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Warning This Web page has been archived on the Web.


Principles of Effective Correctional Programs


This chapter provides a brief outline of principles of effective correctional treatment. The principles recognize the importance of individual differences in criminal behaviour. A truly interdisciplinary psychology of criminal conduct (PCC: Andrews & Bonta, 1998) has matured to the extent that progress has been made with reference to the achievement of two major scientific standards of understanding. In brief, individual differences in criminal activity can be predicted and influenced at levels well above chance and to a practically significant degree. The following principles of effective treatment draw heavily upon that knowledge base. This does not imply that the research base is anywhere near complete with reference to most issues. Rather, all of the following principles are subject to further investigation, including even those principles with relatively strong research support at this time. Also, principles not even hinted at here are expected to be developed and validated in the coming months and years.

To date, PCC has advanced because it is specific about what it attempts to account for, that is, individual differences in criminal behaviour including reoffending on the part of adjudicated offenders. It has advanced also because it recognizes that the risk factors for criminal conduct may be biological, personal, interpersonal, and/or structural, cultural, political and economic; and may reflect immediate circumstances. PCC does not limit its view to the biological, the personal, or to differential levels of privilege and/or victimisation in social origin as may be indexed by age, race, class and gender. This PCC does not purport to be a psychology of criminal justice, a psychology of social justice, a sociology of aggregated crime rates, or a behavioural or social science of social inequality, of poverty, or of a host of other legitimate but different interests.

In applications of PCC, however, these many other legitimate but different interests may not only be of value but may well be paramount. For example, within criminal law and justice systems, principles of retribution and/or restoration may be considered paramount and hence any correctional treatment efforts, if offered at all, must be offered and evaluated within the retributive and/or restorative context. Similarly, the effects of human service efforts may be evaluated within the context of institutional and/or community corrections. Moreover, ideals of justice, ethicality, decency, legality, safety and cost-efficiency are operating in judicial and correctional contexts as they are operating in other contexts of human endeavour. Thus, the principles of effective human service reviewed here are presented in the context of seeking ethical, legal, decent, cost-effective, safe, just and otherwise normative human service efforts aimed at reducing reoffending.

The phrase “otherwise normative” covers a vast area and is included in recognition of the fact that under some political conditions the values and norms of some privileged groups may be dominant no matter how weak the connection between compliance with their norms and the enhancement of peace and security. For example, sentencing according to criminal law and the principle of specific deterrence continues to occur in Canada and other countries even though there is no consistent evidence that reoffending is reduced through increases in the severity of negative sanctioning. Similarly, principles of effective human service in a justice context may be applied even when the sanctions themselves have been handed down with little concern for reducing reoffending (for example, under a pure just desert sanction) or as an attempt to provide restitution for the victim (for example, under a restorative justice disposition).

The following principles have to do with clinically relevant programming and with setting, staff, implementation and integrity issues. The first set of principles, however, restate and underscore the importance of the theoretical and normative issues referred to in the opening paragraphs. The research evidence is appended along with some relevant references to earlier reviews of principles.


Principle 1

Base your intervention efforts on a psychological theory of criminal behaviour as opposed to a biological, behavioural, psycho-logical, sociological, humanistic, judicial or legal perspective on justice, social equality or aggregated crime rates. When the interest is reduced reoffending at the individual level, theories that focus on some other outcome are of reduced value because they are less likely to identify relevant variables and strategies. The average effects on reduced reoffending of interventions based on alternatives to a psychology of crime have been negative or negligible (See Endnote). In brief, if you are interested in individual differences in criminal activity (for example, reducing reoffending) work from a theory of criminal behaviour.

Principle 2

The recommended psychological perspective is a broad band general personality and social learning approach to understanding variation in criminal behaviour including criminal recidivism. This perspective identifies the eight following major risk factors for criminal behaviour:

  • attitudes, values, beliefs, rationalisations and cognitive emotional states specifically supportive of criminal behaviour;
  • immediate interpersonal and social support for antisocial behaviour;
  • fundamental personality and temperamental supports such as weak self-control, restless aggressive energy and adventurous pleasure seeking;
  • a history of antisocial behaviour including early onset;
  • problematic circumstances in the domains of home, school/work, and leisure/recreation;
  • substance abuse. (Principles 5-8).

The general personality and social learning perspectives also identify the major behavioural influence strategies such as modelling, reinforcement and cognitive restructuring in the context of a reasonably high quality interpersonal relationship (Principle 9, 16) The behavioural base of this perspective also suggests that treatment is best offered in the community-based settings in which problematic behaviour occurs (Principle 4). In addition, the behaviour of workers in correctional settings is also under the influence of cognition, social support, behavioural history and fundamental personality predisposition and hence the emphasis placed on the selection, training and supervision of workers (Principle 16, 17).

Principle 3

Introduce human service strategies and do not rely on the principles of retribution or restorative justice and do not rely on principles of deterrence (specific and/or general) and/or on incapacitation. More-over, seriously consider and introduce but do not rely upon other principles of justice and normative appropriateness such as professional credentials, ethicality, legality, decency, and efficiency. Rather, reductions in reoffending are to be found through the design and delivery of clinically relevant and psychologically appropriate human service under conditions and settings considered just, ethical, legal, decent, efficient, and otherwise normative. In brief, the task assigned by the human service principle of effective service is to design and deliver effective human service in a just and otherwise normative context. The principles of effective human service do not vary greatly with such considerations, although the justice and normative contexts themselves may vary tremendously. The setting factor of community versus institutional corrections, however, does lead to a separate principle.

Principle 4

Community-based services are preferred over residential/ institutional settings but, if justice or other concerns demand a residential or custodial placement, community-oriented services are recommended. Community-oriented services refer to services facilitating return to the community and facilitating appropriate service delivery in the community. The principles of relapse prevention provide guidance for clinically relevant community-oriented services. When services are community-based, a supplementary consideration is to favour home and school-based services rather than agency-based services. For example, the best of the family interventions are not delivered in agency offices but in the natural settings of home and community.


Principle 5 -- Risk

More intensive human services are best reserved for higher risk cases. Low risk cases have a low probability of recidivism even in the absence of service. With the lowest risk cases, justice may be served through just dispositions and there is no need to intro-duce correctional treatment services in order to reduce risk. Indeed, a concern in working with the lowest risk cases is that the pursuit of justice does not inadvertently increase risk through, for example, increased association with offenders and/or the acquisition of pro-criminal attitudes and beliefs. Additionally, recognize that well controlled outcome studies have yet to find reduced reoffending when human service is delivered to the highest risk cases such as very high risk egocentric offenders with extended histories of antisocial behaviour. There is the possibility that psychopaths may put any new skills acquired in treatment to antisocial use (see Principle 10, specific responsivity). At this time, however, there are no well-controlled outcome studies of clinically appropriate treatment with psychopaths.

Principle 6 -- Target Criminogenic Need

Treatment services best attempt to reduce major dynamic risk factors and/or to enhance major protective or strength factors. Criminogenic needs are dynamic risk factors that when reduced are followed by reduced reoffending and/or protective factors that when enhanced are followed by reduced reoffending. Following the major risk factors, the most promising targets include moving antisocial cognition and cognitive emotional states such as resentment in the less antisocial direction, reducing association with antisocial others and enhancing association with anticriminal others, and building self-management, self-regulation and problem solving skills. A history of antisocial behaviour can not be eliminated but new less risky behaviours may be acquired and practised in risky situations (as in relapse prevention programs). Rewards for non-criminal behaviour may be enhanced in the settings of home, school/work and leisure. In the home, the major intermediate targets are enhanced caring, nurturance and mutual respect in combination with monitoring, supervision and appropriate discipline. Similarly, reduced substance abuse may shift the pattern of rewards such that the non-criminal is favoured. The less promising intermediate targets of change include enhancing self-esteem and reducing personal distress without touching personal and interpersonal supports for crime, increasing fear of official punishment, and a focus on other weak risk factors. In summary, for adherence with the need principle, emphasize the reduction of criminogenic need and do not rely upon or emphasize the reduction of noncriminogenic need.

Principle 7 -- Multimodal

Target a number of criminogenic needs. The meta-analyses now make it clear that a number of the criminogenic needs of high-risk cases are best targeted.

Principle 8 -- Assessing risk and dynamic factor

Adherence to the principles of risk and criminogenic need depend upon the reliable and valid assessment of risk and need. The best instruments sample the major risk factors and can provide evidence of validity with younger and older cases, men and women, and different ethnic groups in a number of justice and correctional contexts. Assessments of risk best sample the eight risk factors as well as very specific indicators when specialized outcomes are sought. The latter specific indicators, for example, would include deviant sexual arousal and cognitive and/or social support for sexual offending when reduced sex offending is the desired outcome. Similarly, attitudinal and social support for battering would be specific risk factors when reduced family violence is the desired outcome. Please do not confuse seriousness of the current offence with risk of reoffending. Seriousness of the offence is an aggravating factor at time of sentencing but not a major risk factor.

Principle 9 -- General responsivity

Responsivity has to do with matching the style, modes and influence strategies of service with the learning styles, motivation, aptitude and ability of cases. Generally, offenders are human beings and hence the principle suggests use of the most powerful influencing strategies that have been demonstrated with human beings. Consistent with the general personality and social learning perspective, these most powerful approaches are structured behavioural, social learning and cognitive behavioural influence strategies. These fundamentals include reinforcement, model-ling, skill acquisition through reinforced practice in the context of role playing and graduated approximations, extinction, and cognitive restructuring. Reinforcement, extinction, modelling effects and the attractiveness of the setting of change are all enhanced by high quality interpersonal relationships characterized as open, warm, non-hostile, non-blaming and engaging. Structuring activities include anticriminal modeling and reinforcement, skill building through structured learning, problem solving, advocacy and brokerage, and the effective use of authority (see Principle 16, staff considerations).

Principle 10 -- Specific responsivity and strengths

Specific responsivity factors include personality, ability, motivation, strengths, age, gender, ethnicity/race, language, and various barriers to successful participation in service. The personality set, for example, includes interpersonal anxiety (avoid heavy confrontation), interpersonal and cognitive immaturity (use structured approaches), psychopathy (keep very open communication among all workers) and low verbal intelligence (be concrete). Motivational considerations suggest matching treatment style and goals with level of motivation for change (from not even thinking of change though currently involved in change activities). The relationship principle noted under general responsivity is widely applicable but many feminist scholars stress in particular quality of interpersonal interactions in working with female offenders. Aboriginal writers support the introduction of a spiritual component when working with Aboriginal offenders. When working with reluctant cases the general rule of high quality interpersonal inter-actions is underscored as is the removal of concrete barriers such as inconvenient timing and location of service. Make use of personal, interpersonal and circumstantial strengths in planning and delivering service. Some of these helpful strengths are problem-solving skills, respect for family, a particularly prosocial friend or being happily employed in delivering effective service.

Principle 11 -- Assess responsivity and strength factors

Sophisticated assessment instruments are available for assessment of some of the personality factors and a new generation of risk/need scales are introducing routine assessment of strength and other responsivity factors. Generally, however, watch for particular strengths and for particular barriers for individual cases and for particular groups such as women and minorities.

Principle 12 -- After care, structured follow-up, continuity of care, and relapse prevention

This is introduced as a principle on its own because of the need to stress ongoing monitoring of progress and to intervene when circumstances deteriorate or positive opportunities emerge. Generally, and particularly for residential programs, it is important that programming be community oriented and attend to family, associates and other social settings. Going beyond Principle 4, Principle 12 stresses specific and structured after care and follow-up activity and requires co-ordination of applications of all of the previous principles. At a minimum, in the tradition of relapse prevention, high-risk situations and circumstances are identified and low-risk alternative responses are practiced.

Principle 13 -- Professional discretion

In a few cases, with documented reasons, deviations from the general principles may be introduced. For example, for some young people and their families, it may be recommended that facilitating a move out of a particular apartment building in a particularly high crime area is a priority intermediate goal. Similarly, a major mental disorder such as schizophrenia may move from the minor risk set to the major set when specific symptoms include antisocial thoughts that others are out to get the person and should be “got” first.

Principle 14

Create and record a service plan and any modification of plans through re-assessment of risk/need and progress. The service plan describes how the human service principles of risk, need, general responsivity, specific responsivity, multimodal service, aftercare and professional discretion will be addressed in working with a particular case.


Principle 15 -- Integrity in program implementation and delivery

Integrity has to do with whether the human service activities were introduced and delivered as planned and designed, and indeed whether the delivery of services achieved intermediate objectives. Integrity is enhanced when a highly specific and concrete version of a rational and empirically sound theory is employed. Specificity enhances the opportunity for clarity in who is being served, what is being targeted, and what style, mode and strategy of service is to be used. Specificity readily yields the production of training and program manuals in printed, taped or other formats. Integrity is enhanced when workers are selected, trained, and clinically supervised with particular reference to the attitudes and skills required for effective service delivery. Integrity is enhanced when the clinical supervisor has been trained and has access to highly relevant consultation services. In addition, specificity implies an understanding of when treatment comes to an appropriate end or an understanding of the appropriate closing of the case. The latter implies that service personnel and researchers know when dosage has been adequate and/or when treatment has been delivered successfully and/or when intermediate targets have been achieved. Thus, integrity may be enhanced through the monitoring of service process and monitoring of the achievement of intermediate objectives. At the highest levels of integrity, when clinical supervision or other styles of monitoring identify problematic circumstances (or unanticipated service opportunities) actions are initiated to modify the service plan and to overcome barriers and build on strengths. Involvement of researchers in the design and/or delivery of service amplifies integrity. In summary and in checklist format, integrity depends upon all of the following:

  1. Specific version of a rational and empirically sound theory
  2. Selection of workers
  3. Training of workers
  4. Clinical supervision of workers
  5. Trained clinical supervisors
  6. Consultation services for clinical supervisors
  7. Printed/taped program manuals
  8. Monitoring of intermediate service process
  9. Monitoring of intermediate change
  10. Action to maximize adherence to service process and enhance appropriate intermediate gain
  11. Adequate dosage/duration/intensity
  12. Involve a researcher in the design, delivery and evaluation of service -- in particular, involve a researcher interested in service process, intermediate outcome and ultimate outcome in the design and delivery of service.
  13. Other

Implementation and integrity issues involve staff and management issues to such a degree that their importance is underscored through statements of separate principles of staff and management considerations.

Principle 16 -- Attend to Staff

The selection, training and clinical supervision of staff each best reflect the particular attitudes, skills and circumstances that are supportive of the delivery of the service as planned. Reflecting the general social learning and general responsivity principles, staff skill and cognition supportive of effective practice fall into the five general core practice categories of relationship/interaction skills, structuring/contingency skills, personal cognitive supportive of human service, social support for the delivery of clinically appropriate service, and other considerations.

Relationship. Indicators of relationship skills include some combination of the following: being respectful, open, warm (not cold, hostile, indifferent), caring, non-blaming, flexible, reflective, self confident, mature, enthusiastic, understanding, genuine (real), bright and verbal, and other indicators including elements of motivational interviewing strategies (express empathy, avoid argumentation, roll with resistance). Recall from the general responsivity principle that the effectiveness of modelling, reinforcement and even expressions of disapproval are all enhanced in the context of high quality interpersonal relationships.

Structuring. Indicators of structuring skills include some combination of the following social learning/cognitive behavioural strategies reformulated with particular reference to core effective practices. Modelling anticriminal alternatives to procriminal attitudes, values, beliefs, rationalizations, thoughts, feelings and behavioural patterns; anticriminal differential reinforcement; cognitive restructuring; structured learning skills; the practice and training of problem solving skills; core advocacy/brokerage activity; and effective use of authority. More generally expressed, some indicators are being directive, solution focused, contingency based and, from motivational interviewing, developing discrepancy and supporting beliefs that the person can change his or behaviour (supporting prosocial self efficacy).

Personal cognitive supports. Some specific indicators including:

  • a knowledge base favouring human service activity;
  • a belief that offenders can change;
  • a belief that core correctional practices work;
  • a belief that personally they have the skills to practice at high levels both in terms of relationship and structuring;
  • a belief that important others value core practice and value; and
  • a belief that reducing recidivism is a worthwhile pursuit.

Social support for effective practice. The two major indicators are association with others who practice and support clinically relevant treatment, and relative isolation from anti-treatment others and from others who promote unstructured, non-directive, client-centered practice and/or isolation from others who promote intensive service for low risk cases and promote the targeting of non-criminogenic needs.

Other. Credentials and other factors will be relevant in so far as they tap into the core practices. Obviously, the area of staff considerations is a major area for future research.

A program scores high on staff considerations when:

  1. staff are selected with reference to high level functioning on the relationship, structuring, cognitive and social support dimension of effective correctional practice;
  2. staff receive preservice and inservice training that supports high levels of core practice;
  3. staff receive on-the-job clinical supervision that is concerned with high level functioning in core practice;
  4. staff are actually observed to be functioning at high levels in their exchanges with offenders.

Principle 17 -- Attend to management

Effective managers are assumed to be generally good managers with, additionally, the above-noted relationship and structuring skills along with the knowledge base and their own social sup-port system favourable to clinically relevant and psychologically informed human service. It is management that is responsible for implementing the core principles and creating the supports for creating and maintaining integrity. Effective management will take the steps required to develop program champions inside and outside of the agency. Effective management will reward high functioning staff and have programs and sites accredited.

Principle 18 -- Attending to broader social arrangements

The effective prevention and correctional treatment agency in a public manner will locate crime reduction efforts in the con-text appropriate to local and surrounding conditions. In brief, the correctional agency will be able to clearly locate treatment in locally appropriate contexts of public safety, restorative justice, etc. Similarly, the primary prevention agency will be able to locate their crime prevention efforts in the locally appropriate context of child welfare, family service, mental health, community development, etc.. However, if the host agency is preoccupied with punishment, restoration or child welfare etc. -- if the host agency is not understanding of or interested in clinically relevant approaches to reduced antisocial behaviour -- effectiveness will be reduced.

The endnote supplements the statement of principles with supportive citations, research illustrations and notes on gaps in the research.

1 Carleton University, Department of Psychology


Akers, R. L. (1973). Deviant behavior: A social learning app roach. Belmont, CA: Wadsworth.

Andrews, D. A. (1979). The dimensions of correctional counseling and supervision process in probation and parole. Toronto, ON: Ontario Ministry of Correctional Services.

Andrews, D. A. (1980). Some experimental investigations of the principles of differential association through deliberate manipulations of the structure of service systems. American Sociological Review, 45, 448-462.

Andrews, D. A. (1982). A personal, interpersonal and community-reinforcement perspective on deviant behavior (PIC-R). Toronto, ON: 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), 11-18.

Andrews, D. A. (1995a). Report on an expanded exploration of appropriate correctional treatment. Paper presented at the American Society of Criminology annual meeting, Boston, MA, 1995.

Andrews, D. A. (1995b). The psychology of criminal conduct and effective treatment. In J. McGuire (Ed.), What works: Reducing reoffending: Guidelines from research and practice (pp. 35-62). Chichester, UK: John Wiley & Sons.

Andrews, D. A. (1995c). Assessing program elements for risk reduction: The Correctional Program Assessment Inventory (CPAI). Paper presented at “Research to Results”, a conference of IARCA (now ICCA), Ottawa, ON, October 11-14th.

Andrews, D. A. (1996). Behavioral, cognitive behavioral and social learning contributions to criminological theory. Paper presented at the American Society of Criminology annual meeting, Chicago, Illinois, November.

Andrews, D. A., & Bonta, J. (1998). The psychology of criminal conduct. (2nd edition) Cincinnati, OH: Anderson.

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

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., & Carvell, C. (1998). Core correctional treatment -- Core correctional supervision and counseling: Theory, research, assessment and practice. Ottawa, ON: Carleton University.

Andrews, D. A., & Dowden, C. (under review). Managing correctional treatment for reduced recidivism: A meta-analytical review of program integrity. Manuscript submitted for publication.

Andrews, D. A., Dowden, C., & Gendreau, P. (1999). Clinically relevant and psychologically informed app roaches to reduced reoffending: A meta-analytic study of human service, risk, need, responsivity, and other concerns in justice contexts. Unpublished manuscript, Ottawa, ON: Carleton University.

Andrews, D. A., Gordon, D. A., Hill, J., Kurkowski, K. P., & Hoge, R. D. (1993). Program integrity, methodology, and treatment characteristics: A meta-analysis of effects of family intervention with young offenders. A paper based on a presentation at the meetings of the American Society of Criminology, 1992.

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, 369-404.

Antonowicz, D. H., & Ross, R. R. (1994). Essential components of successful rehabilitation programs for offenders. International Journal of Offender Therapy and Comparative Criminology, 38, 97-104.

Cleland, C. M., Pearson, F., & Lipton, D. S. (1996). A meta-analytic approach to the link between needs-targeted treatment and reductions in criminal offending. American Society of Criminology annual meeting, Chicago, IL, November, 1996.

Dowden, C. (1998). A meta-analytic examination of the risk, need and responsivity principles and their importance within the rehabilitation debate. Unpublished master's thesis, Ottawa, ON: Carleton University, Department of Psychology.

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

Garrett, C. J. (1985). Effects of residential treatment of adjudicated delinquents: A meta-analysis. Journal of Research in Crime and Delinquency, 22, 287-308.

Gendreau, P. (1996). The principles of effective intervention with offenders. In A. Hartland (Ed.), Choosing correctional options that work. Newbury Park, CA: Sage.

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

Gendreau, P., & Goggin, C. (1997). Correctional treatment: Accomplishments and realities. In P. Van Voorhis, M. Braswell, & D. L. Lester (Eds.), Correctional counseling and rehabilitation. Cincinnati, OH: Anderson.

Gendreau, P., & Ross, R. R. (1979). Effectiveness of correctional treatment: Bibliography for cynics. Crime and Delinquency, 25, 463-489.

Gendreau, P., & Ross, R. R. (1987). Revivification of rehabilitation: Evidence from the 1980's. Justice Quarterly, 4, 349-408.

Grant, J., & Grant, M. Q. (1959). A group dynamics approach to the treatment of non-conformists in the Navy. Annals of the American Academy of Political and Social Sciences, 322, 126-135.

Henggeler, S. W., Schoenwald, S. K., Borduin, C. M., Rowland, M. D., & Cunningham, P. B. (1998). Multisystemic treatment of antisocial behavior in children and adolescents. New York, NY: The Guilford Press.

Hill, J. K., Andrews, D. A., & Hoge, R. D. (1991). Meta-analysis of treatment pro-grams for young offenders: The effect of clinically relevant treatment on recidivism, with controls introduced for various methodological variables. Canadian Journal of Program Evaluation, 6, 97-109.

Izzo, R., & Ross, R. (1990). A meta-analysis of rehabilitation programs for juvenile delinquents: A brief report. Criminal Justice and Behavior, 17, 134-142. Lipsey, M.W. (1989). The efficacy of intervention for juvenile delinquency: Results from 400 studies. Paper presented at the 41st annual meeting of the American Society of Criminology, Reno, Nevada.

Lipsey, M.W. (1992). Juvenile delinquency treatment: A meta-analytic inquiry into the variability of effects. In T. D. Cook, H. Cooper, D. S. Cordray, H. Hartmann, L. V. Hedges, R. J. Light, T. A. Louis, & F. Mosteller (Eds.), Meta-analysis for explanation: A casebook (pp. 83-127). New York, NY: Russell Sage Foundation.

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 (pp. 63-78). Chichester, UK: John Wiley & Sons.

Lipsey, M. W. & Wilson D. B. (1997). Effective intervention for serious juvenile offenders: A synthesis of research. A paper prepared for the OJJDP Study Group on Serious and Violent Juvenile Offenders. Vanderbilt University.

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

Lösel, F. (1996). Effective correctional programming: What empirical research tells us and what it doesn't. Forum on Corrections Research, 8 (3), 33-36.

Lösel, F. (1998). The importance of offender programming: German and international evaluations. A paper presented at the International Beyond prisons Symposium, March 15-19, 1998, Donald Gordon Centre, Queen's University, Kingston, ON.

Mayer, J. P., Gensheimer, L. K., Davidson, W. S., & Gottschalk (1986). Social learning treatment within juvenile justice: A meta-analysis of impact in the natural environment. In S. J. Apter & A. Goldstein (Eds.), Youth violence: Programs and prospects. Elmsford, NY: Pergamon.

McGuire, J., & Priestly, P. (1995). Reviewing what works: Past, present and future. In J. McGuire (Ed.), What works: Reducing reoffending -- Guidelines for research and practice (pp 3-34). Chichester, UK.

Palmer, T. (1974, March). The Youth Authority's community treatment project. Federal Probation, 3-14.

Palmer, T. (1975). Martinson revisited. Journal of Research in Crime and Delinquency, 12, 133-152.

Patterson, G.R. (1982). Coercive family process. Eugene, OR: Castalia.

Trotter, C. (1999). Working with involuntary clients: A guide to practice. London, UK: Sage Publications.

Van Voorhis, P., Braswell, M., & Lester, D. L. (1997). Correctional counseling and rehabilitation. Cincinnati, OH: Anderson.

Warren, M. (1971). Classification of offenders as an aid to efficient management and effective treatment. Journal of Crime, Law, Criminology, and Police Science, 62, 239-258.


Comments, some references and some meta-analytic research findings for principles of effective prevention and correctional treatment

Clinically relevant and psychologically informed human service recognizes the importance of individual differences in criminal behaviour, the major importance of immediate personal and interpersonal factors, the more distal significance of broad structural factors, and the importance of differences in approaches to treatment. For years it has been hypothesized that clinically relevant and psychologically informed correctional treatment services could significantly and meaningfully reduce criminal recidivism rates (Andrews, 1979, 1980, 1982, 1989; Andrews, Bonta, & Hoge, 1990; Gendreau & Ross, 1979, 1987; Grant & Grant 1959; Palmer, 1974, 1975; Warren, 1971). Now, meta-analytic reviews of controlled outcome studies support not only the value of correctional treatment but of clinically and psychologically appropriate treatment in particular (Andrews, 1995a; Andrews & Bonta, 1998: Resource Note 10.1; Andrews, Gordon, Hill, Kurkowsky, & Hoge 1993; Andrews, Zinger, Hoge, Bonta, Gendreau, & Cullen 1990; Antonowicz & Ross 1994; Cleland, Pearson, & Lipton 1996; Garrett, 1985; Hill, Andrews, & Hoge 1991; Izzo & Ross, 1990; Lipsey, 1989, 1992, 1995; Lipsey & Wilson, 1997: Lösel, 1995, 1996, 1998; Mayer, Gensheimer, Davidson, & Gottschalk, 1986). Now, many scholars and practitioners have provided evidence-based guidelines for appropriate service and the level of agreement among the guidelines, while not perfect, is substantial (Andrews, 1995b; Andrews, Bonta, & Hoge, 1990; Gendreau, 1996; McGuire & Priestly, 1995; Lipsey, 1995; Lösel, 1995, 1998; Van Voorhis, Braswell, & Lester, 1997). [from Andrews, Dowden & Gendreau, 1999]

The meta-analytic findings reported herein are based on analyses of the Carleton University data bank (Andrews, Zinger et al., 1990; Andrews, Dowden, & Gendreau, 1999; Dowden, 1998; Dowden & Andrews, 1999). The mean effect sizes reported below may be interpreted as the difference in percentage rate of recidivism between treatment and comparison groups. For example an effect size of 0.20 reflects a recidivism rate of 40% in the treatment group (50 minus 20/2) and 60% in the comparison group (50 plus 20/2). A positive difference reflects relative success in that the mean recidivism rate of the treatment group was lower than that of the comparison group. A negative difference reflects relative failure in that the mean recidivism rate of the comparison group was lower than that of the treatment group. For example, with an effect size of -0.10, the recidivism rate would be 55% in the treatment group (50 plus 10/2) and 45% in the comparison group (50 minus 10/2).

Principles 1 and 2: The general personality and social learning perspective on criminal behaviour and prevention programming is the most promising perspective whether the context is restorative justice, retributive justice, or outside of the justice system, and/or whether the setting is community-based or residential/ custodial. Similarly, the perspective applies across categories of age, gender, race/ethnicity and social class (for examples of these perspectives see: Akers, 1973; Andrews, 1982, 1996; Andrews & Bonta, 1994, 1998; Henggeler et al., 1998; Patterson,1982). Even without applying the clinically relevant and psychologically informed principles, it is obvious from the meta-analyses that programs with an immediate personal and interpersonal focus do much better in terms of reduced reoffending than do programs based on broad social location and/or social reaction perspectives. A mild average reduction in reoffending for personality and social psychological approaches (0.10, k = 325) compares very favourably with the mild mean increase in reoffending found for the more sociological approaches (-0.03, k = 49) and with the mean increase found for deterrence programs intending to increase fear of official punishment (-0.05. k = 43).

Principle 3: The evidence favouring a human service approach in the justice context is now overwhelming. Programs based on retribution, just desert, deterrence, and/or restorative justice by themselves do not yield impressive reductions in reoffending. In the Carleton University sample, the average effect of 101 tests of the effects of increases in the severity of the punishment is a mild increase in recidivism (-0.03, k = 101). This average effect is dismal in comparison with the modestly positive mean effect of 0.12 for human service when offered in the justice context of diversion, community corrections and/or institutional corrections (k = 273). The University of New Brunswick, St John group, (See Chapter 3 of this Compendium) looked even more closely at the effects of sanctions and the findings are devastating for those who emphasize retribution, deterrence, restoration and/or diversion without the delivery of human service. The aver-age effect of community sanctions (in 140 tests involving over 50,000 offenders) was zero (ranging from -0.07 for scared straight through 0.04 for fines). The average effect of incarceration relative to a community sanction was -0.07 (k = 103, N = 267,804) and the average effect of more vs less custodial time was -0.03 (k = 222, N = 68,248). Please note that the studies of incarceration did not include estimates of incapacitation effects and hence the already mild mean negative effects on reoffending may be overestimating the negative effects.

Principle 4: The meta-analytic evidence in regard to community-based programming suggests that the principles of effective human service are best introduced before the research findings relevant to community settings are reviewed.

Principle 5: Support for the risk principle is now moderate to strong. The support increases as you move up from studies of the effects of sanctions through studies of human service in general to studies of human service that is consistent with the need and general responsivity principles. The average effect of criminal sanctions is mildly negative with low risk cases (-0.05, k = 34) and with higher risk cases (-0.02, k = 256). The mean effect of human service, however, is much more positive with higher risk cases (0.14, k = 211) than with lower risk cases (0.07, k = 62). When the human service is in adherence with the need and/or general responsivity principles, the effects of risk become quite substantial. For example, with adherence to need, the mean effect of service is 0.19 (k = 169) for higher risk cases compared to -0.01 (205) for lower risk cases. The comparable figures with adherence to general responsivity are 0.23 (77) and 0.04 (297) for higher and lower risk cases respectively.

Our understanding of the risk principle, however, is still limited by the relatively few studies that actually report separate effects for the lower and higher risk cases. Still more limited is knowledge of treatment effects among the lowest, low, middle, high, and very high-risk cases (including psychopaths).

Principles 6 and 7: Support for the need principles increased dramatically with the completion of Dowden's (1998) MA thesis (Andrews, Dowden, & Gendreau, 1999). Using the Andrews and Bonta (1994, 1998; Andrews, 1989) classification of more promising and less promising targets (that is, their lists of criminogenic and non-criminogenic needs) there was a clear association between the number of criminogenic needs targeted and reduced recidivism (0.55, k = 374). In dramatic contrast, the effect sizes decreased with the number of non-criminogenic needs targeted (-0.18, k =374). This dramatic difference underscores that the multimodal principle refers only to increases in the number of criminogenic needs targeted -- increases in the number of non-criminogenic needs targeted contribute to reduced effect sizes. Indeed, the mean effect sizes were negative when the number of non-criminogenic needs targeted exceeded the number of criminogenic needs. The mean effect sizes increased directly with how many criminogenic needs targeted exceeded the number of non-criminogenic needs. When the number of non-criminogenic needs targeted were subtracted from the number of criminogenic needs the difference scores ranged from -3 to +6 across 374 tests of treatment. The corresponding mean effect sizes were as follows for the difference scores of -3 through plus 6: -07 (-3, k = 9), -05 (-2, k = 14), -0.00 (-1, k = 93), -0.00 (0, k = 91), 0.14 (1, k = 71), 0.19 (2, k = 27), 0.22 (3, k = 40), 0.25 (4, k = 17), 0.32 (5, k = 7), 0.51 (6, k = 5). Our simple measure for exploration of principle 6 was a difference score of 1 or more compared to a difference score of 0 or less. The corresponding mean effect sizes were 0.19 (k = 169) and -0.01 (k = 205).

As strong as the above-noted findings may be there are serious limitations and gaps in knowledge. Ultimately, what is required are experimental investigations in which the effects of treatment on recidivism can be shown to be reduced through statistical controls for measured changes in the needs targeted. Currently there are very few studies that allow such explorations. Similarly, experimental tests of some types of need with some types of cases have rarely been conducted. For example, experimental tests of programs that target the low self-esteem of women offenders are so rare that we have yet to find a single one. Similarly, tests of particular dynamic risk factors in programs focusing upon sex offending and other types of violent offending are few in number.

Principle 9: Support for social learning and cognitive behavioural influence strategies is readily found in all but one of the meta-analyses of the effects of correctional treatment. The Carleton University meta-analyses have supported the principle of general responsivity in re-analyses of the studies reviewed by the one negative review as well as in three additional sets of tests of treatment. Overall, the mean effect size in 77 tests of social learning/cognitive behavioural strategies was 0.23 (k = 77) compared with 0.04 for 297 tests of other intervention strategies. The general responsivity principle has also been stated with reference to the relationship and structuring aspects of correctional treatment. Research on these statements of general responsivity are presented in the comments on the staff principle (Principle 16).

Principles 3, 5, 6, 7, and 9 in combination: The findings are very clear. Mean effect size increases directly with adherence to the principles of human service, risk, need and general responsivity. The average effect size for tests of criminal sanctions without the delivery of human service and tests of human service that adhered to not one of risk, need and responsivity was -0.02 (k = 124). The mean effect size for human service pro-grams that adhered to at least one of risk, need and general responsivity was equally as unimpressive at 0.02 (k = 106). Human service programs adhering to at least two of the principles of clinically relevant and psychologically informed human service yielded a mean effect size of 0.18 (k = 84). Adherence to all three of the human service principles yielded a mean effect size of 0.26 (k = 60). The evidence suggests that adherence to the principles of clinically relevant and psychologically informed human service is rewarded by substantial reductions in recidivism.

Principle 10: In brief, a meta-analytic review of specific responsivity is required and additional primary studies of differential treatment are required. Few question the idea that treatment strategies are best matched with case characteristics, and yet studies of recidivism rates as a function of variation in both case characteristics and treatment strategies are so few that conclusions are not yet possible from the meta-analyses. Interestingly, 12 tests of treatment did target particular barriers to treatment and an above-average mean effect size was found.

Principles 8 and 11: The human service principles of risk, need and responsivity are of particular importance because the findings of assessments now may be linked directly with the practical decisions required when clinicians and managers wish to maximize the positive effects of treatment. Note that assessments of risk may now be used not to justify enhanced punishment and control but to guide the intensity of human service efforts. Assessments of criminogenic need identify the appropriate intermediate targets of service and responsivity assessments suggest individualized treatment strategies. Discussions of reliable and valid assessment instruments may be found in Gendreau, Little, and Goggin (1996) and Andrews and Bonta (1998).

Principle 12: This principle emphasizes the value of after care, structured follow-up, continuity of care, and a community orientation through an emphasis on relapse prevention. Elements of a relapse prevention orientation were evident in only 18 of the tests of treatment but those few tests did yield an above average mean effect size. A community-based focus was associated with enhanced effect sizes as evidenced by the findings supportive of targeting associates, family and school/work (the need principle).

Principles 13 and 14: To date, the principles of professional discretion and case planning/recording have not been explored through meta-analytic summaries of links with the effects of treatment. However, emerging studies with the Correctional Program Assessment Inventory (Andrews, 1995c; Gendreau & Goggin, 1997) are showing that ongoing programs that have implemented systematic risk/need assessment and reveal adherence to the principle present promising recidivism rates. Ongoing CPAI research also speaks to Principle 15.

Principle 15: Meta-analytic tests of implementation and integrity are generally supportive of the importance of theoretical specificity, staff selection, training and supervision, printed/taped training or skill manuals, small program units (as inferred from small sample studies), involvement of researcher, and duration of service (Andrews & Dowden, under review). Monitoring of service process and/or intermediate change was not found to link with effect size. The value of consultation services for clinical supervisors is unexplored. A weakness in this literature is that programs with indications of integrity tend also to be among the best representatives of clinically relevant and psychologically informed treatment. With integrity and clinically appropriate treatment so highly correlated, it is difficult to show that integrity greatly enhances effect size. It, however, is know that indicators of integrity are unrelated to outcome when treatment is clinically inappropriate. In other words, there is no meta-analytic evidence at all that introducing clinically inappropriate treatment with high levels of integrity is of any value. There are two great needs in this area: an increase in the number of primary studies focusing on implementation and integrity, and increased attention to reporting on integrity in all controlled outcome studies.

Principles 16, 17, and 18. Scoring of the selection, training and clinical supervision of staff becomes moves beyond general integrity (Principle 15) so that the issue becomes selection, training and supervision with particular reference to the demands of the general responsivity principle. Recall, according to general social learning theory (Andrews, 1980; Andrews & Bonta, 1998) and general social learning influence strategies (Andrews, 1979; Andrews & Carvell, 1998), two dimensions are crucial. The two dimensions are quality of the interpersonal relationship and the structuring skills of the worker. Without evidence that staff were selected according to relationship or structuring skills, the mean effect size was 0.05 based on the vast majority of the tests of treatment (k = 327). However, when one or both of the core dimensions were considered, the mean effect sizes varied from 0.25 to 0.36 (k = 47 in total).

Specific elements of structuring in practice yielded mean effect sizes of 0.31 for high level reinforcement (k = 15), 0.30 for high levels of clinically appropriate disapproval (k = 8), 0.30 for structured skill training (k = 38), 0.28 for clinically appropriate modeling (k = 37), 0.26 for clinically appropriate use of authority (k = 15), 0.25 for problem solving (k = 45), and 0.13 for advocacy brokerage (k = 53). Coding according to core correctional practices (CCP) now constitutes an enhanced coding of general responsivity. In the future, the elements of CCP, soon to also include elements of motivational interviewing and cognitive restructuring, may be scored as selection factors, training factors, clinical supervision factors, and as observed elements of treatment process.

The general applicability of these elements of effective practice have been underscored by Trotter (1999) who has produced a model of social work practice with involuntary clients. He focuses on role clarification (authority), pro-social model-ling and reinforcement, problem solving and relationship.

Two additional staff considerations follow directly from a general personality and social learning perspective on human behaviour. Staff performance on indicators of clinically relevant practice would reflect their relevant skills, behavioural history and personal predispositions. Additionally, performance reflects cognitive supportive of such practice and social support for clinically relevant practice. These two factors remain virtually unexplored.

Research that links management concerns and broader social arrangements to actual impact on recidivism is lacking.

We all look forward to an expanded set of principles with stronger research support. Personally, I think that some major advances are soon going to come from studies of female offenders, aboriginal offenders, treatment in restorative justice contexts, treatment in forensic mental health contexts, and primary prevention in non-justice children's and family services. The intellectual energy and expanding public support for experimentation in those areas is very impressive and very promising.


Previous PageTop Of Page Table Of ContentsNext Page