Effective Correctional Treatment and Violent Reoffending: A Meta-Analysis
Dowden, Craig, Andrews, Donald A., Canadian Journal of Criminology
Designing and implementing effective correctional programs for offender populations is of importance to criminal justice practitioners. In this regard, the amount of attention given to the "what works" rather than "nothing works" perspective has steadily increased over the past two decades. More importantly, the methodologies used to examine the effectiveness of correctional treatment have improved significantly. Traditionally, narrative literature reviews were used when researchers were interested in summarizing the findings of this body of literature. Several problems existed with this approach such as the subjectivity of the process and the limited scope of the technique and, thus, researchers searched for more systematic and quantitative approaches to provide appropriate summaries of this research literature. Consequently, meta-analytic techniques were introduced into the field of correctional treatment.
What is a meta-analysis?
A meta-analysis is the statistical aggregation of the results from a large collection of independent studies for the purposes of integrating the findings. The results from each of these studies are converted into a common metric, termed an effect size, to enable cross-study comparison. An effect size is an estimate of the size of the relationship between two variables (Rosenthal 1991) and is the primary unit of analysis within a meta-analysis. The ability of meta-analytic techniques to aggregate systematically and compare the findings across several studies is one of the primary advantages of this technique.
Garrett (1985) pioneered the introduction of meta-analysis into the rehabilitation literature with her exploration of the effectiveness of rehabilitation programs for juvenile offenders. Since this seminal paper, several additional meta-analyses have been conducted (Andrews, Zinger, Hoge, Bonta, Gendreau, and Cullen 1990; Cleland, Pearson, and Lipton 1996; Hill, Andrews and Hoge 1991; Lipsey 1995; Losel 1995; Pearson, Lipton, and Cleland 1997; Whitehead and Lab 1989). These reviews have provided considerable knowledge construction in the area of `what works' in offender treatment.
Andrews and his colleagues (1990) found that the clinically relevant and psychologically informed principles of human service, risk, need, and general responsivity were the strongest correlates of reduced reoffending. Meta-analyses conducted since the Andrews et al. (1990) publication have rendered considerable support for the utility of these principles in reducing offender recidivism (Andrews et al. 1990; Antonowicz and Ross 1994; Dowden and Andrews 1999a; Dowden and Andrews 1999b; Hill, Andrews and Hoge 1991; Lipsey 1995; Losel 1995).
Recently, meta-analytic evidence has also emerged supporting the applicability of these principles for diverse offender populations. More specifically, these meta-analyses have demonstrated that the principles of effective correctional treatment apply to female offenders (Dowden and Andrews 1999a), young offenders (Dowden and Andrews 1999b), as well as within community and institutional settings (Andrews et al. 1990; Dowden 1998; Hill, Andrews, and Hoge 1991). Despite these recent findings, a meta-analysis has not been conducted on the applicability of these principles in reducing violent reoffending. In fact, to our knowledge, no meta-analytic review to date has examined whether any forms of correctional treatment are effective in reducing violent recidivism.
Principles of effective correctional treatment: Risk, need, and general responsivity
The principles of risk, need, and general responsivity answer questions relating to the who, the what, and the how of correctional intervention. Although these principles have been well developed elsewhere within the correctional treatment literature (Andrews 1995; Andrews and Bonta 1998; Andrews, Bonta, and Hoge 1990), a brief overview of these principles will be provided below since they play a crucial role in the present investigation. …