WORKING WITH OFFENDERS
Approaches to working with offenders have inevitably changed as our understanding of antisocial and criminal behaviour has developed, moving from psychodynamic psychotherapy, through group therapy, to behaviour modification. Yet there are those who see little merit in using treatment to reduce offending. (Hollin, 2001, has documented the struggle between proponents of treatment and advocates of punishment.) But since the mid 1990s there has been a renewed interest in the treatment approach, stimulated by a clutch of studies using meta-analysis. Using meta-analysis to inform treatment programmes Meta-analysis allows inspection of the aggregated findings from a group of studies around a common theme. Its use in studies into the effects of offender treatment has had a profound effect on recent practice.
Offender treatment meta-analyses draw the critical distinction between clinical and criminogenic outcome variables. In this context, ‘clinical outcomes’ refers to changes in some dimension of personal functioning, such as psychological adjustment, attitudes or social competence. On the other hand, ‘criminogenic outcomes’ refers specifically to measures concerned with crime, such as selfreported delinquency, official reconvictions and type of offence. As a broad generalization, treatment of offenders (as with other populations) tends to produce beneficial clinical outcomes (Lipsey & Wilson, 1993). But a significant contribution of the meta-analyses has been to highlight influences on criminogenic outcomes (in other words, those characteristics of treatment interventions that produce a reduction in offending). Several meta-analytic studies have sought to identify the practical recommendations that can be taken from this empirical research (see McGuire, 2002, for an overview). The first major conclusion is that there is an overall reduction in reoffending after treatment – in the region of 10 per cent (Lipsey, 1992; Lösel, 1996). The second conclusion is that some interventions have a significantly greater effect than others – the most effective producing more than 20 per cent reduction in reoffending (Lipsey, 1992). As the evidence accumulates, a broad consensus has been reached regarding the characteristics of treatments that impact on offending:
1. Indiscriminate targeting of treatment programmes is counterproductive in reducing recidivism. Medium- to high-risk offenders should be selected and programmes should focus on criminogenic targets: that is, treatments should be concerned with those aspects of the offender’s thinking and behaviour that can be shown to be directly related to their offending.
2. The type of treatment programme is important, with stronger evidence for structured behavioural and multimodal approaches than for less focused approaches. (The term ‘multimodal’ means using a variety of treatment techniques to address a range of targets for change, as discussed below with reference to Aggression Replacement Training.)
3. The most successful studies, while behavioural in nature, include a cognitive component, i.e. they encourage the offender to focus on their attitudes and beliefs.
4. Treatment programmes should be designed to engage high levels of offender responsivity: that is, the style of treatment should engage the offender to make him or her responsive to treatment and, at the same time, be responsive to the needs of different offenders such as juvenile or adult offenders or male and female offenders.
5. Treatment programmes conducted in the community have a stronger effect than residential programmes. While residential programmes can be effective, they should be linked structurally with community-based interventions.
6. The most effective programmes have high treatment integrity, in that they are carried out by trained staff, and treatment initiators are involved in all the operational phases of the treatment programmes.
The translation into practice of these principles derived from meta-analysis has become known as the What Works form of treatment programmes (McGuire, 1995). [What Works generic name given to a recent approach to offender treatment, which is based on findings from metaanalyses of the offender treatment literature]
The possibilities raised by the What Works principles have been recognized in the UK at a government policy level (Vennard, Sugg & Hedderman, 1997) and have significantly influenced work with offenders in prison and on probation. The development of national programmes for working with offenders has become a major initiative, seeking to capitalize on the possibilities raised by What Works (Lipton et al., 2000).
Offending behaviour programmes – an example Aggression Replacement Training (ART) is an excellent example of a programme approach to working with offenders. [Aggression Replacement Training (ART) research-based programme for working with violent offenders] ART was developed in the USA during the 1980s as a means of working with violent offenders.
This training programme has proved to be an effective way of reducing aggressive behaviour (Goldstein & Glick, 1987; 1996). ART has continued to be developed as the evidence base grows and practice techniques become more refined (Goldstein, Glick & Gibbs, 1998).
ART consists of three components, delivered sequentially, and so would qualify as a multimodal programme:
1. Skillstreaming involves the teaching of skills to replace out-of-control, destructive behaviours with constructive, prosocial behaviours. Social skills are taught in terms of step-by-step instructions for managing critical social situations. For example, offenders might be taught conflict negotiation skills for use in situations where previously they would have used aggression.
2. Anger control training first establishes the individualspecific triggers for anger, then uses the anger management techniques of
(i) enhancing awareness of internal anger cues,
(ii) teaching coping strategies,
(iii) skills training,
(iv) self-instruction and
(v) social problem solving.
Thus, offenders are taught to recognise their own feelings of anger and then helped to develop strategies, using new skills and enhanced self-control, to control anger and hence reduce aggression.
3. Moral reasoning training is concerned with enhancing moral reasoning skills and widening social perspectivetaking. This is achieved through self-instruction training, social problem solving and skills training. The focus here is on increasing the offenders’ understanding of the effects of their actions on others people, thereby enhancing the values that young people have for the rights and feelings of others.
[Don A. Andrews (1941– ) is Professor of Psychology at Carleton University, Ottawa. He is arguably the most influential psychologist currently working in the field of offender rehabilitation. At a time when the practice of offender rehabilitation was vigorously challenged, Andrews was one of its staunchest defenders. He is a fierce critic of those criminologists who have questioned the place of psychology in understanding crime and criminal behaviour. A strong advocate of evidence-based practice and theoretical integrity, he developed a risk assessment instrument (the Level of Service Inventory) which is widely used by practitioners in the criminal justice systems of several different countries. His book, with James Bonta, The Psychology of Criminal Conduct, is a fine example of both his forthright style and the outstanding quality of his work.]
The effectiveness of rehabilitation revisited - A research issue
Not surprisingly, there was significant opposition to the notion that ‘nothing works’ in offender rehabilitation. The literature included in the original review was re-examined by several researchers, who all reached different conclusions. Other researchers assembled different sets of relevant studies, which showed (they claimed) that treatment was effective. But still the broadly accepted position in government policy and community practice was that nothing works, and policies for managing offenders become increasingly punitive. (Even Martinson’s later article published in 1979, recanting much of his earlier views, failed to have any impact.) The complex task of making sense of a large body of literature using a narrative review is always liable to lead to disagreement. The development and refinement in the mid 1980s of the technique of meta-analysis presented a more systematic and objective alternative to the narrative review as a means of making sense of the findings of a large body of literature. The largest and most influential meta-analysis of the offender treatment literature to date was conducted by Lipsey (1992).
Design and procedure:- The first step in Lipsey’s work was to establish the eligibility criteria for studies to be included in the meta-analysis. There were six criteria used in making decisions about inclusion, ranging from the nature of the outcome variables to the type of research design. The next step was to gather the research studies together using searches of bibliographic databases. Lipsey noted that these searches produced ‘more than 8000 citations’ (p. 89) of potential relevance to the study. Once the individual studies had been collected and passed through the eligibility criteria, they were coded for analysis. For the 443 studies included in the meta-analysis, Lipsey used a 154-item coding scheme, incorporating study characteristics such as type of treatment, research design, length of treatment, type of outcome measure, and so on. Once coded, the data represented the characteristics and findings of the 443 individual studies to be statistically analysed using metaanalytic procedures.
Results and implications:- In meta-analysis a key outcome is effect size, which can be calculated in several ways but represents the outcome of the comparison between treatment and no treatment. It is also possible to calculate whether an effect size is statistically significant. So with regard to recidivism, a positive effect size would indicate that treatment reduced offending, while a negative effect size would indicate that treatment increased offending. The magnitude of the effect size indicates the numerical difference in recidivism between treated and untreated offenders. A meta-analysis allows comparisons of the effect size of, say, different treatment types or treatment effects in different settings. Lipsey reported an overall small positive effect size (a statistical measure of the impact of the treatment), so while it would not be true to say that ‘nothing works’, neither could an overwhelmingly strong case be made for treatment. Importantly, meta-analysis also allows researchers to identify the characteristics of ‘high effect’ treatments (those treatments that produce a significantly high reduction in recidivism compared to no treatment). For example, Lipsey’s analysis strongly indicated that structured treatments, generally using cognitive–behavioural methods of treatment, gave greater positive effects in reducing recidivism than treatments based on non-directive counselling. The impact of Lipsey’s work, taken in conjunction with other meta-analyses, can be seen in a large-scale resurgence in methods of offender treatment. This renewed interest has attracted significant government funding under the banner of ‘What Works’, with renewed endeavours in both research and practice.
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 et al. (eds), Meta-analysis for Explanation: A Casebook, New York: Russell Sage Foundation.
Martinson, R., 1979, ‘New findings, new views: A note of caution regarding sentencing reforms’, Hosfra Law Review, 7, 242–58.
Approaches to working with offenders have inevitably changed as our understanding of antisocial and criminal behaviour has developed, moving from psychodynamic psychotherapy, through group therapy, to behaviour modification. Yet there are those who see little merit in using treatment to reduce offending. (Hollin, 2001, has documented the struggle between proponents of treatment and advocates of punishment.) But since the mid 1990s there has been a renewed interest in the treatment approach, stimulated by a clutch of studies using meta-analysis. Using meta-analysis to inform treatment programmes Meta-analysis allows inspection of the aggregated findings from a group of studies around a common theme. Its use in studies into the effects of offender treatment has had a profound effect on recent practice.
Offender treatment meta-analyses draw the critical distinction between clinical and criminogenic outcome variables. In this context, ‘clinical outcomes’ refers to changes in some dimension of personal functioning, such as psychological adjustment, attitudes or social competence. On the other hand, ‘criminogenic outcomes’ refers specifically to measures concerned with crime, such as selfreported delinquency, official reconvictions and type of offence. As a broad generalization, treatment of offenders (as with other populations) tends to produce beneficial clinical outcomes (Lipsey & Wilson, 1993). But a significant contribution of the meta-analyses has been to highlight influences on criminogenic outcomes (in other words, those characteristics of treatment interventions that produce a reduction in offending). Several meta-analytic studies have sought to identify the practical recommendations that can be taken from this empirical research (see McGuire, 2002, for an overview). The first major conclusion is that there is an overall reduction in reoffending after treatment – in the region of 10 per cent (Lipsey, 1992; Lösel, 1996). The second conclusion is that some interventions have a significantly greater effect than others – the most effective producing more than 20 per cent reduction in reoffending (Lipsey, 1992). As the evidence accumulates, a broad consensus has been reached regarding the characteristics of treatments that impact on offending:
1. Indiscriminate targeting of treatment programmes is counterproductive in reducing recidivism. Medium- to high-risk offenders should be selected and programmes should focus on criminogenic targets: that is, treatments should be concerned with those aspects of the offender’s thinking and behaviour that can be shown to be directly related to their offending.
2. The type of treatment programme is important, with stronger evidence for structured behavioural and multimodal approaches than for less focused approaches. (The term ‘multimodal’ means using a variety of treatment techniques to address a range of targets for change, as discussed below with reference to Aggression Replacement Training.)
3. The most successful studies, while behavioural in nature, include a cognitive component, i.e. they encourage the offender to focus on their attitudes and beliefs.
4. Treatment programmes should be designed to engage high levels of offender responsivity: that is, the style of treatment should engage the offender to make him or her responsive to treatment and, at the same time, be responsive to the needs of different offenders such as juvenile or adult offenders or male and female offenders.
5. Treatment programmes conducted in the community have a stronger effect than residential programmes. While residential programmes can be effective, they should be linked structurally with community-based interventions.
6. The most effective programmes have high treatment integrity, in that they are carried out by trained staff, and treatment initiators are involved in all the operational phases of the treatment programmes.
The translation into practice of these principles derived from meta-analysis has become known as the What Works form of treatment programmes (McGuire, 1995). [What Works generic name given to a recent approach to offender treatment, which is based on findings from metaanalyses of the offender treatment literature]
The possibilities raised by the What Works principles have been recognized in the UK at a government policy level (Vennard, Sugg & Hedderman, 1997) and have significantly influenced work with offenders in prison and on probation. The development of national programmes for working with offenders has become a major initiative, seeking to capitalize on the possibilities raised by What Works (Lipton et al., 2000).
Offending behaviour programmes – an example Aggression Replacement Training (ART) is an excellent example of a programme approach to working with offenders. [Aggression Replacement Training (ART) research-based programme for working with violent offenders] ART was developed in the USA during the 1980s as a means of working with violent offenders.
This training programme has proved to be an effective way of reducing aggressive behaviour (Goldstein & Glick, 1987; 1996). ART has continued to be developed as the evidence base grows and practice techniques become more refined (Goldstein, Glick & Gibbs, 1998).
ART consists of three components, delivered sequentially, and so would qualify as a multimodal programme:
1. Skillstreaming involves the teaching of skills to replace out-of-control, destructive behaviours with constructive, prosocial behaviours. Social skills are taught in terms of step-by-step instructions for managing critical social situations. For example, offenders might be taught conflict negotiation skills for use in situations where previously they would have used aggression.
2. Anger control training first establishes the individualspecific triggers for anger, then uses the anger management techniques of
(i) enhancing awareness of internal anger cues,
(ii) teaching coping strategies,
(iii) skills training,
(iv) self-instruction and
(v) social problem solving.
Thus, offenders are taught to recognise their own feelings of anger and then helped to develop strategies, using new skills and enhanced self-control, to control anger and hence reduce aggression.
3. Moral reasoning training is concerned with enhancing moral reasoning skills and widening social perspectivetaking. This is achieved through self-instruction training, social problem solving and skills training. The focus here is on increasing the offenders’ understanding of the effects of their actions on others people, thereby enhancing the values that young people have for the rights and feelings of others.
[Don A. Andrews (1941– ) is Professor of Psychology at Carleton University, Ottawa. He is arguably the most influential psychologist currently working in the field of offender rehabilitation. At a time when the practice of offender rehabilitation was vigorously challenged, Andrews was one of its staunchest defenders. He is a fierce critic of those criminologists who have questioned the place of psychology in understanding crime and criminal behaviour. A strong advocate of evidence-based practice and theoretical integrity, he developed a risk assessment instrument (the Level of Service Inventory) which is widely used by practitioners in the criminal justice systems of several different countries. His book, with James Bonta, The Psychology of Criminal Conduct, is a fine example of both his forthright style and the outstanding quality of his work.]
The effectiveness of rehabilitation revisited - A research issue
Not surprisingly, there was significant opposition to the notion that ‘nothing works’ in offender rehabilitation. The literature included in the original review was re-examined by several researchers, who all reached different conclusions. Other researchers assembled different sets of relevant studies, which showed (they claimed) that treatment was effective. But still the broadly accepted position in government policy and community practice was that nothing works, and policies for managing offenders become increasingly punitive. (Even Martinson’s later article published in 1979, recanting much of his earlier views, failed to have any impact.) The complex task of making sense of a large body of literature using a narrative review is always liable to lead to disagreement. The development and refinement in the mid 1980s of the technique of meta-analysis presented a more systematic and objective alternative to the narrative review as a means of making sense of the findings of a large body of literature. The largest and most influential meta-analysis of the offender treatment literature to date was conducted by Lipsey (1992).
Design and procedure:- The first step in Lipsey’s work was to establish the eligibility criteria for studies to be included in the meta-analysis. There were six criteria used in making decisions about inclusion, ranging from the nature of the outcome variables to the type of research design. The next step was to gather the research studies together using searches of bibliographic databases. Lipsey noted that these searches produced ‘more than 8000 citations’ (p. 89) of potential relevance to the study. Once the individual studies had been collected and passed through the eligibility criteria, they were coded for analysis. For the 443 studies included in the meta-analysis, Lipsey used a 154-item coding scheme, incorporating study characteristics such as type of treatment, research design, length of treatment, type of outcome measure, and so on. Once coded, the data represented the characteristics and findings of the 443 individual studies to be statistically analysed using metaanalytic procedures.
Results and implications:- In meta-analysis a key outcome is effect size, which can be calculated in several ways but represents the outcome of the comparison between treatment and no treatment. It is also possible to calculate whether an effect size is statistically significant. So with regard to recidivism, a positive effect size would indicate that treatment reduced offending, while a negative effect size would indicate that treatment increased offending. The magnitude of the effect size indicates the numerical difference in recidivism between treated and untreated offenders. A meta-analysis allows comparisons of the effect size of, say, different treatment types or treatment effects in different settings. Lipsey reported an overall small positive effect size (a statistical measure of the impact of the treatment), so while it would not be true to say that ‘nothing works’, neither could an overwhelmingly strong case be made for treatment. Importantly, meta-analysis also allows researchers to identify the characteristics of ‘high effect’ treatments (those treatments that produce a significantly high reduction in recidivism compared to no treatment). For example, Lipsey’s analysis strongly indicated that structured treatments, generally using cognitive–behavioural methods of treatment, gave greater positive effects in reducing recidivism than treatments based on non-directive counselling. The impact of Lipsey’s work, taken in conjunction with other meta-analyses, can be seen in a large-scale resurgence in methods of offender treatment. This renewed interest has attracted significant government funding under the banner of ‘What Works’, with renewed endeavours in both research and practice.
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 et al. (eds), Meta-analysis for Explanation: A Casebook, New York: Russell Sage Foundation.
Martinson, R., 1979, ‘New findings, new views: A note of caution regarding sentencing reforms’, Hosfra Law Review, 7, 242–58.
No comments:
Post a Comment