Restorative Justice Meta-Analysis

7/30/2019 Restorative Justice Meta-Analysis

http://slidepdf.com/reader/full/restorative-justice-meta-analysis 1/33

S E R V I N G C A N A D I A N S

Department of J usticeCanada

Ministère d e la J usticeCanada

RESEARCH ANDSTATISTICS DIVISIONMETHODOLOGICALSERIES

The Effectiveness of Restorative Justice Practices: A Meta-Analysis




THE EFFECTIVENESS OFRESTORATIVE JUSTICE PRACTICES:

A META-ANALYSIS

Jeff LatimerSenior Research OfficerResearch and Statistics Division

Craig DowdenPrincipal ResearcherThe Action Group

&

Danielle Muise

Research AssistantResearch and Statistics Division

Research andStatistics Division

2001

The views expressed herein are solely those of the authors and do not

necessarily reflect those of the Department of Justice Canada.



T H E E F F E C T I V E N E S S O F R E S T O R A T I V E J U S T I C E P R A C T I C E S :

A M E T A - A N A L Y S I S

ii | The Effectiveness of Restorative Justice Practices: A Meta-Analysis

Foreword

Iam pleased to introduce The Effectiveness of Restorative Justice Practices: A Meta-Analysis. This is

the first in a series of publications fromthe Research and Statistics Division that

will profile innovative policy researchmethods. Through this series readers

will be introduced to some of the moreinteresting and progressive method-ological work being undertaken by the

Division. Each report in the series willaddress a justice-related policy researchquestion which has been examinedusing a different research method.

In this current report, the authors havetackled a rather difficult question — Is restorative justice an effective response to criminal behavior? — using one of the most comprehensive quantitative

About the Research and Statistics Division

The Research and Statistics Division is staffed by social science researchers drawn from a broadrange of disciplines including criminology,sociology, anthropology, education, statistics,political science, psychology, and social work.

We conduct social science research in support of

the activities and programs of the Department of Justice Canada. We also provide statistical data,methodological services and analytical advice andundertake public opinion research and compre-hensive environmental analyses.

We recognize that to be useful, research must beaccessible. In an effort to make our research moreaccessible we have created new products tailoredto the needs of a diverse group of users, such asa research series, Qs&As, fact sheets, and thismethodological series.

For further information on our researchactivities, please visit our Web site athttp://canada.justice.gc.ca/ps/rs.

research methods. Simply put, meta-analysis refers to an analysis of analyses.It is a statistical analysis of a collection

of studies for the purposes of integrating the various and, oftentimes, discrepantfindings from a body of literature. Thisreport is a good example of our efforts tosupport the Department in its evidence-based decision making process.

I would like to acknowledge the contri-bution made by Jeff Latimer, SeniorResearch Officer, in carrying out this

project from conception to final prod-uct, as well as co-authoring the report.

We would welcome any feedback onthe series.

Roberta J. Russell, Ph.D.Director, Research and Statistics DivisionDepartment of Justice Canada




The Effectiveness of Restorative Justice Practices: A Meta-Analysis | iii

Foreword ii

Acknowledgements iv

List of Tables iv

List of Figures iv

1.0 Introduction 1

2.0 Meta-Analysis 3

3.0 Method 5

3.1 L i tera tu re Rev iew:

Study Iden t i f i cat i on Cr i ter i a 5

3 .2 Da ta Co l l ect i on :

Cod i ng Procedu res 6

3 .3 Da ta Ana l ysi s:

Effect Si ze Cal cu l at i on s 7

4.0 Results 9

4.1 Vi ct im Sat i sfact i on 9

4.2 Of fender Sat i sfact i on 11

4.3 Rest i tu t i on Compl i an ce 12

4.4 Reci d i vi sm 14

5.0 Discussion 17

5.1 Sel f-Sel ect i on Bi as 17

5.2 Appropr i a te Treatm en t 18

5.3 Moderat i n g Var i abl es 18

5.4 Add i t i onal Resear ch Issues 19

6.0 Research Recommendations 21

7.0 Conclusion 23

8.0 Bibliography 25

Table of Contents





iv | The Effectiveness of Restorative Justice Practices: A Meta-Analysis

Acknowledgements

W e would like to sincerely thank the following individuals whoprovided constructive com-

ments on an earlier draft of this report:Roberta Russell, David Daubney, Danny Graham, Ivan Zinger, Kwing Hung andDaniel Antonowicz.

Further acknowledgements go to thenumerous individuals who posed

thought-provoking questions andoffered instructive comments on ourpresentation of this meta-analysis atthe “Restorative and Community Justice:Inspiring the Future” conference, in

Winchester, England on March 30, 2001.

List of Tables

Table 1.Pr i m ar y Vari ables

i n Meta-Anal ysi s 6

Tabl e 2. Descri pti ve Program /

Study Character i st i cs 9

Tabl e 3. Mod erat or Anal yses

for Vi ct im Sat i sfact i on 11

Tabl e 4.1. Mod erat or Ana lyses

for Offend er Sat i sfact ion

(Wi th Ou t l i er ) 13

Tabl e 4.2. Mod erat or Ana lyses

for Offend er Sat i sfact ion

(Wi thou t Ou t l i er ) 13

Tabl e 5. Mod erat or Anal yses

for Reci d i vi sm 15

List of FiguresFigure 1. Dist r i but i on of

Effect Si ze Esti m at es

(VICTIM SATISFACTION) 10

Figure 2. Dist r i but i on of


(OFFENDER SATI SFACTION) 12


Effect Si ze Esti m at es (RESTITUTION COMPLIANCE) 14



(RECIDIVISM ) 15




all the parties with a stake in a particularoffence come together to resolve collec-tively how to deal with the aftermath of

the offence and its implications for thefuture” (cited in Braithwaite, 1999, p. 5).

Essentially, the restorative justice para-digm begins with the premise that crimeis a violation of people and relationships(Zehr, 1990) rather than merely a viola-tion of law. The most appropriateresponse to criminal behaviour, there-fore, is to repair the harm caused by the

wrongful act (Law Commission, 2000). As such, the criminal justice systemshould provide those most closely affected by the crime (the victim, theoffender and the community) an oppor-tunity to come together to discuss theevent and attempt to arrive at someunderstanding about what can be doneto provide appropriate reparation.

According to Llewellyn and Howse(1998), the main elements of the restora-tive process involve voluntariness, truthtelling and a face-to-face encounter.Consequently, the process should becompletely voluntary for all participants;the offender needs to accept responsibil-ity for the harm and be willing to openly and honestly discuss the criminal behav-

iour; and the participants should meetin a safe and organized setting to collec-tively agree on an appropriate methodof repairing the harm.

Models of restorative justice can begrouped into three categories: circles,conferences and victim-offendermediations (VOM). While somewhatdistinct in their practices, the princi-

ples employed in each model remainsimilar.1 A restorative justice programmay be initiated at any point in the

1.0 Introduction

Current activity at governmentaland community levels suggeststhat restorative justice, in its

many forms, is emerging as an increas-ingly important element in mainstreamcriminological practice. While first dis-cussed in the 1970s by both Barnett(1977) and Eglash (1977) in the contextof restitution, restorative justice hasbeen more clearly integrated into crim-

inological thinking through such worksas Braithwaite (1989), Marshall (1985),Umbreit (1994a) and Zehr (1990). Ratherthan focussing on the traditional reha-bilitation versus retribution debate,many researchers and policy makersnow consider restorative justice and,more precisely the concept of restora-tion, as a valid third alternative (Zehr,1990). Numerous countries have adoptedrestorative approaches, including Canada, England, Australia, Scotland,New Zealand, Norway, the United States,Japan and several European countries(Hughes & Mossman, 2001).

Despite the increased attention givento restorative justice, the concept stillremains somewhat problematic to define

as numerous responses to criminalbehaviour may fall under the “restora-tive umbrella.” The term has been usedinterchangeably with such concepts ascommunity justice, transformative jus-tice, peacemaking criminology and rela-tional justice (Bazemore & Walgrave,1999). Although a universally acceptedand concise definition of the term has

yet to be established, Tony F. Marshall’s

definition appears to encompass themain principles of restorative justice:“Restorative justice is a process whereby

The Effectiveness of Restorative Justice Practices: A Meta-Analysis | 1





criminal justice system and need notbe used simply for diversionary pur-poses. Currently, there are five identified

entry points into the criminal justicesystem where offenders may be referredto a restorative justice program:

1. police (pre-charge)2. Crown (post-charge)3. courts (pre-sentence)4. corrections (post-sentence)5. parole (pre-revocation)

Proponents of restorative justice claimthat the process is beneficial to bothvictims and offenders by emphasizing recovery of the victim through redress,vindication and healing, and by encour-aging recompense by the offenderthrough reparation, fair treatment andhabilitation (Van Ness & Strong, 1997).In the process of coming together torestore relationships, the community is also provided with an opportunity toheal, reconstitute and strengthen itself through the reintegration of victims andoffenders (Llewellyn & Howse, 1998).

Despite the intuitive appeal of restora-tive justice, it is imperative to fully eval-uate the impact of this approach onseveral important outcomes. Previous

evaluation research focussing on thisarea has ranged from purely anecdotalaccounts to more rigorous designs using comparison groups and, in some cases,random assignment into control andtreatment groups (Bonta, Wallace-Capretta, & Rooney, 1998). These studieshave examined the impact of restorative

justice on victim and offender satisfac-tion, restitution compliance, recidivism,procedural fairness, as well as several

others.

Given that the field of restorative justiceresearch has been maturing, a needexisted to aggregate the present body of empirical knowledge. In this regard,several authors have recently providedcomprehensive literature reviews of thisarea of research (Braithwaite, 1999;Latimer & Kleinknecht, 2000; Marshall,

1999). Summarizing the researchthrough narrative or qualitativeapproaches, however, may fail to objec-tively analyse the available data anddraw the appropriate conclusions.Cooper and Rosenthal (1980) directly tested the reliability of synthesizing literature through narrative reviewsby providing test subjects with a set of seven studies that measured the rela-tionship between two variables. Despitethe fact that the set of studies showeda clear statistically significant relation-ship between the variables, 73 percentof the reviewers found limited or nosupport for the hypothesis. This sug-gested that traditional narrative reviewssuffer a considerable loss of power andthat the incidence of Type II errors may

be common. In addition, the criteria forselecting literature for a narrative review are rarely systematic and consistent.The introduction of meta-analytic tech-niques, however, has marked a majorstep forward in summarizing researchby providing a more objective method of aggregating knowledge.

2 | The Effectiveness of Restorative Justice Practices: A Meta-Analysis

1For a more detailed discussion of the nature and principles of restorative justice, and the core programmodels, see Restorative Justice in Canada: A Consultation Paper (May 2000) developed by the Federal/Provincial/Territorial Working Group on Restorative Justice and available from the Department of Justice Canada (http://canada.justice.gc.ca/en/ps/voc/rjpap.html).




research synthesis compared to tradi-tional narrative reviews as the formerare “more systematic, more explicit,

more exhaustive, and more quantita-tive” (Rosenthal, 1991, p.17). Meta-analytic techniques have been usedacross such diverse fields as educa-tion and medicine, and have morerecently been adopted within the socialsciences (Lipsey & Wilson, 1993). Inthe area of criminal justice research inparticular, meta-analytic studies haveinvestigated the prediction (Bonta, Law,

& Hanson, 1998; Dowden & Brown, inpress; Gendreau, Little, & Goggin, 1996;Hanson & Bussière, 1998) and treat-ment (Andrews et al., 1990; Dowden& Andrews, 1999, 2000; Latimer, 2001;Lipsey, 1995; Losel, 1995; Whitehead& Lab, 1989) of criminal behaviour.

Critics argue that one of the major limi-tations of meta-analytic techniques isthat the sampling procedures are biasedin favour of including predominantly published studies. It is surmised thatthe probability of publishing a study isincreased by the statistical significanceof the results so that published studiesare not actually representative of theentire body of research that has beenconducted in that area. Consequently,

a calculated effect size, based exclu-sively on published studies, may beoverestimating the relationship. Coinedthe “file drawer problem” (Rosenthal,1991, p. 103), this suggests that if unpublished studies were includedin the meta-analysis, the effect sizeestimate would be smaller.

Bonta, Wallace-Capretta, and Rooney

(1998) conducted a preliminary meta-analysis of programs that containedelements of restorative justice and

2.0 Meta-Analysis

Similar to traditional quantitativeresearch methods, the meta-analytic process involves three

basic steps:

1. literature review — identifying andgathering relevant research studies

2. data collection — extracting datathrough pre-determined coding procedures

3. data analysis — analysing theaggregated data using statisticaltechniques

A meta-analysis can be understood asa statistical analysis of a collection of studies that aggregates the magnitudeof a relationship between two or morevariables (Glass, McGaw, & Smith, 1981).These studies may differ on severalimportant characteristics, such as oper-ationalization of independent anddependent variables, sample size, sam-ple selection techniques and designquality. Meta-analytic statistics candescribe the typical strength of theeffect under investigation, the degreeof statistical significance, the variability,as well as provide researchers the oppor-

tunity to explore and identify potentialmoderating variables. The outcome of a meta-analysis is an effect size, whichcan be interpreted as the estimatedeffect of the independent variable onthe dependent variable. For example,an average effect size estimate of +.10translates into the independent variableaccounting for a 10 percent change inthe dependent variable (Rosenthal, 1991).

Meta-analytic reviews are generally regarded as a superior method of






exclusively focussed on their role inreducing offender recidivism. Theresults revealed that these programs

yielded mild reductions in re-offending (+.08). However, the authors used a very broad operational definition of restora-tive justice as they included both court-ordered restitution and community service programs. This definition issomewhat problematic as it fails to fully

incorporate some of the fundamentalprinciples of restorative justice —namely, the voluntary nature of both

offender and victim participation andthe face-to-face encounter. A needtherefore existed to quantitatively aggregate the findings of the literatureusing a more precise definition of restorative justice.





tion compliance and recidivism, asthese were the only ones that were suf-ficiently available to be subjected to a

meta-analysis. Furthermore, these fouroutcomes are clear and quantifiabledeterminants of the effectiveness of restorative justice.

3.1 Literature Review:Study Identification Criteria

To gather eligible studies for themeta-analysis, we conducted a com-

prehensive search of the restorative justice literature over the last 25 years.The studies were primarily drawn fromthe Internet, social science journals, andgovernmental and non-governmentalreports. A secondary search was con-ducted using the bibliographies of the identified studies and by contact-ing researchers active in the field toidentify new, unpublished and/orundiscovered research. An explicit setof criteria was established in orderto select studies for inclusion in themeta-analysis:

1. The study evaluated a restorative justice program that fell withinour working definition.

2. The study used a control group or

a comparison group that did notparticipate in a restorative justiceprogram.

3. At least one of the following fouroutcomes was reported for both thetreatment and control/comparisongroup — victim satisfaction,offender satisfaction, restitutioncompliance and/or recidivism.

4. Sufficient statistical information

was reported in order to calculatean effect size.

3.0 Method

Following the techniques of Rosenthal (1991), a meta-analysis

was designed to test the effec-tiveness of restorative justice practices.One of the major issues in conducting this form of research is agreeing ona definition of restorative justice .Generally, it is much easier to identify a non-restorative approach than itis to provide a precise definition of

what constitutes restorative justice.For the purpose of this meta-analysis,the following operational definition

was developed: restorative justice is a voluntary, community-based response to criminal behaviour that attempts to bring together the victim, the offender and the community in an effort to address the harm caused by the crimi-nal behaviour .

While this may be open to debate, anoperational definition is necessary toconduct research. Therefore, for thepresent meta-analysis, programs thatcontained “restorative” elements, suchas restitution or community service, butdid not attempt to bring together thevictim, the offender and the community,

were not considered. This definitionprovided us with a guide for the study selection process and ensured that we

were examining a consistent responseto criminal behaviour.

We also needed to identify appropriateoutcomes that were measurable andlinked directly to the goals of restora-tive justice. Although several outcome

measures have been used, we selectedvictim and offender satisfaction, restitu-






3.2 Data Collection:Coding Procedures

The standardized information con-tained in Table 1 was drawn from eachstudy using a pre-designed coding manual. In designing a coding manual,the definition of certain variables can beproblematic. For example, several stud-ies chose to operationalize recidivismdifferently. In keeping with standardmeta-analytic practice, we acceptedmultiple definitions of recidivism (i.e. a

new criminal conviction, a new criminalcharge, pre-post test offending). We alsoaccepted two definitions of restitutioncompliance (proportion of offenders

who repaid their restitution and pro-portion of total restitution dollarsrepaid by offenders).

For an overall mean effect size, in cases where multiple control/comparisongroups were used in a single study, wecombined the results to generate a sin-gle effect size for each program. In addi-tion, where multiple follow-up periods

were reported in a single study, weselected the longest at-risk period.To examine the impact of follow-uplength and the use of different control/comparison groups, we did, however,

also code multiple effect sizes for eachprogram. The results of the two coding methods will be presented separately.

Since a large proportion of programsaccepted referrals along multipleentry points, we coded both the earli-est and the latest entry points in thecriminal justice system. This providedus with two methods of conducting

analysis on the entry point of the pro-gram and its subsequent impact oneach outcome.

While we did identify those programs

that randomly assigned participantsto treatment and control groups, itshould be noted that this is somewhatmisleading. While participants areinitially assigned to either group,restorative justice participation by definition is voluntary so participantscan choose to withdraw from a program.Consequently, the problem of self-selection bias, which random assign-

ment strives to eliminate, remains asthe attrition rate in many studies wasquite high.


Table 1. Primar y Variables in Meta-Analysis

RESEARCH ARTICLE I NFORMATION

Year of the study Author(s) of the study Type of publicationCountry in which research was conducted

PROGRAM CHARACTERISTICS

Restorative justice modelEntry point in the criminal justice systemTraining, selection criteria, experience

and educational background of the mediatorEligibility criteria for offender participationExistence of training manuals or procedural

guidelines

PARTICI PANT CHARACTERISTICS

Criminal history of offendersOffence types

Age, gender and ethnicity of offenders Victim/offender relationship

OUTCOME MEASURES

Victim satisfaction ratesOffender satisfaction ratesRestitution compliance ratesRecidivism rates

METHODOLOGICAL CHARACTERISTICS

Sample sizeRandom assignment to control and treatment

groupsLength of follow-up for recidivismCharacteristics of control/comparison groupUse of an independent evaluator




3.3 Data Analysis:Effect Size Calculations

The relationship between participationin a restorative justice program andeach of the four outcomes (victim satis-faction, offender satisfaction, restitutioncompliance and recidivism) was calcu-lated from the raw statistics reported

within each study. The phi coefficient(Pearson’s r product moment correlationapplied to dichotomous data) was usedas the effect size estimate. If the neces-

sary data were not contained in anindividual study, but a non-significantrelationship between participation ina restorative justice program and theoutcome was reported, the effect size

was recorded as zero.

Once the effect sizes from each of thestudies were calculated, we conducteda series of analyses across each of ourfour outcome measures of interest. First,the overall mean effect size, along withthe corresponding confidence intervalsand standard deviation (SD), was calcu-lated. It should be noted that both the

weighted and unweighted mean effectsizes were calculated but only theunweighted estimates were used ininterpreting the results and in the

moderator analyses. This was donebecause, as stated previously, we had toestimate the actual number of victims,thus reducing the reliability of the

weighted estimates. Furthermore,the weighted mean effect sizes wereonly marginally lower or higher thanthe unweighted effect sizes and wouldnot have made a significant differenceto the results of the analysis.

We also determined whether the overalldifference between the restorative

To effectively compare victim andoffender satisfaction between restora-tive and traditional approaches, it was

necessary to create a binary satisfactionvariable. This was achieved by coding positive measures of satisfaction as“satisfied” whereas neutral and nega-tive responses were collapsed into an“unsatisfied” category. For example, if a study employed a five-point scale tomeasure satisfaction (i.e. very satisfied,somewhat satisfied, neutral, somewhatdissatisfied, very dissatisfied), we

selected the top two categories as indi-cating satisfaction and considered thelast three as unsatisfied.

In certain studies, the actual numberof victims was not indicated but thestudy reported the percentage of satis-fied versus unsatisfied victims. In thesecases, we assumed the number of vic-tims was equal to the number of offend-ers in order to calculate an effect size.In meta-analytic work, there is usually a trade-off between the comprehen-siveness of the research and the preci-sion of the coding techniques due to thereporting practices contained in moststudies.

To test the reliability of the coding

procedures, a second individual codedsix randomly selected studies contain-ing a total of 15 effect sizes. The generalrate of agreement between the codersranged from 47 percent to 100 percent,

with an overall rate of agreement of 91 percent. In cases of coder disagree-ment, both coders discussed the dis-crepancy until a consensus was reachedand this decision was then entered as

the final code. Those variables that fellbelow 80 percent agreement were notincluded in the analysis.






programs and the non-restorativecontrol/comparison groups was statis-tically significant by conducting a one-

sample t-test. This determines if themean effect size is significantly differentfrom zero (a zero effect size would indi-cate that participation in restorative

justice had no effect on the subsequentoutcomes). Additional analyses wereconducted to explore whether certain

variables, such as demographic or study characteristics, had a moderating impacton effect size magnitude. For example,

if adequate information was available, we explored whether the age of thestudy sample (adult versus youth) hada significant effect on program outcome.This provided us with a mechanism

whereby specific program impactscould be isolated for further study.





justice program resulted in higher vic-tim satisfaction ratings when comparedto a comparison group in all but one of the 13 programs examined.

It should be noted that the one negativeresult was found in the only program thatoperated at the post-sentence (or correc-tions) entry point. Compared to victims

who participated in the traditional jus-

tice system, victims who participated inrestorative processes were significantly more satisfied (t (12) = 3.89, p < 0.01).

4.0 Results

T wenty-two unique studies thatexamined the effectiveness of 35 individual restorative justice

programs generated 66 effect sizes. A summary of specific study characteris-tics is presented in Table 2. The frequen-cies presented in Table 2 are based uponthe 35 programs with the exception of the type of outcome measure and study source, which are based upon 66 effect

sizes and 22 unique studies respectively.

The vast majority of the effect sizes werederived from programs that targetedpredominantly male (94%), young (74%)offenders. Interestingly, a large propor-tion of the effect sizes was drawn fromstudies that were not published in peer-reviewed academic journals (55%),

which, as discussed previously, is typi-cally not the case in meta-analytic work.

As shown in Table 2, studies commonly included one or more of the following outcome measures: victim satisfaction,offender satisfaction, restitution com-pliance and recidivism reduction. Eachof these issues will be discussed accord-ingly in the following subsections.

4.1 Victim Satisfaction

The overall mean effect size for the13 tests of treatment that explored theimpact of restorative justice program-ming on victim satisfaction was +0.19(SD=.18) with a 95 percent confidenceof +0.30 to +0.08 (see Figure 1). Althoughthe effect sizes ranged from +0.44 to

-0.19, the latter was the only negativevalue found in the distribution. In other

words, participation in a restorative


Table 2. Descriptive Program/

Study Characteristics

VARIABLE FREQUENCY (%)

Resto ra ti ve Justi ce Model

Conferencing 8 (22.9) Victim-offender mediation 27 (77.1)

Entr y Point

Pre-charge 7 (20.0)Post-charge 6 (17.1)Pre-sentence 1 ( 2.9)Post-sentence 1 ( 2.9)Mixed 20 (57.1)

Outcome Measur e Victim satisfaction 13 (19.7)Offender satisfaction 13 (19.7)Restitution compliance 8 (12.1)Recidivism 32 (48.5)

Gender

Predominantly male (>70%) 33 (94.3)Mixed 2 ( 5.7)

Ethnic i ty

Predominantly Caucasian (>70%) 14 (40.0)

Other 2 ( 5.7)Mixed/unspecified 19 (54.3)

Age Grou p

Adult 9 (25.7) Youth 26 (74.3)

Stu dy Sou rce

Published 10 (45.5)Unpublished 12 (54.5)



Given the relatively wide range of effectsizes, additional analyses were con-ducted to explore whether characteris-tics of the study sample or methodologi-cal considerations could explain thisvariability. Initially, we had hoped toexplore a relatively large number of potential moderators, such as gender,ethnicity, criminal history, offencetype, etc. The relative homogeneity

of the offenders in the studies, how-ever, as well as the large amount of missing data, rendered many of theseanalyses untenable. On the other hand,this homogeneity increases our confi-dence in the generalizability of the find-ings to this population. Therefore, theanalyses focus on six factors: randomassignment, offender age, publicationsource, restorative justice model, entry

point and control/comparison grouptype.

As indicated previously, we used two coding methods for capturing

information on thecontrol/comparisongroup used in the studies.First, we combined multi-ple control/comparisongroups from the samestudy to calculate a singleeffect size; second, wecalculated individualeffect sizes for each

control/comparisongroup. For the compar-ison of control groups inTable 3 (and subsequentTables), we used the lattercoding technique. Thisallowed us to comparerestorative justice pro-grams with individuals

who were referred to a restorative justiceprogram but refused participation ver-sus all other control group types (i.e.probation, court, prison).

The mean effect sizes for each valueof the moderator variable, along withtheir corresponding significance tests,are presented in Table 3. Althoughthese variables did not yield significant

between-group differences, studiesusing non-randomized comparisonsand studies published in academic jour-nals displayed a higher mean effect sizethan their counterparts. In addition,

VOM models tended to display highervictim satisfaction rates than confer-encing models when compared to thenon-restorative approaches. The lack of significance between moderating

variables might be due to the low num-ber of effect sizes.




0 1 2 3 4 5 6 7 8

.40 to .49

.30 to .39

.20 to .29

.10 to .19

0.0 to 0.9

-.10 to -.01

-.20 to -.11

-.30 to -.21

E f f e c t s i z e

Number of effect sizes

Figure 1. Distribution of Effect Size Estimates(VICTIM SATISFACTION)




two negative effect sizes contributing tothis result, the -0.71 was a clear outlier.Moreover, given that the sample sizeused in this outlier study was extremely small (n=7), we removed the study fromthe analysis. This increased the meaneffect size to +0.17 and substantially reduced the standard deviation (SD=.13).Furthermore, and more importantly,

removal of this study resulted in theconfidence interval not including zero,thus suggesting that these programshave a moderate to weak positiveimpact on offender satisfaction. Thedifference in offender satisfactionbetween restorative and non-restorativeparticipation also becomes significant(t (11) = 4.52, p < 0.01). Interestingly, the-.71 effect size was drawn from the same

post-sentence entry point program asthe only negative victim satisfactioneffect size.

4.2 Offender Satisfaction

The overall mean effect size for the13 tests of the impact of restorative

justice programming on offender satis-faction was +0.10 (SD=.28), while theeffect sizes ranged from +0.31 to -0.71(see Figure 2). While offenders whoparticipated in restorative justice pro-

grams displayed higher satisfaction with the process than their compar-isons, the one-sample t-test indicatedthat this difference was not statistically significant. Since the 95 percent con-fidence interval included zero, thisfurther decreased our confidencethat these programs have had any discernible impact on offendersatisfaction.

This conclusion is mitigated, however,by the finding that although there were


Table 3. Moderator Analyses for Victim Satisfaction

EFFECT SIZE T value

VARIABLE N Unweighted (p)

AGE Youth 8 .20 -.09 Adult 5 .19 (ns)

RANDOM ASSIGNM ENT Yes 3 .14 .94No 10 .21 (ns)

STUDY SOURCE Published 3 .30 -1.42Unpublished 10 .16 (ns)

ENTRY POINT (ear li est) Pre-charge 6 .16 -.31Other entry points 6 .18 (ns)

ENTRY POINT (l at est) Pre-charge 3 .15 -.32Other entry points 9 .18 (ns)

MODEL Conferencing 4 .14 .94 V-O mediation 9 .21 (ns)

CONTROL GROUP Non-participation 9 .22 .62Other control 6 .18 (ns)

v-o = victim-offender

ns = not significant



pensation for the harmcaused by the criminalactivity and the offenders

would be actively accept-ing responsibility. Theresults of the studies thatincluded a measure of restitution complianceare reported below.

Only eight studies exam-ined the impact of restora-tive justice programming

on restitution compliance. Although this numbermay seem small, it may have been, in part, dueto the inclusion criteriafor this meta-analysis(i.e. the study used a com-parison group). Overall,the mean effect size of +0.33 (SD=.24) was quite

high, indicating that offenders whoparticipated in restorative justice pro-grams tended to have substantially higher compliance rates than offend-ers exposed to other arrangements.Furthermore, there was a great deal of variability in the effect sizes found inthese studies, with values ranging from+0.63 to -0.02 (see Figure 3). Compared

to the comparison/control groups notparticipating in a restorative justiceprogram, offenders in the treatmentgroups were significantly more likely to complete restitution agreements(t (7) = 3.87, p < 0.01).

The small number of effect sizes (k=8)made conducting the moderator analy-ses inappropriate.

To account for this substantial discrep-ancy, we presented the results both withthe outlier ( Table 4.1) and without theoutlier (Table 4.2). Given this extremeoutlier, interpreting these results wasinappropriate as the conclusions wouldbe drastically different in each casebased upon the inclusion or exclusionof one value.

4.3 Restitution Compliance

One of the potential advantages of arestorative justice approach is that itcould be more effective in ensuring offender compliance with restitutionagreements. This would be a significantcontribution as the victims would havea greater likelihood of receiving com-




0 1 2 3 4 5 6 7 8

.40 to .49

.30 to .39

.20 to .29

.10 to .19

0.0 to 0.9

-.10 to -.01

-.20 to -.11

-.30 to -.21

E f f e c t s i z e


*Does not include -.71 outlier

Figure 2. Distribution of Effect Size Estimates(OFFENDER SATISFACTION*)





Table 4.1. Moderator Analyses for Offender Satisfaction (With Outlier)

EFFECT SIZE T Value

VARIABLE N Unweighted (p)AGE Youth 8 .15 -.47

Adult 5 .05 (ns)


STUDY SOURCE Published 1 .08 .00Unpublished 12 .11 (ns)

ENTRY POINT (ear li est) Pre-charge 8 .15 .17Other entry points 5 .03 (ns)



CONTROL GROUP Non-participation 8 .10 .02Other control 7 .10 (ns)


s = not significant

Table 4.2. Moderator Analyses for Offender Satisfaction (Without Outlier)

EFFECT SIZE T Value


AGE Youth 8 .15 1.71 Adult 4 .22 (ns)

RANDOM ASSIGNM ENT Yes 4 .09 1.33No 9 .21 (ns)

STUDY SOURCE Published 1 .08 .00

Unpublished 11 .18 (ns)

ENTRY POINT (ear li est) Pre-charge 8 .15 -1.02Other entry points 4 .22 (ns)

ENTRY POINT (l at est) Pre-charge 5 .09 -1.79Other entry points 7 .23 (ns)


CONTROL GROUP Non-participation 7 .21 1.93Other control 7 .10 (ns)







4.4 Recidivism

Arguably, one of the most importantoutcome variables for any form of crimi-nal justice intervention is recidivism.Considerable public and institutionalsupport for correctional programming

rests on its ability to reduce future crim-inal activity. Therefore, the ability of restorative justice programs to reducerecidivism was felt to be particularly important for this meta-analysis.

The overall mean effect size for the32 tests that examined the effectivenessof restorative justice programming inreducing offender recidivism was +0.07

(SD=.13) with a 95 percent confidenceinterval of +0.12 to +0.02. Although theeffect sizes ranged from +0.38 to -0.23,

more than two thirdsof the effect sizes werepositive (72%). In other

words, restorative justiceprograms, on average,

yielded reductionsin recidivism comparedto non-restorativeapproaches to criminalbehaviour. In fact, com-pared to the comparison/control groups that didnot participate in a

restorative justice pro-gram, offenders in thetreatment groups weresignificantly more suc-cessful during the follow-up periods (t (31) = 2.88,p < 0.01).

One of the major areas of debate in the correctionaltreatment literature isthe impact of different

methodological and demographiccharacteristics on program effective-ness. Subsequently, we conducted mod-erator analyses to explore the impactsof several variables on recidivism reduc-tion. The results of these analyses arepresented in Table 5 and are discussed

below.

As stated previously, one of the primary criticisms lodged against meta-analysisis its predominant reliance on publishedstudies and the subsequent problemof potential publication bias. This issuehas been addressed in the present meta-analysis by conducting searches of governmental and non-governmental

reports, graduate theses and disser-tations and by directly contacting researchers active in the field for


0 1 2 3 4 5 6 7 8

.40 to .49

.30 to .39

.20 to .29

.10 to .19

0.0 to 0.9

-.10 to -.01

-.20 to -.11

-.30 to -.21

E

f f e c t s i z e


.50 to .59

.60 to .69

Figure 3. Distribution of Effect Size Estimates(RESTITUTION COMPLIANCE)





Table 5: Moderator Analyses for Recidivism

EFFECT SIZE T value


AGE Youth 24 .06 .60 Adult 8 .10 (ns)


STUDY SOURCE Published 12 .12 -1.73

Unpublished 20 .04 (ns)

ENTRY POINT (ear li est) Pre-charge 16 .07 .17Other entry points 16 .06 (ns)



CONTROL TYPE Non-participation 9 .02 -1.73Other control 31 .12 (ns)



0 1 2 3 4 5 6 7 8 9 10 11 12

.40 to .49

.30 to .39

.20 to .29

.10 to .19

0.0 to 0.9

-.10 to -.01

-.20 to -.11

-.30 to -.21

E f f e c t s i z e


Figure 4. Distribution of Effect Size Estimates(RECIDIVISM)





unpublished research. Nevertheless, wedirectly tested the impact of publicationsource on effect size. Inspection of

Table 5 reveals that the mean effect sizefor studies from published sources wassomewhat higher than the mean effect

size found in unpublished sources. This,in combination with the differencereported in victim satisfaction rates,

lends support to the “file-drawer”problem in meta-analytic work.





argued that there was no treatmenteffect on recidivism from participationin restorative justice beyond a self-

selection effect.

This is an inherent problem in restora-tive justice research. It is not possibleto truly randomly assign participants totreatment and control conditions. Oncean individual is forced to participate in arestorative justice program, most wouldargue that the program is no longer truly restorative. Given this, we believe that

an alternative method of determining the effectiveness of restorative justice isnecessary. We recommend administer-ing questionnaires designed to measureparticipants’ motivation prior to pro-gram participation. This would allow researchers to examine the motivationof the control group, restorative justiceparticipants and those who refused par-ticipation. A research design such as this

would provide a comparison of highly motivated, moderately motivated andunmotivated individuals in each group.If the satisfaction or recidivism rates, forexample, were improved in the restora-tive justice group, and motivation

was controlled for in the analysis, we would be more convinced that there isa treatment effect from participation

in restorative justice processes.

Notwithstanding the issue of self-selection bias, the results of thismeta-analysis, at present, representthe best indicator of the effectivenessof restorative justice practices (i.e. thoseindividuals who choose to participatein restorative justice programs findthe process satisfying, tend to display

lower recidivism rates and are morelikely to adhere to restitutionagreements).

5.0 Discussion

Generally, compared to traditionalnon-restorative approaches,restorative justice was found to

be more successful at achieving eachof its four major goals. In other words,based on the findings of this meta-analysis, restorative justice programsare a more effective method of improv-ing victim/offender satisfaction, increas-ing offender compliance with restitu-

tion, and decreasing the recidivismof offenders when compared to moretraditional criminal justice responses(i.e. incarceration, probation, court-ordered restitution). In fact, restorativeprograms were significantly more effec-tive than these approaches in all fouroutcomes (with the exclusion of theoffender satisfaction outlier).

5.1 Self-Selection Bias

The positive results of this meta-analysisare tempered, however, by the self-selection bias evident in controlledoutcome studies on restorative justiceprograms. Restorative justice, by its very nature, is a voluntary process. This cre-ates a treatment group of participants

(both offenders and victims) who havechosen to participate in the programand may therefore be more motivatedthan the control group. This concern iselevated by the high rate of attrition

within many of the studies in this meta-analysis. McCold and Wachtel (1998),for example, found clear differences inthe recidivism rates of restorative justiceparticipants (20%) versus individuals

who refused participation in the pro-gram (48%) versus the comparisongroup (35%). In fact, these authors






5.2 Appropriate Treatment

While the effects of restorative justice

participation on recidivism remainssomewhat uncertain due to the self-selection bias, many argue that it isnaïve to believe that a time-limitedintervention such as a VOM will have adramatic effect on altering criminal anddelinquent behaviour (Umbreit, 1994b).

Additional factors, such as anti-socialpeers, substance abuse and crimino-genic communities, which have been

linked to criminal behaviour (Hawkinset al., 1998; Lipsey & Derzon, 1998), arenot adequately addressed in the restora-tive process. Andrews and Bonta (1998)have also identified several crimino-genic needs that they maintain areimperative to address in the treatmentof offenders in order to reduce recidi-vism effectively. Generally, they identi-fied anti-social attitudes, pro-criminalassociates, personality factors, family factors and low levels of educational andemployment attainment. In fact, previ-ous meta-analytic work conductedby Dowden (1998) and Andrews et al.(1990) found that “appropriate” correc-tional treatment (i.e. those programsthat adhered to the clinically relevantprinciples of risk, need and responsiv-

ity 2

) displayed an appreciably highermean effect size (+.26 and +.30, respec-tively) for recidivism compared to thefindings for restorative justice programs(+.07) presented here. In other words,although restorative justice programsmay yield reductions in recidivismcompared to more traditional criminal

justice responses to crime, they did

not have nearly as strong an impacton re-offending as psychologically informed treatment.

It has been argued, however, thatrestorative justice and rehabilitativetreatment are rather complementary approaches (Crowe, 1998). The utiliza-tion, therefore, of both restorative jus-tice processes and “appropriate treat-ment” as a comprehensive response tocriminal behaviour would be a valuableand theoretically directed experiment.

This combination would enable bothapproaches to capitalize on theirstrengths and minimize their weak-nesses. More specifically, the restora-tive processes could increase victim/offender satisfaction and restitutioncompliance while the rehabilitativeprocesses could have a significantimpact on recidivism.

5.3 Moderating Variables

Although we did not discover any signif-icant differences in reported outcomesbased on model type (i.e.VOM versusconferencing), VOM models did havehigher victim satisfaction ratings andoffender satisfaction ratings (excluding the outlier). The low number of effect

sizes, however, coupled with the lack of a significant difference, does not allow for a firm conclusion. Theoretically, thelarge number of participants in a con-ference compared to a VOM could becontributing to the reported lowervictim/offender satisfaction ratings.There is concern in the restorative

justice literature that coming to a


2For a detailed description of the principles of risk, need and responsivity and their importance ineffective correctional treatment, see Andrews and Bonta (1998).




the criminal history of the offenders(i.e. first-time offenders versus repeatoffenders), the specific offences

(i.e. minor versus serious offences,property versus violent offences),and the relationship between offendersand victims (i.e. family, neighbour,stranger).

In general, we were unable to provide anadequate explanation for the large rangeof reported effect sizes in each of theoutcomes. It is possible that the signifi-

cant factors in determining a more suc-cessful restorative justice program arethose that were not reported in theliterature (i.e. facilitator characteristics,offence types, criminal history). Tofacilitate a better understanding of theeffectiveness of restorative justice, werecommend that future studies reportoutcomes, such as recidivism or satis-faction, separately for groups of offend-ers using such variables as gender, age,criminal history, offence types and rela-tionship between victim and offender.In addition, we recommend that studiesprovide more detailed information onthe processes used within the restora-tive justice programs and the facilitators.

5.4 Additional Research Issues

Another issue that future studies may wish to explore is the effect that offendercompliance with restitution agreementshas on victim satisfaction. The restric-tions of meta-analytic procedures pre-clude such an analysis. Morris andMaxwell (1998), however, did reportthat the reason most frequently reportedfor victim dissatisfaction in an evalua-

tion of a family group conference pro-gram in New Zealand was a failure toreceive the appropriate restitution.

satisfactory agreement in a session issometimes difficult, particularly whenthere are numerous individuals from

various backgrounds participating inthe process (Hooper & Busch, 1996).This possibility is certainly increasedin the case of conferences.

Interestingly, we did not discover differ-ences among programs that operatealong different entry points to the crimi-nal justice system, with the exception of offender satisfaction. This difference is

not convincing, however, because of theoutlier issue and the fact that the entry points were difficult to code (more thanhalf of the programs had mixed entry points). Thus, we recommend thatresearchers more explicitly code theentry point for their program and/orseparate the analyses by this variable.

There were several questions that we were unable to answer due to a lack of data reported in the literature. Forexample, we were interested in explor-ing whether the characteristics of thefacilitator had a significant moderating impact on restorative justice programeffectiveness. However, very few, if any of the studies, provided informationconcerning the education, professional

background or training of the facilita-tors. This is particularly noteworthy asfacilitators within restorative justiceprograms can have a significant impacton the outcome of a session. Support forthis assertion may be found within thecorrectional treatment literature whereprogram staff characteristics and behav-iours have been found to have a signifi-cant impact on program effectiveness

(Dowden & Andrews, under review).There was also rather limited data onadditional important variables, such as






More empirical research into the restitu-tion conditions (i.e. type of restitution,size of restitution, length of time given

to comply) that lead to successful com-pliance would also be appropriate.Moreover, the same type of analysiscould be completed on restitution con-ditions and victim/offender satisfaction.

And finally, there is no research in theliterature that examines the longer termeffects for victims who participate in a

restorative justice process. An exami-nation of whether victims still feel thatthey have experienced some closure andhealing six months or a year after therestorative process would be beneficial.





■ more specific reporting practices when presenting outcomes (i.e. by age groups, gender, criminal history,

offence type);■ more detailed information in

research reports/articles on theprocesses involved in the programsand program facilitators;

■ the effectiveness of a combination of restorative justice and “appropriatetreatment” approaches;

■ the effectiveness of restorative justice for female offenders;

■ controlled evaluations of circle sen-tencing models and healing circles;

■ the relationship between restitutionand satisfaction; and

■ follow-up research into the longerterm effects on victims.

6.0 Research

Recommendations

Surprisingly, given the current levelof restorative justice activity inCanada, there were only three

Canadian studies and no appropriateempirical evaluations of circle sentenc-ing models or healing circles. This isdue, in some part, to our selectioncriteria, which required the use of a

control/comparison group but alsobecause the field of restorative justiceresearch in Canada is not as well devel-oped as it is in other countries. In addi-tion, as with a large proportion of crimi-nal justice research, there was a dearthof information on the effectiveness of restorative justice for female offenders.

As such, we are recommending thatfuture Canadian research focus on thefollowing issues:

■ more evaluations of restorative justice programs using randomly assigned treatment and controlconditions with an examinationof restorative justice participants’motivation in order to address theinherent self-selection bias;









not as strong as “appropriate correc-tional treatment,” restorative justicedoes appear to reduce recidivism for

those who choose to participate. Finally,offenders in restorative justice programsreport moderate increases in satisfac-tion compared to offenders in the tradi-tional system.

The proliferation of restorative justiceprogramming worldwide is, therefore,not surprising. Both the theory andempirical research tend to offer sup-

port for such a response to criminalbehaviour. The next critical step forboth research and program develop-ment is to obtain a better understand-ing of the effect of self-selection biasthat currently diminishes our confi-dence in these results. To more defini-tively claim restorative justice as aneffective response to criminal behav-iour, we need to be able to address thislimitation inherent in restorative justiceresearch methods.

7.0 Conclusion

The traditional criminal justicesystem, which has been oftencriticized as too formal, punitive

and adversarial, is clearly changing.The large increase in the number of restorative justice programs operating in Canada is undoubtedly having animpact on criminal justice theory andpractice. We are currently in a periodof substantial change; but, as the results

of this meta-analysis indicate, we aremoving in a positive direction. Theaddition of restorative justice programshas enhanced victim satisfaction in aprocess that was, by its very nature,rather unsatisfactory. Moreover, thisresponse to criminal behaviour has astrong impact by encouraging moreoffenders to take responsibility for theiractions and repair through restitutionsome of the harm they have caused. And

while the gains made in recidivism are









Crowe, A.H. (1998). Restorative justice andoffender rehabilitation: A meeting of theminds. Perspectives: Journal of the American Probation and Parole Association 22: 28–41.

*Davis, R.C. (1982). Mediation: The Brooklynexperiment. In R. Tomasic and M. Feeley (eds.), Neighbourhood Justice: Assessment of an Emerging Idea. New York: Longman Inc.

Dowden, C. (1998). “A Meta-analyticExamination of the Risk, Need andResponsivity Principles and TheirImportance within the RehabilitationDebate.” Unpublished M.A. thesis. Ottawa:

Psychology Department, Carleton University.Dowden, C., and Andrews, D.A. (1999). What

works for female offenders: A meta-analyticreview. Crime and Delinquency 45: 438–452.

___________________________________________________ . (2000). Effective correctional treat-ment and violent re-offending: A meta-analysis. Canadian Journal of Criminology 42: 449–476.

___________________________________________________ . (under review). “The Importance of Staff Characteristics in Delivering EffectiveCorrectional Treatment: A Meta-analyticReview of Core Correctional Practice.”Manuscript submitted for publication.

Dowden, C., and Brown, S. (in press). Therole of substance abuse factors in predicting recidivism: A meta-analysis. Psychology,Crime, and Law.

Eglash, A. (1977). Beyond restitution:Creative restitution. In J. Hudson and

B. Galaway (eds.), Restitution in Criminal Justice. Lexington, Mass.: D.C. Heath.

*Evje, A., and Cushman, R.C. (2000). A Summary of the Evaluations of Six California Victim Offender Reconciliation Programs. The Judicial Council of California

Administrative Office of the Courts, Centerfor Families, Children & the Courts.

Federal/Provincial/Territorial Working Group on Restorative Justice. (2000).

Restorative Justice in Canada: A Consultation Paper. Ottawa: Department of Justice,Canada.

8.0 Bibliography

*Indicates study used in meta-analysis Andrews, D.A., and Bonta, J. (1998).Psychology of Criminal Conduct (2nd ed.).Cincinnati, Oh.: Anderson Publishing.

Andrews, D.A., Zinger, I., Hoge, R., Bonta, J.,Gendreau, P., and Cullen, F. (1990). Doescorrectional treatment work? A clinically relevant and psychologically informedmeta-analysis. Criminology 28: 369–404.

Barnett, R. (1977). Restitution: A new para-digm of criminal justice. Ethics 87: 279–301.

Bazemore, G., and Schiff, M.F. (1996).Community justice/Restorative justice:Prospects for a new social ecology for com-munity corrections. International Journal of Comparative and Applied Criminal Justice 20: 311–335.

Bazemore, G., and Walgrave, L. (1999).Restorative Juvenile Justice: Repairing the

Harm of Youth Crime. Monsey, N.Y.: CriminalJustice Press.

Bonta, J., Law, M., and Hanson, R.K. (1998).The prediction of criminal and violentrecidivism among mentally disorderedoffenders: A meta-analysis. Psychological Bulletin 123: 123–142.

*Bonta, J., Wallace-Capretta, S., and Rooney,J. (1998). Restorative Justice: An Evaluation of the Restorative Resolutions Project.

Ottawa: Solicitor General Canada.

Braithwaite, J. (1989). Crime, Shame and Reintegration. Cambridge: CambridgeUniversity Press.

___________________________________________________ . (1999). Restorative justice: Assessing optimistic and pessimistic accounts. InM. Tonry (ed.), Crime and Justice: A Review of Research 25: 1–127.

Cooper, H., and Rosenthal, R. (1980). Statisti-

cal versus traditional procedures for sum-marising research findings. Psychological Bulletin 87: 442–449.






*Fercello, C., and Umbreit, M. (1998). Client Evaluation of Family Group Conferencing in 12 Sites in 1st Judicial District of Minnesota.Center for Restorative Justice & Mediation,

School of Social Work, University of Minnesota.

Gendreau, P., Goggin, C., and Cullen, F.(1999). The Effects of Prison Sentences on Recidivism. Ottawa: Corrections ResearchBranch, Solicitor General of Canada.

Gendreau, P., Little, T., and Goggin, C. (1996). A meta-analysis of the predictors of adultoffender recidivism: What works! Criminol-

ogy 34: 575–607.Glass, G., McGaw, B., and Smith, M. (1981).Meta-analysis in Social Research. Newbury Park, Calif.: Sage.

Hanson, R.K., and Bussière, M. (1998).Predicting relapse: A meta-analysis of sexualoffender recidivism studies. Journal of Consulting and Clinical Psychology 66:348–362.

Hawkins, J.D., Herrenkohl, T., Farrington,D.P., Brewser, D., Catalano, R.F., and Harachi,T.W. (1998). A review of predictors of youthviolence. In R. Loeber and D.P. Farrington(eds.), Serious and Violent Juvenile Offenders: Risk Factors and Successful Interventions.London: Sage.

Hooper, S., and Busch, R. (1996). Domesticviolence and restorative justice initiatives:The risks of a new panacea. Waikato Law Review, Special Issue: Domestic Violence 4: 1–19.

Hughes, P., and Mossman, M.J. (2001).“Re-thinking Access to Criminal Justicein Canada: A Critical Review of Needs andResponses.” Unpublished paper for theDepartment of Justice Canada.

LaPrairie, C. (1995). Altering course: New directions in criminal justice — Sentencing circles and family group conferences.

Australian and New Zealand Journal of Criminology, Special Issue: Crime,Criminology and Public Policy : 78–99.

Latimer, J. (2001). A meta-analytic exami-nation of youth delinquency, family treat-ment and recidivism. Canadian Journal of Criminology 43: 237–253.

Latimer, J., and Kleinknecht, S. (2000). The Effects of Restorative Justice Programming: A Review of the Empirical Research Literature.Ottawa: Research and Statistics Division,Department of Justice Canada.

Law Commission. (2000). From Restorative Justice to Transformative Justice: Discussion Paper. Ottawa: Law Commission of Canada.

Lipsey, M. (1995). What do we learn from

400 research studies on the effectivenessof treatment with juvenile delinquents?In J. McGuire (ed.), What Works: Reducing Reoffending — Guidelines from Research and Practice. Chichester, Eng.: John Wiley & Sons.

Lipsey, M., and Derzon, J. (1998). Predictorsof violent or serious delinquency in adoles-cence and early adulthood: A synthesis of longitudinal research. In R. Loeber andD.P. Farrington (eds.), Serious and Violent

Juvenile Offenders: Risk Factors and Successful Interventions. London: Sage.

Lipsey, M., and Wilson, D. (1993). Theefficacy of psychological, educational,and behavioral treatment: Confirmationfrom meta-analysis. American Psychologist 48: 1181–1209.

Llewellyn, J., and Howse, R. (1998). Restora-tive Justice: A Conceptual Framework. Ottawa:Law Commission of Canada.

Losel, F. (1995). The efficacy of correctionaltreatment: A review and synthesis of meta-evaluations. In J. McGuire (ed.), What Works: Reducing Re-offending — Guidelines from Research and Practice. Chichester, Eng.:John Wiley & Sons.

Marshall, T. (1985). Alternatives to Criminal Courts: The Potential for Non-judicial Dispute Resolution. Brookfield, Ver.: Gower

Publishing Co.





Rosenthal, R. (1991). Meta-analytic Procedures for Social Research. Newbury,Calif.

*Roy, S. (1993). Two types of juvenile restitu-tion programs in two Midwestern counties:

A comparative study. Federal Probation 57: 48–53.

*Schneider, A.L. (1986). Restitution andrecidivism rates of juvenile offenders:Results from four experimental studies.Criminology 24: 533–552.

*Sherman, L.W., and Strang, H. (2000).Recidivism Patterns in the Canberra

Reintegrative Shaming Experiments (RISE).Centre for Restorative Justice, ResearchSchool of Social Sciences, AustralianNational University.

*Strang, H., Barnes, G.C., Braithwaite, J.,and Sherman, L.W. (1999). Experiments in Restorative Policing: A Progress Report on the Canberra Reintegrative Shaming Experiments (RISE). Australian Federal Police and

Australian National University.

*Umbreit, M.S. (1989). Crime victims seeking fairness, not revenge: Toward restorative

justice. Federal Probation 53: 52–57.

___________________________________________________ . (1994a). Victim Meets Offender: The Impact of Restorative Justice and Mediation.Monsey, N.Y.: Criminal Justice Press.

*___________________________________________________ . (1994b). Crime victims confronttheir offenders: The impact of a Minneapolismediation program. Research on Social Work

Practice 4: 436–447.

*Umbreit, M.S., and Coates, R.C. (1993).Cross-site analysis of victim-offender media-tion in four states. Crime & Delinquency 39: 565–585.

*Umbreit, M.S., Coates, R.C., Kalanj, B.,Lipkin, R., and Petros, G. (1995). Mediation of Criminal Conflict: An Assessment of Programs in Four Canadian Provinces — Executive Summary Report. Center for

Restorative Justice & Mediation, Schoolof Social Work, University of Minnesota.

___________________________________________________ . (1999). Restorative Justice: An Overview. London: Home Office, ResearchDevelopment and Statistics Directorate.

*McCold, P., and Stahr, J. (1996). “BethlehemPolice Family Group Conferencing Project.”Paper presented to the American Society of Criminology Annual Meeting, Chicago,November 20–23, 1996.

*McCold, P., and Wachtel, B. (1998).Restorative Policing Experiment: The Bethlehem Pennsylvania Police Family Group Conferencing Project — Summary.Community Service Foundation, Pipersville,

Penn.*Moore, D., and Forsythe, L. (1995). A New

Approach to Juvenile Justice: An Evaluation of Family Conferencing in Wagga Wagga. A report to the Criminology Research Council,Centre for Rural Social Research, CharlesSturt University, Wagga Wagga NSW,

Australia.

Morris, A., and Maxwell, G. (1998).Restorative justice in New Zealand: Family

group conferences as a case study. Western Criminology Review 1: 1–16.

*Niemeyer, M., and Shichor, D. (1996). A pre-liminary study of a large victim/offenderreconciliation program. Federal Probation 60: 30–34.

*Nuffield, J. (1997). Final Report — Evaluation of the Adult Victim-Offender Mediation Program, Saskatoon Community Mediation Services. Saskatoon, Sask.:Saskatchewan Justice.

*Nugent, W.R., and Paddock, J.B. (1995).The effect of victim-offender mediation onseverity of re-offence. Mediation Quarterly 12: 353–367.

*___________________________________________________ . (1996). Evaluating the effects of avictim-offender reconciliation program onre-offence. Research on Social Work Practice 6: 155–178.






*Umbreit, M., and Fercello, C. (1997). Interim Report: Client Evaluation of the Victim/ Offender Conferencing Program in Washington County (MN). Center for

Restorative Justice & Mediation, Schoolof Social Work, University of Minnesota.

*Umbreit, M.S., Warner Roberts, A., Kalanj,B., and Lipkin, R. (1996). Mediation of Criminal Conflict in England: An Assessment of Services in Coventry and Leeds — Executive Summary. Center for Restorative Justice &Mediation, School of Social Work, University of Minnesota.

*URSA Institute. (1993). Final Evaluation Report: Community Involvement in Mediation of First and Second Time Juvenile Offenders Project of the Community Board Program of San Francisco. Washington, D.C.:U.S. Department of Justice.

Van Ness, D., and Strong, K.H. (1997).Restoring Justice. Cincinnati, Oh.: AndersonPublishing.

Whitehead, J.T., and Lab, S.P. (1989). A meta-analysis of juvenile correctional treat-ment. Journal of Research on Crime and Delinquency 26: 276–295.

*Wynne, J. (1996). Leeds mediation and repa-ration service: Ten years’ experience withvictim-offender mediation. In B. Galaway and J. Hudson (eds.), Restorative Justice: International Perspectives. Monsey, N.Y.:Criminal Justice Press.

Zehr, H. (1990). Changing Lenses: A New Focus for Crime and Justice. Scottsdale,Penn.: Herald Press.

Zellerer, E. (1996). Community-based justiceand violence against women: Issues of gen-der and race. International Journal of Comparative and Applied Criminal Justice 20: 233–244.

Restorative Justice Meta-Analysis

Documents