Correctional Service Canada
Symbol of the Government of Canada

Common menu bar links

Research Reports

Warning This Web page has been archived on the Web.

Persistently Violent (Non-sexual) Offenders: A Program Proposal

N° R-42
Research Division
Correctional Research and Development
Correctional Service of Canada

November 1995

Table of Contents


Executive Summary


Conceptual Model


Consultation Group

Program Design:

Research Framework



Appendix A: Preliminary Assessment Battery

Appendix B: Responsivity Factors

Appendix C: Admission Criteria

Program Rules


     The consultants and their area of expertise are as follows: i) Dr. Ronald Blackburn, Director of Research, Ashworth Hospital, Liverpool, England - psychopathy, assessment and treatment in correctional and forensic settings; ii) Dr. Joseph Newman, Department of Psychology, University of Wisconsin, Madison, Wisconsin - psychopathy, impulsivity; iii) Dr. Raymond Knight, Department of Psychology, Brandeis University, Massachusetts - typology derivations, assessment of sexual offenders, statistics; Dr. Howard Barbaree, Head, Forensic Division, Clarke Institute of Psychiatry, Toronto - assessment and treatment of sexual offenders, treatment responsivity, and statistics. Additionally, Dr. Sharon Kennedy and Mr. Michael Bettman, CSC psychologists with expertise in the area of assessment and treatment of violent offenders, participated in the meeting and are members of the advisory council. At the conclusion of the experts' meeting, there was consensus to establish a collaborative research agenda over the course of the project, which should ensure a very high quality final product.

Executive Summary

     This proposal outlines relevant treatment targets and presents a research framework to evaluate a program for persistently violent (nonsexual) offenders. The proposed program is compatible with existing Correctional Service of Canada programs, but is more intensive than most, reflects an alternate conceptual framework which has been employed in other settings, and has a rigorous assessment component. The proposal recommends providing intervention over a 3 year period at multiple sites, with a matched treatment control (anger management) and untreated groups to permit the evaluation of program effectiveness on institutional violence and post-release recidivism.

     The program incorporates novel and multiple measures of interpersonal style (personality), impulsivity and aggressive beliefs. Responsivity or treatability factors relate to an offender's ability and motivation to respond to both the program content and intervention model. These factors are included in the assessment protocol to investigate which factors may potentially influence an offender's response to treatment, an increasingly important consideration for all programs. The results of these investigations will therefore be applicable to a range of offender programming, including anger control, the treatment of sexual offenders, and core programming.

     Comparisons of types of violent offenders and treatment outcome will be considered.

     The program, an intensive intervention, is intended to be more than an educational exercise. It is expected that offenders must demonstrate skill acquisition and application of these skills in order for successful participation in the proposed program. Targets reflected in the treatment components include: aggressive beliefs; anger and arousal; cognitive distortions regarding violence, impulsivity, and self-regulation. Violent offenders are a diverse group who will vary with respect to their needs in these areas.

     The specific research questions and methodology, a preliminary assessment battery, overview of treatment components, and consultation process are described in the proposal.


     The identification and effective treatment of persistently violent offenders is of paramount concern to the Correctional Service of Canada (CSC) for the protection of Canadian society. This mandate is reflected in Corporate Objectives 1 and 2 which relate to the assessment, treatment and management of violent offenders, and reduction of violence in institutions, respectively. Recent training in the Service incorporating contemporary risk assessment knowledge has been completed with a view towards improving our ability to recognize and respond to offender risk. That such assessment includes both the identification of risk situations and the appropriate management of offenders' level of risk indicates that intervention is an important part of the risk management continuum. Several initiatives have culminated in providing a body of literature which should guide programming for persistently violent (non-sexual) offenders : i) a review of strategies for managing sexual and high- risk offenders (Motiuk, Belcourt, & Bonta, 1995; National Sex Offender Strategy, CSC, 1995); ii) literature reviews on the assessment and treatment of violent offenders (Blackburn, 1993; Rice, Harris, Quinsey, & Lang, in press; Serin, 1994); iii) issues affecting treatment responsivity in offenders (Forum on Corrections Research, 1995); and, iv) several recent summaries of effective correctional programming, including meta-analytic procedures, are now available (e.g., Andrews, Zinger, Hoge, Bonta, Gendreau & Cullen, 1990; Gendreau, in press; Gendreau, Little & Coggin, 1995; Lipsey & Wilson, 1993).

     For the purposes of this initiative, persistently violent offenders are defined by any of: a history of seriously violent offenses either as a juvenile or adult; a record of institutional violence, (e.g., threats or assaults against staff or offenders); frequent use of a weapon during crimes; and/or, evidence of explosive behaviour, not attributable to psychosis or overt mental disorder. The intent is to identify a group of offenders who engage in repetitive violence and choose their victims relatively indiscriminately. This high-risk group of offenders would then be provided alternate treatment opportunities for comparison. Specifically, the objective is to offer an intensive program for this group of serious offenders. Existing intake assessment protocols, criminal history records, police reports, and court records will be used to identify potential candidates for programming.

     A variety of correctional programs have been developed in an attempt to address the needs of violent offenders in general. These include: personality disorder programs; sexual offender treatment; cognitive living skills; anger and emotions control; substance abuse; and domestic violence programs. Further, anger control programs have been provided to offenders for the past decade, albeit with few published findings. As well, a survey is currently underway to create an inventory of programs for violent offenders, including suggested "best practices". It should be noted that these programs differ in their focus, intensity and direction, and at this time, their efficacy remains largely speculative.

     For maximum efficacy in treating persistently violent (non-sexual) offenders we must distill from the assessment and treatment literature a promising strategy that addresses their unique treatment needs. Such a strategy should consider theoretical issues, reflect contemporary assessment technology, address compliance and responsivity factors, and maintain methodological rigor. The development of a theoretical model would provide an overall framework for intervention and permit the testing of specific hypotheses about intervention and its relation to outcome. Likewise, the incorporation of contemporary assessment advances is critical for measuring appropriate treatment targets and determining the extent of treatment gain. Although offender risk and need assessment knowledge has made considerable gains over the past decade, measurement developments have remained, with only a few exceptions, limited. Finally, the inclusion of responsivity factors, (e.g., age, ethnicity, motivation) is important because high risk offenders have often proven to be resistant to treatment. Even limited treatment progress for persistently violent (non-sexual) offenders may have significantly greater impact on recidivism rates than modest treatment gains for lower risk offenders.

     The development of a specific program for these offenders is critical because simple participation in a treatment program has proven insufficient for making substantive changes on a variety of attitudinal and behavioral measures that sustain violence. Fortunately, the literature on correctional treatment identifies the characteristics of good program deliverers that should be incorporated into the treatment of highly resistant offenders (Gendreau, in press; Miller & Rollnick, 1991). Methodological sophistication will, however, be required to investigate the parameters of treatment effectiveness. Optimally, determining which offenders respond best to which components of a particular intervention would permit better matching of offenders to treatment resources. Although a pilot initiative, it is imperative to reflect the highest level of program integrity from the current literature. Equally important is the development of a comprehensive research agenda to parallel the delivery of services. Such a research program would yield data that should then inform subsequent decisions regarding such issues as the modification of assessment strategies, treatment content, and selection criteria.

     The purpose, then, of this project would be the following: i) to develop and evaluate the efficacy of a conceptual model program for persistently violent (non-sexual) offenders; ii) to investigate the relationship between dynamic treatment variables and recidivism; and, iii) to advance assessment strategies for persistently violent (nonsexual) offenders for incorporation into existing anger control programs, providing a more standardized assessment for the allocation of treatment resources. The treatment program, by design and intensity, is intended to be more than a psycho- educational exercise. Not only should offenders demonstrate skills acquisition post-treatment, but there should be a mechanism for measuring skills application.

Conceptual Model

     Violence is a complex phenomenon with multiple determinants. However, there is increasing evidence that persistently violent (nonsexual) offenders share characteristics which predispose them to resolve problems violently or to use violence to gain a particular outcome. Several models (Novaco, 1994; Serin & Kuriychuk, 1994; Zillman, 1988) highlight the importance of arousal and cognitive style in determining whether an individual will respond violently. These models identify, from a theoretical perspective, treatment targets that must be assessed and modified, if a treatment program is to be effective. These treatment targets include: aggressive beliefs, anger and arousal, cognitive distortions regarding violence, impulsivity, and self-regulation deficits. Further, it is anticipated that persistently violent (non-sexual) offenders will vary regarding the relative importance of each of these concepts. That is, violent offenders are heterogeneous and diverse, requiring prescriptive intervention.

     Briefly, these models emphasize the central role of aggressive beliefs or cognitive schema in offenders' vigilance towards antecedent events and their regulation of arousal and impulsivity. Anger is an important antecedent to violence (Novaco, 1994), but, not all violence is precipitated by increased arousal or anger. That is, some offenders' violence is instrumental, or used to gain a particular outcome. Cognitive style appears to be relevant for the identification of offenders whose egocentricity and lack of concern for others is central to their interactions, including instrumental violence. The literature on schema (Novaco & Welsh, 1989; Serin & Kuriychuk, 1994) and social-information processing deficits (Dodge, 1986) suggests that persistently violent individuals demonstrate cognitive distortions regarding interpersonal conflict, both real or imagined. These schema result in their belief that others act with malevolence and lead to preemptive aggression and egocentric views, (i.e., righteous use of violence). Efforts to measure these deficits include hypothetical vignettes or stories (Dodge, 1986), interpersonal rating systems (Agee, 1990), and self-report questionnaires (Bettman, 1995; Slaby & Guerra, 1988; Serin & Kuriychuk, 1993).

     Impulsivity is also considered to be an important determinant of violence in offenders in that deficits relate to the control or self-regulation of behaviour, i.e., poor behavioral inhibition (Barratt, 1994). Impulsivity has been viewed as either an inability to be reflective or the interval between a particular event and the individual's response. Aggression has been linked to impulsiveness across various domains: motor (acting without thinking); cognitive (making quick decisions); and non-planning (unconcern for consequences). Management of impulsivity or improved self-regulation, then, incorporates recognition of increased arousal and the use of competing strategies, improved problem-solving, and clearer consideration of consequences. Self-regulation, then, involves more than simply increasing the time until an individual responds, it implies withholding certain responses. Further, impulsivity is in part a function of the method of measurement used, and is considered two dimensional, either cognitive or behavioral (Barratt, 1994; White, Moffit, Caspi, Bartusch, Needles, & Stouthamer Loeber, 1994). Newman's investigation of self-regulation in offenders has led to the identification of a group of offenders whose disinhibition occurs most often when they are involved in goal-directed behaviors, or impaired because of high emotional states or substance abuse (Newman, 1990; Newman & Wallace, 1993a). This research on self-regulation suggests that offenders may vary regarding the best strategy for them to control their impulsive behaviour (Newman & Wallace, 1993a, 1993b). The issues of schema, arousal, and disinhibition need to be incorporated into a theoretical model for violent offenders and guide intervention (Novaco, 1994; Serin & Kuriychuk, 1994).


     An important consideration in the assessment and treatment of violent offenders is their heterogeneity. Offenders differ in terms of demographic factors, but also developmental history, criminal history, criminogenic needs, and skills levels. For violent offenders, additional issues such as degree of planning, anger (undercontrol), hostility (overcontrol), victim injury, weapon interest, developmental patterns of violence, motivational cues, and victim affiliation must also be considered. Offenders with the same index offense could therefore have different treatment needs, although some overlap would occur. Given this heterogeneity and the diversity of proximal cues or high risk situations, adopting a relapse prevention framework may be helpful in individualizing assessment and treatment efforts. This approach has been successfully applied with offender populations in the areas of substance abuse (Annis, 1986) and sexual offending (Marques, Day, Nelson, & West, 1994; Pithers, 1990). Such an approach is also consistent with the principles of effective risk management. Identifying high risk situations as treatment targets permits review and monitoring of these situations when the offender is under supervision in the community. Assessment of offenders' understanding of high risk situations (institutional or community), relapse prevention principles, and competence in responding to these situations can be adapted from materials developed for sex offenders (California Department of Mental Health, Sex Offender Treatment and Evaluation Project, 1988). For this proposal, relapse prevention principles will be reflected in the program content and reinforced through weekly individual sessions and homework logs/assignments.

     Increasingly, cognitive behavioral approaches are dominating intervention efforts with offenders (Andrews & Bonta, 1994; Blackburn, 1993). Moreover, less structured approaches have not proved effective with offender populations (Gendreau, in press). Best practices for good correctional programming have been described by Andrews and his colleagues (Andrews & Bonta, 1994) and have been supported by recent meta-analytic reviews of the treatment literature (Andrews, Zinger, Hoge, Bonta, Gendreau, & Cullen, 1990; Gendreau, Little, & Goggin, 1995). These components have also been endorsed by advocates of treatment for mentally disordered offenders (Rice & Harris, 1993; Rice, Harris, Quinsey, Harris, & Lang, in press).

     A review of programs designed to address the treatment needs of violent offenders yields reasonable consensus regarding the areas that should be targeted in a treatment program (Blackburn, 1993; Rice, Harris, Quinsey, & Cyr, 1990; Serin, 1994). Perhaps, because of offender heterogeneity and differing theoretical perspectives guiding offender programming, there is little consensus regarding the relative importance of these content areas and the preferred order of presentation in a treatment program. Clinically, it would seem desirable to provide components of programs that are incremental, such that knowledge and mastery of one skill is prerequisite to the next step in the treatment process. Treatment should also be prescriptive, (i.e., related to an offender's specific treatment needs). More specifically, treatment should only be provided where there is evidence that the offender has particular deficits or needs. Pre-treatment assessment of program requirements, then, is critical to ensure the appropriate allocation of resources.

Assessment of Treatment Targets

     The psychological constructs highlighted earlier from conceptual models can be directly translated into treatment targets. These include: anger (definition, anger cues, anger triggers, relaxation training, and cognition's); cognitive processing (aggressive beliefs, expectancies and appraisals, attribution biases); impulse control; problem-solving; substance abuse; social and communication skills; assertiveness training; empathy (victim impact); moral reasoning, and criminal attitudes and associations. These latter two appear to be important components in the maintenance of violent criminal behaviour. Peer support for antisocial and violent behaviour must be challenged and alternate support systems developed as part of the follow-up process to maintain those gains derived from the treatment program (Andrews & Bonta, 1994; Andrews, Young, Wormith, & Searle, 1973). These content areas or treatment targets then, identify specific skills deficits which must be addressed in the treatment program. It is important that offenders demonstrate both skills acquisition, (i.e., new learning) and skills application, (i.e., generalization and use of these skills in different contexts).

     Multiple offender-specific assessment strategies for each of these targets need to be developed. In some cases measures exist, but reliance on self-report measures of treatment needs and gains have been a notable weakness (Serin, 1994). A preliminary list of measures for each treatment target is presented in Appendix A. This is an exhaustive battery, but computerization of many of the self-report scales will decrease the time for testing, enhance compliance, and facilitate data entry. Such inclusiveness is considered important to meet the research function of this treatment program proposal. In addition, assessment strategies should include behavioral measures, detailed histories, consideration of interpersonal style in various settings and by different raters within the institution, synthesis of criminal history variables, and self-report. Peer ratings might also be helpful, but may be difficult to operationalize. Computerized assessment, pencil and paper tests, clinical interviews, and role plays should all be incorporated into a broad assessment battery. Optimally, there should be convergence among different approaches of measurement of the same construct.

     There is less agreement regarding the intensity of treatment required in a particular case, but it is inferred that higher-risk, higher-need cases warrant more intensive intervention. The specific inference is that such cases are both more resistant to change and have poorer baseline skills. There may, however, be a ceiling or limit regarding the intensity of intervention and anticipated gains (Serin, 1995). Treatment which is more intensive or of longer duration should not be viewed as a panacea, rather an important management strategy for persistently violent offenders.

     While some argue that sexual and non-sexual offenders have similar treatment needs (Quinsey, Rice, Harris, & Lalumière, 1993), important differences remain relevant to assessment and treatment (Motiuk, Belcourt, & Bonta, 1995; National Sex Offender Strategy, CSC, 1995; Prentky, 1995). Nonetheless, a review of treatment targets for sex offender programs reveals considerable overlap with targets for non-sexual offenders (see National Sex Offender Strategy, CSC, 1995). Further, it has been suggested that those variables which predict recidivism among non-sexual offenders may apply also to sexual offenders (Quinsey, Rice, & Harris, 1995). This implies that gains in the assessment and treatment of persistently violent (nonsexual) offenders could have application to the treatment and management of sexual offenders.

Assessment of Treatment Responsivity

     Treatability has been described as an elusive construct, yet it is central to increasing our understanding about treatment efficacy. While efforts to operationalize what is meant by the term treatability have not been particularly successful (Heilbrun, Bennett, Evans, Offult, Reiff, & White 1992), the relevant parameters have been provided. Some of these, (e.g., response to prior treatment, and motivation for treatment) have been incorporated into a self-report scale (Baxter, Marion, & Goguen, 1995). Additionally, related issues have been included in measures of offender participation in treatment (Ogloff, Wong, & Greenwood, 1990). The appropriateness of the particular treatment has also been considered (Heilbrun, et al., 1992).

     Marques and colleagues have also recognized the importance of treatment gain and have developed behavioral rating scales to reflect increased competence which correlate with post-treatment performance (Marques, et al., 1994). Finally, case workers' sequential ratings of offender's needs are related to release performance (Motiuk & Brown, 1994), suggesting that improvements in the measurement of dynamic factors will improve post-treatment predictions. Appendix B presents responsivity factors, although specific measures will need to be developed for this project. It is important, however, that treatment gain not be confounded by assessment of treatability, such that only those who respond to treatment are considered treatable. Rather, treatment responsivity could be potentially enhanced by pre-treatment orientation sessions. Another way of conceptualizing this dilemma is to view offenders as differing in their level of readiness for treatment (Prochaska, DiClemente, & Norcross, 1992). This could then be controlled for statistically in data analysis.

     Similar to the assessment of treatment targets, responsivity factors need to be measured using various strategies and at multiple times during the treatment program. Process measures may be instructive dynamic factors which help identify those offenders who best respond to particular elements of intervention.

Assessment of Treatment Gain

     The investigation of treatment gain begins with a clear description of what the program is intended to impart, e.g., specific knowledge, skills, and attitudes, which were presumably absent pretreatment. Each of these gains should be assessed for each treatment target (see above), and post-treatment improvement is a measure of treatment gain. Since offenders vary with respect to their initial level, treatment gain must be reflected in relative rather than absolute terms (higher skilled offenders will be less able than lower skilled offenders to show comparable treatment gain). In addition, an absolute measure is also desirable to determine if there is a threshold for competence and maintenance of a particular skill.

     Another measure of treatment gain for violent offenders is reduced institutional violence. Again, relative to baseline (pre-treatment) levels, it is important to determine whether treated offenders decrease the frequency and severity of institutional aggression, both physical and verbal. Institutional records could be augmented by behavioral observations completed by security staff and shop instructors, with appropriate training (although it is recognized this is a time consuming endeavour with difficulty obtaining reliable ratings).

     These in-program measures of treatment gain are considered distinct from the generalization to longer-term effects such as recidivism. Regarding release, several measures of treatment gain are possible. First, the impact of program participation can be assessed by considering the rate of transfers to reduced security, granting of release opportunities (ETAs, UTAs, day parole, full parole), or the withholding of detention. Having an untreated control group or alternate treatment control group would provide valid rates for comparison. Once released, assessment of treatment gain is considered by failure rates. Community performance measures could also include compliance with special conditions and treatment follow-up, type of failure, and time until failure. Some success has also been found using ratings of community adjustment (Motiuk & Brown, 1993). A protocol for the data coding and the sources of data are required, but should reflect existing procedures in the Correctional Research and Development section.

Consultation Group

     In order to ensure the proposal reflects a contemporary contribution to the assessment, treatment, and management of high-risk violent offenders, a consultation process was developed. Accordingly, an earlier version of the proposal was reviewed to ensure the final product be a very high caliber proposal meeting high external and academic standards, as well as operational realities. The methodology proposed was enhanced by the consultation process, but the responsibility for the final product rests with the Project Manager, not the consultants. The consultants have varied backgrounds, permitting a complementary endeavour. Each consultant reviewed the proposal and provided written comments, which were incorporated in this version of the proposal. Additionally, an experts group met to discuss in detail each aspect of the treatment program, the methodology, and research plan.

Program Design

     One of the major obstacles to the evaluation of program effectiveness has been weak methodology (Quinsey, et al., 1993) and this is particularly notable in the area of violent offenders (Serin, 1994). Fortunately, the literature on the treatment of sexual offenders, which is conveniently analogous to the treatment of violent offenders, identifies critical issues to consider (Marques, et al., 1994; Quinsey, et al., 1993). As a pilot project intended to inform subsequent programming for persistently violent offenders, it is vital that the methodology meet contemporary standards. At the same time, meeting such requirements is resource intensive and requires careful planning.

     After much discussion, it appeared there were two clear options regarding the methodology for the design and intent of the proposed program. One consideration was to a priori identify "types" of violent offenders and match them to a specific type of intervention. The types would be proposed by personality theory, or the Big 5 personality types and some prior work on typologies of violent offenders. These types are broadly reflected in the treatment targets described previously. This strategy would maximize investigations of heterogeneity and prescriptive intervention. A concern was that there is no empirical evidence to support this approach and potentially some offenders would be poorly matched to a particular treatment. A second strategy was to conceptualize an alternate treatment regime (Cognitive Mediation Program) and contrast it with the current programming (Living Skills - Anger and Emotions Management Program) at the same site. Offenders must be randomly assigned to either of the programs, which would be of similar length. Offenders in each program would be matched according to risk level (SIR score) and age. Questions relating to heterogeneity and responsivity to treatment would have to be considered after the fact, yet this strategy was superior in that it ensured a program of proven clinical integrity for both the treatment and treatment control groups. This would mean that the pilot aspect of the program would not interfere with operational requirements. Importantly, subsequent investigations of program efficacy, offender compliance, and offender heterogeneity would also be possible. Further, alternate control (untreated) subjects and cohort comparisons could still be considered. The research design is presented in Figure 1.

     The problem of offender noncompliance or treatment drop-out has practical and methodological implications High rates of noncompliance are one indication that a program lacks effectiveness, at least for those who fail to remain in treatment. Different rates of noncompliance for the treatment and alternate treatment control programs creates unique methodological problems. That is, failure to consider this issue risks suggestions that treatment only works for cases who really didn't need it. Options to address this potential concern include: i) Debrief treatment drop-outs to determine rates for each group and those factors contributing to noncompliance; ii) Complete analyses twice, once with treatment drop-outs included and once with them excluded; iii) Compare treatment drop-outs with treatment completers on relevant variables; iv) Compare failure rates for similar offenders (e.g., risk levels, age, index offense) in control and treatment groups, for treatment drop-outs and completers.

Selection of Candidates

     Admission criteria are presented in Appendix C. Being within two years of release will facilitate recidivism follow-up (unless a participating offender is subsequently detained). The program manual has yet to be developed, but will reflect work by Slaby and Guerra (1990) which is a problem-solving approach. Data collection will be a multiwave, longitudinal design and careful documentation of compliance rates and treatment dropouts will be completed. Assessments will occur pre-treatment, during treatment (multiple times for process measures), and post-treatment. The untreated, treatment control (Anger and Emotions Management Program) and treatment (Cognitive Mediation Program) groups would complete the full assessment pre and post-treatment. In addition to a standardized battery for these phases, specific measures for program content will be considered for each assessment wave. For instance, a detailed social and criminal history and clinical evaluation will only occur at pre- treatment. Additionally, knowledge and performance measures of competency, (i.e., skills acquisition and application) for each treatment target will be developed to ensure offenders understand program content and can apply it to their situation. Abbreviated assessments will occur at 6 month follow-up intervals, including in the community for those offenders released.

Site Selection

     Prior to commencement, a program outline will be required reflecting the underlying philosophy. Sites will have to be selected and treatment staff recruited. Although it is required that the staff have considerable clinical expertise and correctional experience, additional training will be required to ensure the program is delivered in a standardized manner. While the treatment targets have been identified and a cognitive-behavioral framework proposed, the clinical staff should be co-developers of the treatment manual (several existing manuals are available as a starting point). This will incorporate their collective experience and provide a sense of ownership in the project. The "best practices" yielded from the to-be- completed inventory will also be considered.

     Regarding the setting characteristics, a number of researchers recommend that a treatment program be within a specialized area in an institution to maximize therapeutic interaction, to ensure consistent monitoring by staff, and to limit the negative peer interactions with other offenders not in the treatment program (Cooke, 1989; Ogloff, Wong, & Greenwood, 1990). Such a requirement has operational implications, yet is preferred. As this program is intended for medium or maximum security offenders, only a Regional Psychiatric Center (RPC) could easily meet such a requirement. If this were a requirement of site selection it may limit its potential exportation to other settings. Nonetheless, it may be feasible for potential sites to allocate a small range or unit for the program, which over time would become exclusively a treatment range. Assignment of a range would facilitate staff training and reliability of behavioral observations. Potential sites should be approached regarding the viability of such an accommodation.

Qualifications of Staff

     It is proposed that a pilot treatment program be provided at three separate sites, selected from different regions, over a period of three years. Each program will require two program deliverers or clinicians, one Ph.D. level psychologist and one social worker/nurse/case management officer, preferably at the postgraduate level. Preferably, Service clinicians will be seconded to this project for the duration. The budget should include costs to back-fill with a term replacement.

Program Description and Duration

     The program will be six months in duration, with provision for follow-up, post-treatment in the form of weekly group meetings. For the 3 year demonstration period, attendance will be closed admission. Once the group of 10 offenders begins, others cannot be admitted until a new group starts. The program will be 4 half-days of group work and a minimum of two individual session per week. Each clinician will serve as the primary therapist for half the group. The program deliverers will meet weekly for half a day to review the group and discuss individual sessions. They will co-author post-treatment reports.

     Other considerations are treatment follow-up sessions, e.g., relapse prevention component for those offenders not released; varying level of community supervision/intervention; and, components or factors which contribute to successes. It is likely that at the end of Year I, formal follow-up in the form of weekly relapse prevention groups will be required. Graduates, correctional staff, and decision-makers will wish to have a vehicle to ensure treatment gains are sustained. As offenders are transferred to reduced security, there will be an additional need for follow-up in the form of relapse prevention groups. The program could also be provided, as described for institutional settings, in the community as a separate pilot site. Regardless, as sufficient offenders begin to be released to the community, post-release programming for treatment continuity must be developed and funded. This is consistent with attempts to develop management strategies for high risk offenders released to the community. A highly structured program would be available with clear guidelines regarding the sharing of information and supervision requirements.

Sample Size

     In order to investigate program effectiveness, a sufficient number of offenders and controls must be available. It is proposed the pilot project be comprised of the following phases: Phase I) pilot single site for one program (6 months) to make revisions and ensure there are no obstacles to full scale implementation at additional sites; Phase II) implementation at all sites, with a review after the first program - end of Year I; Phase III) continuance at sites with data entry for Year I end of Year II; Phase IV) continuance at sites with data entry for Year II - end of Year III. Phase III will permit initial investigations of comparisons between treated and control groups on measures of treatment gain. Phase IV will permit initial outcome research. It is recommended that each site hire a part-time research assistant to complete ongoing assessments and data entry - 60 days per year. Further, the project will require a full-time research assistant beginning at Year II and continuing for up to 3 years (1 year analyses after the last group in Year III).

     For illustrative purposes, assuming 3 sites, by the end of Year I, 40 offenders will have participated in the program (and 40 treated controls assessed). By Year II this will increase to 100 treated offenders, and by Year III to 160. Even with a 10% attrition rate, this will still yield 144 treated and 144 treated control subjects for comparison, with 3 sites.

Overview of Treatment Programs

     The Anger and Emotions Management Program (AEMP) is a 24 session cognitive-behaviorally based program that is delivered over a 12 week period. Program content is highlighted in Figure 2. It is a contemporary program in that it reflects current literature and conceptualizes offender's violence as primarily related to their deficits to identify and manage anger. Cognitions are emphasized, as are skills enhancement in the form of assertiveness and communications training. A relapse prevention component is intended to enhance generalization of treatment gains. Developed in 1993, the program has yet to be systematically evaluated and the assessment strategies of treatment gain remain limited. It is proposed this program be extended to 5 months in duration to control for the length alternate treatment program. This compromise is important so that each program is of comparable length. Otherwise offenders would likely refuse to participate in a longer program, thus attrition rates would be different, significantly impeding interpretations of program effectiveness and comparisons between the treatment program (CMP) and treatment control program (AEMP).

     The Cognitive Mediation Program (CMP) is an extension of work by Guerra and Slaby (1990). The conceptual model is that violent offenders have deficits in social-cognitive skills such as poor problem-solving, attribution of hostility towards others, and self-regulatory difficulties. Arousal and anger are seen to be parenthetic, rather than causal, from an information-processing perspective. In addition, work on resistant clients (Miller & Rollnick, 1991) indicates that engagement is central to issues of compliance and treatment gain. An overview of the treatment targets is presented in Figure 3. Considerable work has already been invested in the development of treatment strategies reflecting the targets in the CMP, yet these require integration and revision prior to implementation.

Research Framework

     Five areas are highlighted which would be systematically investigated during the course of the treatment program, but they would not detract from the intensity or duration of the treatment. Rather, the research initiative is cost effective because it takes advantage of data being collected for the purpose of evaluating program effectiveness and considers additional questions.

1.  The primary goal of this initiative is the evaluation of program effectiveness. This pilot project will investigate process measures and outcome as reflected in the assessment of treatment gain section. While it is clear that intervention is not a panacea, the relative effectiveness of such a treatment program should assist in the selection of specific offenders into treatment and subsequent decisions regarding release.

2.  As part of the initial assessment for the treatment program, a detailed social, clinical and criminal history will be compiled (protocols available upon request). This information will then be employed to develop typologies of violent offenders to empirically reflect the heterogeneity seen clinically. Such typology work with sex offenders has proved helpful in the areas of risk assessment and differential response to treatment. Initially the typologies will include historical information, but current clinical assessments and treatment participation information will be included as dynamic variables.

3.  Treatment responsivity is an important area about which little is known. A major contribution will be the development and potential validation of dynamic variables related to treatment performance and outcome. Such work would be relevant to other offender programs, e.g. substance abuse, cognitive skills, but may have to be revised for particular applications. The interaction between offender typologies and therapist characteristics should also be investigated. A related issue is the role of psychopathy as a moderator variable in treatment compliance and program efficacy.

4.  The investigation of social problem-solving deficits in violent offenders is fundamental to the model underlying the proposed treatment program. The role of impulsivity, anger, and schema will be considered in reviewing accuracy of problem definition and anticipated consequences in problem vignettes (Bettman, 1995). Comparisons between hypothetical versus real problem situations from the offense cycle analysis would also be of interest.

5.  The final research initiative is the development and validation of a multi-method assessment strategy for violent offenders, reflecting the theoretical model noted previously. This assessment battery should inform psychologists regarding risk assessment, in that the relation between particular measures and indices of violence will be demonstrated. The typology data base and the assessment battery could be used to develop a retrospective risk assessment instrument. Follow-up would allow the development of a prospective prediction measure.

     Both the research and clinical information will be maintained in separate confidential files. Summary reports of pre/process/post-treatment test results will be reviewed with the offender and placed on the psychology and case management files. This also applies to the final post-treatment report which will also be sent to the National Parole Board.

Approval and Resources

     This proposal is to be presented to senior administrators for approval and resources. Pursuant to resourcing, several issues require resolution before implementation can proceed.

  • Selection of sites and therapists/program deliverers
  • Finalization of the treatment manual
  • Finalization of the assessment battery


     This project is intended to provide a controlled evaluation of a treatment program for persistently violent (non-sexual) offenders. Both the AEMP and the CMP will be contrasted with a no treatment control group. The interaction between types of violent offenders and various outcomes, including institutional violence and recidivism, will also be investigated. The program content and methodology reflect contemporary knowledge regarding violent offenders and good correctional treatment. Advances in assessment strategies will be shared with other settings with a view towards standardized assessment of treatment requirements for violent offenders. The utility of dynamic or treatment variables in the prediction of violent recidivism will be considered for incorporation into future versions of Offender Intake Assessment, re-assessment, and risk assessment training for correctional staff. Lastly, investigations of the diversity in the violent offender population and differential treatment effects will be completed. This may inform subsequent initiatives regarding hierarchical treatment for violent offenders.


Agee, V. (1990). Rating scale for aggressive juveniles. Unpublished manuscript

Andrews, D.A. and Bonta, J. (1994). The psychology of criminal conduct. Cincinnati: Anderson Publishing.

Andrews, D.A., Zinger, L., 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, 369-404.

Andrews, D.A., Young, J.G., Wormith, J.S., Searle, C.A., & Kouri, M. (1973). The attitudinal effects of group discussions between young criminal offenders and community volunteers. Journal of Community Psychology, 1, 417-422.

Annis, H. (1986). A relapse prevention model for treatment of alcoholics. In W. E. Miller and N. Heather (Eds.), Treating addictive behaviors (pp. 407-435). New York: Plenum.

Barratt, E.S. (1994). Impulsiveness and aggression. In J. Monahan and H.J. Steadman (Eds.), Violence and mental disorder: Developments in risk assessment (pp. 61-79). Chicago: University of Chicago Press.

Baxter, D.J., Marion, A.J., & Goguen, B. (1995). Predicting treatment response in correctional settings. Forum in Corrections Research, 7, 38-41.

Bettman, M. (1995). Cognition and Criminal Violence. Manuscript in Preparation.

Blackburn, R. (1993). The Psychology of Criminal Conduct. Chichester, England: Wiley & Sons Ltd.

California Department of Mental Health, Sex Offender Treatment and Evaluation Project. (1988). Sex offender situational competency test. Unpublished Test.

Cooke, D.J. (1989). Containing violent prisoners: An analysis of the Barlinnie Special Unit. British Journal of Criminology, 29, 129-143.

Correctional Service of Canada. (1995). National Sex Offender Strategy, CSC. Ottawa, Canada: Author.

Dodge, K.A. (1986). A social-information processing model of social competence in children. In M. Perlmutter (Ed.), Minnesota Symposium on Child Psychology. Hillsdale, NJ: Erlbaum.

Gendreau, P. (in press). The principles of effective intervention with offenders. In A.J. Harland (Ed.). Choosing correctional options that work: Defining the demand and evaluating the supply. Thousand Oaks, CA: Sage.

Gendreau, P., Little, T., & Goggin, C. (1995). A meta-analysis of the predictors of adult offender recidivism: Assessment guidelines for classification and treatment Ministry Secretariat, Solicitor General of Canada, Ottawa.

Heilbrun, K.S., Bennett, W.S., Evans, J.H., Offult, R.A., Reiff, H.J., & White, A.J. (1992). Assessing Treatment in Mentally Disordered Offenders: Strategies for improving reliability. Forensic Reports, 5, 85-96.

Lipsey, M., & Wilson, D.B. (1993). The efficacy of psychological, educational, and behavioral treatment: Confirmation from meta-analysis. American Psychologist, 48, 1181-1209.

Marques, J.K., Day, D.M., Nelson, C., & West, M.A. (1994). Effects of cognitive-behavioral treatment on sex offender recidivism: Preliminary results of a longitudinal study. Criminal Justice and Behavior, 21, 28-54.

Motiuk, L.L., Belcourt, R., & Bonta, J. (1995). Managing high- risk offenders: A post-detention follow-up. Research Report R-39. Ottawa: Correctional Service of Canada.

Motiuk, L.L. & Brown, S.L. (1994). Offender needs identification and analysis in community corrections, Forum, 6, 14-16.

Motiuk, L.L. and Brown, S.L. (1993). The validity of offender needs identification and analysis in community corrections. Research Report R-34. Ottawa: Correctional Service of Canada.

Newman, J.P. (1990). Self-regulatory failures in criminal psychopathy. Proceedings of third symposium on violence and aggression. Saskatoon, Sask.

Newman, J.P., & Wallace, J.F. (1993a). Psychopathy and cognition. In K.S. Dobson and P.C. Kendall (Eds.), Psychopathy and Cognition, (pp. 293-349). Orlando, FL: Academic Press, Inc.

Newman, J.P., & Wallace, J.F. (1993b). Diverse pathways of deficient self-regulation: Implications for disinhibitory psychopathology in children. Clinical Psychology Review, 13, 721-740.

Novaco, R.W. (1994). Anger as a risk factor for violence among the mentally disordered. In J. Monahan and H.J. Steadman (Eds.), Violence and mental disorder: Developments in risk assessment. Chicago: University of Chicago Press.

Novaco, R.W., & Welsh, W.N. (1989). Anger disturbances: Cognitive mediation and clinical prescriptions. In K. Howells and C.R. Hollin (Eds.), Clinical Approaches to Violence, (pp. 39-60). London: John Wiley & Sons Ltd.

Ogloff, J.P.R., Wong, S., & Greenwood, A. (1990). Treating criminal psychopaths in a therapeutic community program. Behavioral Sciences and the Law, 8, 81-90.

Pithers, W. D. (1990). Relapse prevention with sexual aggressors: A method for maintaining therapeutic gain and enhancing external supervision. In W. L. Marshall, D. R. Laws, and H. E. Barbaree (Eds.), Handbook of Sexual Assault: Issues, Theories, and Treatment of the Offender. New York: Plenum.

Prentky, R. (1995). Assessment and treatment of rape in the United States. Paper presented at the 4th International Conference on Treatment of Sexual Offenders. Free University, Amsterdam, Amsterdam.

Prochaska, J.O., DiClemente, C.C., & Norcross, J.C. (1992). In search of how people change: Applications to addictive behaviors. American Psychologist, 47, 1102-1114.

Quinsey, V.L., Rice, M.E., & Harris, G.T. (1995). The actuarial prediction of sexual recidivism. The Journal of Interpersonal Violence, 10, 85-105.

Quinsey, V.L., Rice, M.E., Harris, G.T., and Lalumière, M.L. (1993). Assessing treatment efficacy in outcome studies of sex offenders. Journal of Interpersonal Violence, 8, 512- 523.

Rice, M.E. and Harris, G.T. (1993). Treatment for prisoners with mental disorder. In J.H. Steadman and Cocozza (Eds.), Mental illness in America's prisons. Seattle, WA: National Coalition for the Mentally Ill in the Criminal Justice System.

Rice, M.E., Harris, G.T., Quinsey, V.L., & Cyr (1990). Planning treatment programs in secure psychiatric facilities. In D.N. Weisstub (Ed.), Law and mental health: International perspectives, Vol. 5, (pp. 162-230). New York: Pergamon Press.

Rice, M.E., Harris, G.T., Quinsey, V.L., Harris, A.J., & Lang. C. (in press). Treatment of forensic patients. In Sales, B., & Shah, S. (Eds.). Mental health and the law: Research, policy, and practice. Serin, R.C.(1994). Treating violent offenders: A review of current practices. Research Report R-38. Ottawa: Correctional Service of Canada.

Serin, R.C. (1995). Psychopaths and treatment responsivity. Forum on Corrections Research, 7, 23-26.

Serin, R.C. (1993). Survey of Ontario Region anger control programs. Unpublished manuscript.

Serin, R.C. & Kuriychuk, M. (1994). Social and Cognitive Processing Deficits in Violent Offenders: Implications for Treatment, International Journal of Law and Psychiatry, 17, 431-441.

Serin, R.C. & Kuriychuk, M. (1993). Aggressive Beliefs. Unpublished Test.

Slaby, R. G. and Guerra, N. G. (1988). Cognitive mediators of aggression in adolescent offenders: 1. Assessment. Developmental Psychology, 24, 580-588.

White, J.L., Moffit, T.E., Caspi, A., Bartusch, B.J., Needles, D.J., & Stouthamer-Loeber, M.D. (1994). Measuring impulsiveness and examining its relationship to delinquency. Journal of Abnormal Psychology, 103, 192-205.

Zillman, D. (1988). Cognition-excitation interdependencies in aggressive behavior. Aggressive Behavior, 14, 51- 64.

Appendix A

Treatment Targets and Suggested Testing

(Note: This requires further consultation before being finalized)

Target Test(s)
Personality Revised Interpersonal Adjective Scales (Wiggins, 199)
Anger Novaco's Anger Inventory (1994)
Anger Parameters (Novaco)
Anger Knowledge Questionnaire (Serin & Kuriychuk, 1993)
Situations-Reactions Hostility Inventory (Blackburn & Evans, 1985)
Impulsivity Eysenck's 17
Q-sort undercontrol
Psychopathy Revised Psychopathy Checklist (Hare, 1991)
Belligerence Withdrawal Scale (Blackburn)
Cognitive Processing Aggressive Beliefs (Bettman, 1995; Serin & Kuriychuk, 1993)
Card Playing Task (Newman, 1990)
Attribution Vignettes (Serin, 1989)
Hostility Buss Durkee Inventory (Buss & Durkee, 1957)
Needs Antecedents to Crime Inventory (Serin, 1994)
Community Risk/Needs Scale (Motiuk, 1993)
Social Desirability Balanced Inventory of Desirable Responding (Paulus, 1984)
Substance Abuse CLAI (MAST/DAST/Inventory of Drinking Situations)
Empathy Hogan's Empathy Scale
Computerized Task (Newman, 1995)
Victim Empathy (to be developed)
Interpersonal Style Agee (1992)
Chart of Interpersonal Reactions in Closed Living Environments (Blackburn, 1994)
Intelligence Multidimensional Aptitude Test Battery
Moral Reasoning ?
Assertiveness ?
Treatability Serin & Kennedy (1995)
Problem-solving Vignettes (Bettman, 1995)
Criminal Attitudes Criminal Sentiments Scale (Andrews & Wormith, 1984)
Neutralization Scale (Shields, 1994)
Pride in Delinquency Scale (Shields, 1994)
Social Skills ?
Additional Risk Scales Level of Service Inventory - Revised (LSI-R)
Statistical Information on Recidivism
Scale (GSIR)
Risk Assessment Guide (RAG; Webster et al., 1994)
Schema, problem-solving Computerized vignettes, assessment of anticipated consequences
Schema, empathy Digitalized computer presented stimuli, plus response latencies
Impulsivity Response latencies
Empathy Computerized task (Newman, 1995)

Appendix B

Responsivity Factors

1.  Completion of Treatability Section from Psychological Intake Assessment Protocol (considers motivation, prior treatment participation, denial, etc.)

2.  Age

3.  Clinical/social history from PIA

4.  Intelligence (see above)

5.  Psychopathy

6.  Internalization of treatment materials (to be adapted from Miller & Rollnick, 1991)

7.  Attachment (capacity to emotionally interact with another) - to be developed

8.  Measure of treatment gain (knowledge, performance) - draft available for some treatment targets, to be developed

9.  Competency - to be developed from sex offender materials

Appendix C

Admission Criteria

1.  Three or more convictions for assaultive/violent behaviour or

2.  Evidence of persistent institutional violence (assaults, threats towards staff or offenders).

3.  Within 2 years of SRD or WED.

4.  Willing to commit to the full 6 months of the program.

5.  Not presently actively psychotic.

6.  In regular population.

7.  Must agree to group and individual sessions.

8.  Grade 8 literacy level or comparable reading skills.

Program Rules

1.  No violence within group.

2.  Group information remains confidential (exceptions relate to threat of harm or safety issues including within institution).

3.  Regular attendance is mandatory. Missing more than 2 sessions per week without prior approval (e.g., for PFV) or medical reasons for a maximum of 10 sessions per program may result in expulsion.

4.  Assigned homework must be completed.

5.  Informed consent (voluntary and unrelated to release decisions) and group contract (permission to videotape) forms completed.

6.  Weekly behavioral checklists (point system) completed by program deliverers. Monthly checklists completed by other staff.

7.  Post-treatment report shared with offender and CSC/NPB.

Figure 1 -- Research Design

Figure 1 -- Research Design

Figure 2 -- Anger and Emotions Management Program

Figure 2 -- Anger and Emotions Management Program

Figure 3 -- Cognitive Mediation Program

  Figure 3 -- Cognitive Mediation Program