders who are under community supervision.

The CIS is an empirically validated and theoretically based assessment instrument which, predominantly focuses on the criminogenic need areas of offenders.

Criminogenic needs, by definition, are those dynamic risk factors that, when present, are associated with increased levels of criminal activity (Andrews & Bonta, 1998). Non-criminogenic needs on the other hand are also dynamic, but changes in these need areas are unrelated to recidivism.

The CIS is administered to an offender approximately every six months throughout his/her conditional release period. There are twelve need domains included within the CIS. These are academic/vocational skills, employment pattern, financial management, marital/family relationship, companions/significant others, living arrangements, behavioral/emotional stability, alcohol usage, drug usage, mental ability, health and attitude. However, similar categories within these twelve domains are commonly combined to form seven primary need areas, which include associates, attitudes, community functioning, employment, marital/family, personal/emotional, and substance abuse.

Two separate studies explored the utility of these need domains for managing Canadian federal offenders on conditional release. Motiuk and Porporino (1989) reported that during the field tests of the CRNMS that ten of the twelve need dimensions (with the exception of mental ability and health) were significantly correlated with conditional release suspension over a six-month period. These results were essentially replicated in a larger study (Motiuk, 1996) whereby each of the need domains with the exception of health was significantly correlated with postrelease outcome. These studies provided strong evidence that problems in these need areas were related to post-release outcome. Concomitantly, it would appear that increased knowledge of these factors should enhance our management capabilities of offenders in the community.

Although the previous studies provided confirmatory evidence for these needs in the prediction of offender recidivism, meta-analytic studies have become an increasingly popular statistical methodology to examine the strongest predictors of offender recidivism. A meta-analysis is the statistical aggregation of the results from a group of studies in order to provide an overall conclusion. The advantage of metaanalytic techniques over traditional literature reviews is that the conclusions are reached through quantitative statistical analyses rather than through subjective interpretations. The meta-analytic evidence to date has demonstrated that the majority of the domains that are assessed by the CIS are moderate to strong predictors of criminal recidivism (Brown, 1998).

Despite this consistency, even for men and women offenders, there may be important differences between the needs of each. In particular, factors such as childcare/pregnancy and physical/sexual abuse have been noted (Koons et al., 1997). The importance of attending to the specific needs of women offenders and the necessity of gender-specific programming and assessment have also been emphasized (Blanchette, 1997; Chesney-Lind, 1998; Covington, 1998; Koons et al., 1997). Despite this call for gender-specific treatment and assessment, there is still the need to explore the applicability, to women offenders, of the principles of effective correctional practice (i.e. risk, need, and responsivity) as outlined within the broader correctional treatment literature (Koons et al; 1997).

Some evidence has suggested that there is overlap between "what works" for both female and male offender populations. A study conducted by Coulson, Ilacqua, Nutbrown, Giulekas, and Cudjoe (1996) revealed increased risk scores on the Level of Service Inventory (LSI)¹.

A similar pattern of findings has also been reported in the correctional treatment literature for female offenders. Dowden and Andrews (1999), in conducting the first meta-analysis of the correctional treatment literature for female offenders, reported comparable findings to the broader correctional treatment literature. Their results suggested that the principles of effective correctional treatment were sufficiently robust to impact outcome regardless of the gender of the treatment sample.

However, the authors stressed the importance of gender as a specific responsivity factor, noting that research on gender-specific treatment is virtually absent in the literature.

From a general perspective, this review suggests it may be possible to utilize assessment materials developed for male offenders with women offenders, while being cognizant of potential limitations. The purpose of this study is to investigate the applicability of the CIS to federal women offenders who are under community supervision.

METHODOLOGY

Sample

The sample for this study was extracted on May 1, 1999, from the Correctional Service of Canada Offender Management System, an automated database. All the study participants were federal women offenders who:

• had been admitted to federal custody and received a comprehensive intake assessment to identify risk and needs;
• had been released into the community by the study extraction date;
• had available Canadian Police Information Centre records (which documents official offence history); and
• had been administered the CIS at least once during their conditional release period.

Procedure

The twelve need domains of the CIS were aggregated to form seven need domains consistent with the Dynamic Factors Identification and Analysis (DFIA); the risk/needs assessment instrument administered to all incoming federal inmates.

Similar to its predecessor, the CRNMS, the CIS provides an overall rating for the offender on each target domain based on a four-point continuum. The scale ranges from "asset to community adjustment" (not applicable to the personal/emotional and substance abuse domains) to "considerable need for improvement." The two intermediate ratings are "no need for improvement" and "some need for improvement".

For the present investigation, the rating scale was collapsed and the scores were dichotomized to indicate the presence or absence of a particular need. "Asset to community adjustment" and "no need for improvement" were combined to indicate the absence of a particular need. The domains that were scored as "some" or "considerable" need for improvement indicated the presence of need.

Analyses

The initial analyses focused on the demographic characteristics of the entire sample. The variables examined included age; ethnicity, marital status and overall risk level at intake and provided a reasonably comprehensive overview of women offenders currently under community supervision in Canada.

Subsequent analyses focused on the need areas of this sample of offenders over an extended period of community supervision. Four different time periods were sampled (less than 6 months, 6-12 months, 12-24 months, and more than 24 months). It was hypothesized that as the length of time in the community increased, the proportion of offenders who experienced problems in a particular need area would decline. This trend has been found in previous research investigating changes in offender need during community supervision (Motiuk, 1997).

Finally, each of the individual need domains was correlated with recidivism to explore its predictive utility. In other words, would those offenders who scored higher on these need domains experience higher rates of recidivism than those without such needs?

RESULTS

Demographic information

The final sample consisted of 633 women offenders who were on some form of conditional release and had been assessed at least once on the Community Intervention Scale (or its predecessor the CRNMS). Offenders ranged between 21 to 78 years of age (mean age = 38.0 years). Table 1 presents additional demographic information for our sample of women offenders.

Marital Status

The risk levels assigned at intake for this sample of women offenders are presented in Table 2. Not surprisingly, the vast majority (68%) of the women offenders residing in the community were classified as low risk at intake. This follows the risk principle of case classification (Andrews & Bonta, 1998) which states that lower-risk cases should receive minimum intervention and should be placed on community supervision. This is an informed offender management strategy that has been supported by previous primary and meta-analytic research (Andrews, Dowden, & Gendreau, under review; Lipsey, 1995).

Community Intervention Scale Need Areas

Each offender who is under community supervision is assessed using the Community Intervention Scale approximately once every six-months for the duration of her conditional release period. The purpose of this re-assessment is to monitor any changes in an offender's need level to ensure that an appropriate level of supervision be provided, responsive to the offender's current situation. This dynamic risk assessment approach provides maximum flexibility for properly managing an offender who is under supervision.

Clearly, offenders' needs change during their period of conditional release. For example, one need area for an offender is the area of antisocial associates. It is possible that during her conditional release this could become less of a concern, but that substance abuse concerns arise (or vice versa). If the community interventions that were assigned for this offender were based on the original CIS rating, they would fail to target the most important and current criminogenic needs. Continually tracking changes in need level allows parole officers to intervene with offenders in a timely manner. Thus, such dynamic case management strategies should provide the greatest likelihood for success on release.

Changes in Criminogenic Need Across Assessments

The purpose of the next section is to present the number of offenders who experienced a problem in each need domain during the four time periods outlined previously. The proportions of offenders who experienced a problem in each need domain over time are presented in Table 3. It should be noted that an offender could contribute to more than one time period. For example, offenders who were in the community for 24 months or more would contribute to the analyses for each of the four time periods.

Inspection of Table 3 reveals that women offenders experienced problems in many different need domains during their period of community supervision (with the exception of Attitudes). In addition, in most cases, the proportion of offenders who experienced problems in these areas diminished over time. This makes intuitive sense, as those with increased needs are more likely to be revoked or otherwise returned to prison.

Women offenders initially experienced the most problems in the employment domain at the six-month review period. Blanchette and Dowden (1998) reported a similar finding in their study of a smaller sample of federally sentenced women offenders in the community (n = 297). In the present study, the employment domain demonstrated a clear and steady decline over time, with only 53% of offenders having a perceived need in this area after twenty-four months in the community, a reduction of 20%.

The majority (62.3%) of women who were supervised in the community during the study period experienced problems in the personal/emotional domain. More importantly, the proportion of offenders who experienced problems in this domain did not decline appreciably across the different assessment periods. Clearly, these problems did not ameliorate with time in the community.

One additional finding must also be highlighted. It has been argued that one of the primary criminogenic need areas for women offenders is in the area of substance abuse (Bloom & Covington, 1998; Dixon, 1998; Koons et al., 1997; Peugh & Belenko, 1999; Prendergast, Wellisch, & Wong, 1996). Surprisingly, however, relatively few women in this large sample were determined to have a problem in this area. This is particularly the case, especially when you focus in comparison to other need domains and for those women who are in the community for two years or more.

Clearly, federally sentenced women offenders experience problems in several different need areas while under community supervision, although these diminish over time for those who remain in the community for an extended period of time.

Accordingly, analyses were conducted to explore whether women offenders who experienced problems in each need domain were more likely to reoffend.

Need Ratings and Recidivism

This section describes the utility of the CIS need assessments for predicting postrelease outcome. The relationship between the final need rating received by each offender and post-release outcome was examined through chi square analyses. The final need rating was operationally defined as either the rating received prior to revocation (if the offender was revoked) or the most recent one received before the end date of the study. We believed that this definition provided the most accurate assessment as it reflected the most proximal index of an offender's situation.

Interestingly, the results revealed that having a problem in each of the need domains (with the exception of marital/family) was associated with significantly higher rates of return to custody for any reason. Further, offenders who experienced problems in the associates, community functioning, personal/emotional and substance abuse domains were significantly more likely to be returned to custody for a violent offence during their conditional release period than those who did not experience such problems.

*p<.05; **p<.01; ***p<.001

Maximum Need Change and Recidivism

Although correlating the most recent need rating with recidivism provided some interesting information regarding offender risk management in the community, this approach did not take full advantage of the available data. More specifically, the dynamic nature of the data enabled us to explore whether changes in these need areas were related to reoffending. Accordingly, additional analyses were conducted to examine whether the maximum change score received by an offender could significantly predict post-release outcome.

The first step involved assigning a numerical value to each of the need level ratings received by each of the offenders. The following coding values were assigned for each need level:

0 = Asset to community adjustment
1 = No need for improvement
2 = Some need for improvement
3 = Considerable need for improvement

For each offender included in the sample, a change score was calculated by examining the difference in need scores received between two consecutive time periods for each need domain (i.e. time 2 versus time 1, time 3 versus time 2, etc.). Consequently, positive need change scores identified an increased problem in the need domain while a zero or negative score indicated that the need area had not changed or had improved for the offender. Once these change scores were calculated, the maximum need change score was identified. The maximum need change was operationally defined as the largest change score that reflected an increased problem for an offender within a particular need domain (i.e. the largest positive change score).

The maximum need change scores were then correlated with recidivism. This determines whether offenders who experienced increased problems during their supervision period were more likely to recidivate.

The distribution of the maximum change in need scores for the offenders in this sample ranged from "0" to "2". These scores were classified as "no change", "moderate change" and "severe change". Not surprisingly, the maximum change score received had a significant impact on post-release outcome (c²(2) = 10.01, p<.01) with increased change scores being associated with increased levels of recidivism. More specifically, individuals whose need scores did not change experienced the lowest rate of recidivism (5.1%) followed by those with moderate change (11.2%) and those with severe change (19.7%). These differences were statistically significant (c²= 9.25, p<.005). This same trend was maintained when violent reoffending was the outcome measure of interest.

DISCUSSION

The present study explored the need characteristics of federally sentenced women offenders on conditional release. That the majority of offenders under community supervision were low-risk cases suggests the Service is following effective correctional practice. This follows the theoretically based and empirically-validated risk principle of case classification which suggests that lower-risk cases should be assigned the lowest levels of service and supervision (Andrews & Bonta, 1998).

Since the primary purpose of this study was to explore changes in offender needs while they are in the community, several analyses were conducted in this area. The results revealed that women offenders experienced problems in a number of need areas during their conditional release and that these generally dissipated with time in the community. As stated previously, this finding was fairly intuitive and previous research (Motiuk, 1997) has also commented on this trend.

The need domains of the CIS also have important implications in terms of offender risk. The vast majority of the need areas assessed by the CIS were also related to post-release outcome and two of them (associates, personal/ emotional, and substance abuse) were related to violent recidivism. This suggests that knowledge of offender needs is critical for the effective management of offenders in the community.

Another interesting finding was that the proportion of women offenders who experienced substance abuse problems was quite small when compared to the other need domains. This finding, however, replicates research conducted by Blanchette and Dowden (1998) who reported that federally sentenced women offenders under community supervision did not experience as many problems with substance abuse as they did in other need areas such as the employment or personal/emotional domains.

The strong link between offender needs and post-release outcome highlights an extremely important offender management issue regarding the appropriate allocation of correctional resources in the community. Specifically, what strategy should be used to allocate available resources? An example may more effectively illustrate this point. The present research indicated that although a large proportion of women offenders experienced marital/family problems while under community supervision, these did not place them at an increased risk to reoffend. In contrast, very few women had problems in the attitude domain, yet these problems were significantly related to post-release failure. The dilemma, then, is whether resources should be allocated according to need or risk of recidivism? Presumably, providing correctional interventions for the relatively small number of offenders experiencing difficulty in the attitude domain should be more cost-effective. It may be that a hierarchical model of resource allocation is preferred whereby both need and risk are considered. Clearly, additional research is required; however, these are important considerations in terms of policy development.

Finally, the consideration of maximum change scores provided strong utility in the prediction of offender recidivism. Accordingly, this study demonstrates the value of dynamic need assessment given that increases in offender needs were linked with future criminal activity. These types of analyses have been recommended in order to augment the quality of the correctional treatment literature (Andrews, Zinger, Hoge, Bonta, Gendreau, & Cullen, 1990). Only when changes in need areas directly relate to post-release outcome will we move to the next generation in risk appraisals.

DIRECTIONS FOR FUTURE RESEARCH

This study has provided a preliminary overview of the characteristics of federally sentenced women offenders in the community and how standardized appropriate need assessment is related to post-release outcome. Although the current study preliminarily considered changes in need, there are other research questions that need to be addressed in the future.

Subsequent research must make a clear differentiation between the magnitude of the changes in the need levels. For example, an offender who goes from a rating of "no need for improvement" to "considerable need for improvement" should be scored differently than an offender who goes from "some need" to "considerable need." This difference in severity could be very important and may provide additional insight into the relationship between offender needs and recidivism.

An equally important point is that there should be a differentiation made between offenders who change from not having a problem in a particular area (i.e., asset to community adjustment or no need for improvement) and those who experience increased intensity of an existing problem. Clearer links must be made between dynamic need changes and recidivism, based on empirical findings.

Finally, the pattern of changes in offender need must also be longitudinally mapped. Certainly, changes in dynamic factors will likely not take place in a predetermined pattern. Also, it is possible that the severity and expression of certain offender needs varies over time. For example, the present study suggests that while the majority of offender needs decrease over time, problems in the personal/emotional domain do not. An improved understanding of these issues may facilitate community management of offenders. From the present research, offenders who have need ratings at all four time periods could be analyzed separately such that their patterns may be examined more carefully.

CONCLUSION

The present study provided a comprehensive profile of federally sentenced women in the community. The preliminary evidence suggests that the CIS is a valid risk/needs assessment instrument for women offenders and that dynamic needs assessment is a critical case management consideration. It is hoped that future research will make further contributions to dynamic needs assessment, recidivism prediction and effective case management.

REFERENCES

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

Andrews, D.A. & Bonta, J. (1998). The Psychology of Criminal Conduct. Second Edition. Cincinnati, OH: Anderson.

Andrews, D.A., Dowden, C., & Gendreau, P. (under review). Clinically relevant and psychologically informed approaches to reduced re-offending: A metaanalytic 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 psychologically informed meta-analysis. Criminology, 28, 419-429.

Blanchette, K. (1997). Classifying female offenders for correctional interventions. Forum on Corrections Research, 9 (1), 36-41.

Blanchette, K., & Dowden, C. (1998). A profile of federally sentenced women in the community: Addressing needs for successful reintegration. Forum on Corrections Research, 10 (1), 40-43.

Bloom, B., & Covington, S. (1998, November). Gender-specific programming for female offenders: What is it and why is it important? Paper presented at the 50^th Annual Meeting of the American Society of Criminology, Washington, DC.

Brown, S.L. (1998). Dynamic factors and recidivism: What have we learned from the Case Needs Review Project? Forum on Corrections Research, 10 (3), 46-51.

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, 427-439.

Dixon, J. (1998). Relationship of substance abuse by women offenders and recidivism. Unpublished Honours thesis, Department of Psychology, Saint Mary's University, Halifax, Nova Scotia.

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

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

Koons, B.A., Burrow, J.D., Morash, M., & Bynum, T. (1997). Expert and offender perceptions of program elements linked to successful outcomes to incarcerated women. Crime and Delinquency, 43, 512-532.

Lipsey, M.W. (1995). What do we learn from 400 research studies on the effectiveness of treatment with juvenile delinquents? In J. McGuire (Ed.), What works: Reducing reoffending - Guidelines from research and practice. Chichester, England, Wiley & Sons.

Motiuk, L.L. (1996, November). Assessment methods in corrections. Paper presented at the 4^th Annual International Community Corrections Association Research Conference, Austin, TX.

Motiuk, L.L. (1997). The Community Risk/Needs Management Scale: An effective supervision tool. Forum on Corrections Research, 9 (1), 8-12.

Motiuk, L.L., & Porporino, F.J. (1989). Field tests of the Community Risks/Needs Management Scale: A study of offenders on caseload. Research Report No. R-06, Ottawa: Correctional Service of Canada.

Peugh, J., & Pelenko, S. (1999). Substance-involved women inmates: Challenges to providing effective treatment. The Prison Journal, 79, 23-44. Prendergast, M.L., Wellisch, J., & Wong, M.M. (1996). Residential treatment for women parolees following prison-based drug treatment: Treatment experiences, needs and services, outcomes.