rapid eviden e assessment (rea) of …...at the abstract review stage. using this tool, we reviewed...

1

RAPID EVIDENCE ASSESSMENT (REA) OF

‘OFFENDER LEARNING’ IN PRISONS –

FIRST DRAFT FOR DISCUSSION

THE MANCHESTER COLLEGE

18TH MARCH 2016

3

EXECUTIVE SUMMARY

- This is a Rapid Evidence Assessment of offender learning. Rapid Evidence Assessments follow a similar

process to Systematic Reviews: collating descriptive outlines of the available evidence on a topic;

critically appraising the studies identified; sifting out studies of poor quality: and providing an overview

of what that evidence tells us, and what is missing from it. However, REAs are not as comprehensive

as systematic reviews and provide a summary of evidence.

- Educational deficits among prisoners are widely reported. However, recent studies demonstrate that

the popular narrative is not necessarily accurate with prisoners actually outperforming the public up to

Level 1 in numeracy. However, they do consistently fall behind the general population at Level 2

Numeracy and in all levels of literacy.

- Included in this review are papers relating to educational programmes delivered within prison. These

include vocational, academic, basic skills, accredited and unaccredited courses. Excluded are courses

delivered outside of a traditional prison (e.g. halfway houses, community centres, ‘camps’ for

offenders), life skills programmes, work place training (unless specifically towards a qualification) and

psychologically based thinking skills programming.

- Studies were included if they had clear outcome measures for recidivism or employment. However,

these measures vary between studies. There is no clear consensus on the definition of ‘recidivism’ or

‘employment’.

- Previous meta-analyses generally indicate a positive impact of educational participation by reducing

recidivism and increasing employment. However, some previous meta-analyses question the impact

of vocational education (Wilson et al 2000, Hurry et al, 2006) the general consensus seems to be that

education is effective. There is some indication in this literature that the higher the qualification, the

more impact it has; degree level qualifications are the most effective at reducing recidivism. However,

the evidence is not conclusive.

- Initial bibliographic searches returned 4,304 abstracts. Two reviewers then reviewed the abstracts of

these titles and marked them as ‘met’, ‘partially met’ or ‘not met’. Inter-rater reliability was checked

with 2,682 (62%) of these abstracts being reviewed by both reviewers. From these, initially 35 studies

met the criteria. Grey literature searches revealed a further seven studies taking the total to 42 studies

whilst hand searching specific journal revealed no further studies for inclusion. The appraisal of key

systematic reviews revealed a further 11 studies, the majority of which were from the grey literature.

- Two reviewers piloted and refined a data capture tool that provided an overview of the papers included

at the abstract review stage. Using this tool, we reviewed 52 studies and allowed for a more in-depth

analysis of the studies against the inclusion criteria. After more in-depth reviews of full-papers, 28

papers met the criteria for inclusion. These papers are included in the qualitative, descriptive reviews

of findings. However, of these 28, only 18 papers provided sufficient information and robust enough

research design to be included in the meta-analysis.

- The detailed review provided the following insights into the studies available:

4

o Studies are of two types: general and specific. Some papers considered the general impact of

engaging in education on a statewide level. Others looked at specific programmes and their

impact on recidivism. Nearly all of the studies are from the United States with only one coming

from the UK.

o Most of the studies do not provide information regarding race, gender, sentence length, or

criminal conviction of the participants.

o Many of the studies included discuss participants who were released and follow-up a

significant number of years before publication. For example, some studies look at cohorts

released and followed-up in the 1970s and 1980s. As such, the relevance of their findings is

questionable. However, they have been included due to the small number of relevant papers

included.

o There was little to no consistency regarding the follow-up period. Papers varied from 1 year

to 13.5 years of follow-up for participants.

- Meta-analysis on 18 reoffending studies identified that delivering education in prison settings has a

positive impact on recidivism (various forms) when compared to the outcomes of a comparison or

control group. Overall, the pooled odds ratio indicates a reduction in the likelihood of recidivating of

approximately one third (36 percent).

- Meta-analysis on five employment studies identified that education in prison settings has a positive

impact on employment (in various forms) when compared to the outcomes of a comparison or control

group. Three studies had a positive effect and two demonstrated a negative effect on employment.

Overall, odds ratios indicated a 24 percent reduction in likelihood of gaining employment if the prison

engages in prison education. However, this is based on a small number of papers.

6

TABLE OF CONTENTS

Executive Summary 3

1. Introduction 8

2. Background for the Review 9

Educational deficit in prisoners and links with re-offending 10

Education in English prisons 11

Prior reviews 12

3. Outlining the scope of the reivew 17

3.1 Review questions 17

3.2 Defining the intervention, target group and outcomes 17

4. Methods used in the review 19

4.1 Systematic Reviews and Rapid Evidence Assessments 19

4.2 Characteristics of the studies relevant to the objectives of the review 20

4.3 Search strategy 21

5. Results 27

5.1 Characteristics of included studies 27

5.2 Review of Studies included in the meta-analysis 32

5.3 Context of studies included in the meta-analysis 44

5.4 Meta-Analysis 45

6. KEY Findings 50

7. References Error! Bookmark not defined.

8. Appendicies 55

Appendix A: Data Captured 55

Appendix B: Error! Bookmark not defined.

7

Title: Rapid Evidence Assessment (REA) of offender learning in prisons and outcomes on

Recidivism and employment

Reviewers: Mark Ellison, Kirstine Szifris, Dr Rachel Horan, Prof Chris Fox,

Version: 18/03/16

Contact reviewer: Kirstine Szifris, Mark Ellison

Sources of support:

Peter Cox, Barbara McDonough

Contribution of

reviewers:

Chris developed an initial methodology, which Mark further developed and refined.

Rachel, Kirstine and Mark conducted the searches for this review and refined the list

of papers for inclusion. Kirstine developed the data capture tool and Rachel and

Kirstine then extracted relevant information from the papers for inclusion. Mark,

assisted by Rachel, conducted the data analysis and refined the papers for inclusion

in the meta-analysis. Mark developed techniques for conducted meta-analyses in R.

Chris and Kirstine researched the background information that sets this report in

context. All of the reviewers contributed to the writing of this report.

Acknowledgements:

8

1. INTRODUCTION

The purpose of this review is to examine the evidence available regarding the link between education in prison

and post-release outcomes. The review focuses on employment and recidivism as desirable outcomes of

engaging in prison education. The reviewers have undertaken a systematic search of available literature,

applying strict inclusion criteria to develop a list of 28 studies that are relevant to this question. Only 18 of these

studies provided sufficient information, or had robust enough research design, to be taken forward in the meta-

analysis. Overall, this research demonstrates that prison education does impact on recidivism with statistical

analysis equating it to reducing the likelihood of reoffending by approximately a third. However, results

regarding the impact on employment are more complex to interpret. Findings are also mitigated by varying

methods of measuring recidivism and employment, the skew of information relating to the United States and

the focus of many studies on higher-level education.

This report provides background information regarding the role of education in prisons and the current policy

climate in the UK. Section 3 describes how the reviewers have interpreted the terms ‘prison education’ and

defined the research questions. Section 4 provides a detailed account of how we selected the papers for

inclusion and the methods of data extraction. Section 5 presents two sets of findings. The first set is a descriptive

account of the 28 papers that relate to the link between education in prison and post-release outcomes. The

second set focuses on the 18 papers that were found to be suitable for a meta-analysis. A brief summary of

each of these papers is provided followed by a graphical representation of the statistical analysis undertaken.

The report concludes with a discussion of the key findings.

9

2. BACKGROUND FOR THE REVIEW

All prisons in England and Wales provide education. Over the last 15 years, the focus of policy has been on

improving the literacy and numeracy rates of prisoners. This focus can be traced back to the findings of the 2002

report by the Social Exclusion Unit on how to cut re-offending rates, commissioned by the then Prime Minister,

Tony Blair. The Report outlines nine key factors in re-offending with education and employment being top of

the list, alongside physical and mental health, family networks, housing, and drug and alcohol misuse. The Report

found that many prisoners’ basic skills were poor with 80 percent having the writing skills, 65 percent the

numeracy skills, and 50 percent the reading skills at or below the level of an 11-year-old child (Social Exclusion

Unit, 2002). These figures are attributed to a Home Office statistical release but in recent years, scholars have

questioned their accuracy (see for example Nichols, 2016).

In the years since, a variety of reviews and government reports re-emphasized the role of education in

prisons (see for example, Education and Skills Committee, 2005). In September 2015, Michael Gove, the then

(2015), Secretary of State for Justice, announced a review of education in adult prisons. This review, led by

Dame Sally Coates, closed its call for evidence in December 2015 and reported in May 2016. The

recommendations made by Coates (2016) are far-reaching and ambitious. If followed, they would constitute a

fundamental shift in the way education works in prisons in England and Wales and a move away from the

(current) Offenders’ Learning and Skills Service (OLASS) requirements. In the report, Coates recommends that

government move responsibility for education from the department for Business Innovation and Skills1 to the

Ministry of Justice (MoJ) and that Prison Governors should have responsibility for commissioning and delivering

education and be held accountable for outcomes through publication of prison league tables. Her vision of prison

education is one that is embedded across the whole prison regime with learning at the heart of a prison’s

endeavour. Coates points to a drop in those studying Level 3 (AS / A level) and Open University qualifications

and cites the restrictions placed by MoJ on such studies as the reason. She recommends a core prison curriculum

including English, Maths and ICT delivered to Level 3 and above. Coates highlights the need to tailor education

towards employment highlighting the role of local and national employers to help shape vocational education

in prisons. Importantly, Coates recognises that such fundamental changes to the way in which prisons are judged

and measures requires significant training for staff. She dedicates a chapter of her report to this area

emphasising professional development of staff and highlighting the low pay and lack of available promotions of

prison teachers (prison education departments tend to have a flat structure with one head, and the rest

teachers). She argues for more visibility of prison teaching in the sector as a whole with potential placements

for student teachers in prisons and a ‘Teach First’ style scheme for prison officers that would emphasize the role

of officers in promoting and supporting education beyond the education department. She also recognises that

governors would need significant support and training if they are to become commissioners of education.

The renewed focus on prisons by Gove and Cameron, together with the publication of the Coates

Review has highlighted the role of basic skills, vocational education and other educational opportunities in

1 Subsequently renamed after the formation of Prime Minister Theresa May’s government

10

reducing recidivism and increasing employment activities. In the aftermath of the 2016 referendum on the UK’s

membership of the European Union, Gove returned to the backbenches. As such, it remains unclear whether

the new Justice Secretary will continue with this focus. However, Gove’s time as Justice Minister provided an

opportunity to focus on prison education. As such, this review of extant literature provides a timely update to

the discussion of what is known about the relationship between engaging in prison education and post-release

outcomes.

EDUCATIONAL DEFICIT IN PRISONERS AND LINKS WITH RE-OFFENDING

Data regarding the educational deficits of the prison population are complex and varying. As discussed, widely

reported statistics are difficult to trace and cannot be scrutinised for accuracy. This section provides an overview

of some of the research available in this area.

The Surveying Prisoner Crime Reduction (SPCR) is a longitudinal cohort study of 1,457 adult prisoners sentenced

to between one month and four years in custody in the UK in 2005/06 (Minstry of Justice, 2014). Sampled

prisoners were administered (by interview) a detailed questionnaire on reception to prison, prior to release from

prison, and post-release. Only around a third of prisoners (32 %) reported being in paid employment in the four

weeks prior to custody and 51 % reported being in paid employment during the 12 months prior to custody

(Brunton-Smith & Hopkins, 2014). Both of these figures are considerably lower than the UK general employment

rate, which was 75 % in 2006 for those of working age (Ministry of Justice, 2010). Many prisoners reported

needing help finding a job on release (48%), with education (41%) and to improve work-related skills (40%), and

agreed that having a job would help them stop reoffending (68%) (Hopkins, 2012). A lower likelihood of

reconviction has been associated with having been employed in the year before custody (40% compared with

65% not in employment) and having a qualification prior to custody (45% compared with 60% for those who

didn’t have a qualification) whilst a higher rate of reconviction is associated with having been in receipt of

benefits in the year before custody (58% compared with 42% for those not in receipt of benefits) (Hopkins 2012).

The most comprehensive and up-to-date study found by the reviewers comes from the Centre for Education in

the Criminal Justice System (CfE) at the University of Central London who produced a report into the English and

Maths level of prisoners in England (Creese, 2015). Rather than basing their statistics on MoJ data, they

employed the OLASS provider’s data to assess the levels of literacy and numeracy among prisoners. In prisons

in England and Wales, educational providers are required to conduct initial assessments of English and maths

needs for all prisoners entering the estate (unless they have been tested in the 6-months prior to prevent repeat

testing). The CfE used data from four educational providers, including Novus. They compared the levels of

literacy and numeracy in various prisons to the national levels (as measured by the Skills for Life national survey

conducted by BIS in 2012). Their figures show that 85 percent of general population have literacy skills at L1 or

L2 compared to only 50 percent of prisoners (see Table 1). The differentiate at the lower level showing that the

gap between national literacy levels and prison literacy levels steadily increase from just a couple of percent at

Entry level 1 to over 30 percent difference between those achieving level 2 in prison compared to those

11

achieving level 2 nationally. Their findings regarding numeracy however are very different (see Table 2). Up to

level 1, prisoners out-perform national statistics with 79.4 percent prisoners having Entry Level 3 numeracy skills

and above, compared to 76.4 percent of the general population. It should be noted that this gap is much larger

up to Level 1 where prisoners outperform the national average consistently. However, at Level 2, only 9% of

prisoners are recorded as having achieved level 2 whilst 21.8% of the population have.

Table 1: Literacy levels of prison population compared to general population (Source: Creese 2015)

Table 2: Numeracy levels of prison population compared to general population (Source: Creese 2015)

The key message in outlining the varying statistics available on this issue is to note that the statistics regarding

the educational levels of prisoners vary and are wrought with methodological issues. However, there is general

agreement that prisoners are, on average, less well qualified, less skilled and often come from difficult and

confused educational backgrounds. For these reasons, and many others besides, prisons provide education to

their residents. The following section provides an overview of the nature and content of education in prisons in

England and Wales.

EDUCATION IN ENGLISH PRISONS

Education in English prisons is funded by the Skills Funding Agency.

“The Skills Funding Agency manages the Offenders’ Learning and Skills Service (OLASS) to integrate

offender education with mainstream academic and vocational provision. OLASS allows offenders in

custody, according to need, to receive education and training. This in turn enables them to gain the skills

and qualifications they need to get sustainable employment and have a positive role in society” (Skills

Funding Agency, 2015).

Currently, all prisons in England and Wales maintain an education department usually provided by an outside

contractor. Despite this, research into the impact of current education courses on employment and/or

recidivism is limited. The research that is available is primarily from North America with many focussing on

College level education (or degree-level).

12

Michael Gove MP stated that

“We have more than 80,000 adults in our custody. One of the most important things we can do once

they are inside the prison walls is to make sure that they get the literacy and numeracy skills they need

to make them employable and positive contributors to society once released. For those serving longer

sentences, education and training is a key part of their rehabilitation” (Gove, 2015).

This Rapid Evidence Assessment (REA) has been commissioned by the Manchester College. It aims to provide

an overview of quantitative research currently available that examines the role of prison education in reducing

reoffending and increasing employment.

PRIOR REVIEWS

The Ministry of Justice describes evidence on the effectiveness of employment/education programmes in

reducing re-offending as ‘mixed/promising’ (Ministry of Justice, 2013). A number of systematic reviews have

examined the impact of employment and/or education programmes on re-offending and/or employment. We

are aware of 12 reviews that have been undertaken in the last 20 years:

Pearson and Lipton (1999) undertook at meta-analysis of the effectiveness of educational and

vocational programs. Their study is widely quoted but it has not been possible to obtain a copy of this

report. They conclude that: “Although some types of educational and vocational programs appear

promising in terms of reducing recidivism, due to a lack of studies using high-quality research methods

we are unable to conclude that they have been verified effective in reducing recidivism” (quoted in

Gillis, 2000, p. 33).

Phipps, Korinek, Aos and Lieb (1999) provide an overview of evidence for ‘effective’ programming in

prisons. They categorise research into ‘programmes with some evidence of effect’, ‘programmes where

effectiveness cannot be determined’ and ‘programmes with no evidence of effect’. Studies rated 2 or

below on the Maryland Scale were excluded. For each programme reviewed they provide an overview

of the findings, evaluation limitations and the measures employed. They do not provide commentary

on the overall findings or a statistical analysis of combined research.

Bouffard, Mackenzie, Layton and Hickman (Bouffard, Mackenzie, & Hickman, 2000) use the Maryland

scale to assess the rigour of studies that look at vocational education and work release programmes

impact on recidivism. Only the former is of interest here. They classify programmes based on the

studies they review as ‘Works’, ‘doesn’t work’, ‘promising’ and ‘don’t know’. They conclude that

vocational educational programmes ‘work’ in that they are effective at reducing recidivism. Following

a discussion of the poor evidence available regarding the link between vocational education in prison

and recidivism, the results are however inconclusive. This is also caveated with the low level of scientific

rigour in the studies available.

13

Wilson et al. (2000) undertook a meta-analysis of 33 studies of education programs in custodial settings

administered to adults and published after 1975. They rated each study using the Scientific Methods

Scale, often referred to as the Maryland Scale (Farrington, Gottfredson, Sherman, & Welsh, 2004).

Wilson et al. (2000) included only studies that received at least a Level 2 rating. In their meta-analysis,

they demonstrated that participation in academic programs was associated with an average reduction

in recidivism of about 11 percentage points. Participation was also associated with a greater likelihood

of employment, although this was not quantified. Vocational training programmes did not demonstrate

a consistent relationship with recidivism but were associated with increased employment.

Visher et al. (2005) constructed their review to answer the question: “What is the effect of non-

custodial employment services interventions on the subsequent criminal behaviour of ex-offenders?”

They searched for studies in which some of the treatment was delivered in a non-custodial setting

(treatment delivered in a halfway house, group home, or similar facility were eligible). Interventions

could be residential or non-residential and the intervention had to include a job-placement component

or a job-training component (although other components, such as life-skills training, remedial

education, or social-service assistance, may have been included as well). All interventions involved adult

offenders, although some also involved young offenders. The review had a high methodological

threshold: only random assignment experimental designs were considered eligible. Visher et al.

identified eight studies that met their criteria. The first study being implemented in 1971 and the most

recent study being implemented in 1994. All were US-based studies. Meta-analysis of the results

showed that, on average, the employment interventions examined did not reduce arrest among the

treatment group subjects by more than the amount expected by chance.

MacKenzie (2006) updated Wilson et al.’s (2000) meta-analysis. Studies published after 1980 were

included and the methodological threshold was raised to Level 3 on the Scientific Methods Scale. Meta-

analysis found that the odds of not recidivating were 16 percent higher among academic education

programme participants than nonparticipants. This time, their meta-analysis showed that vocational

programme participation was associated with a reduction in recidivism with the odds of recidivating 24

percent lower among vocational program participants than nonparticipants. The analysis of

employment outcomes was not updated.

Aos et al. (2006) conducted a large meta-analysis of 291 evaluations of offender rehabilitation

programmes for adults. They limited their sample to studies conducted from 1970 onward and included

only studies that received at least a Level 3 rating on the Scientific Methods Scale. They identified seven

evaluations of basic adult education programs in prison and meta-analysis showed a 5.1 percent

reduction in recidivism. They also identified three studies of vocational education in prison showing a

12.6 percent reduction in recidivism.

14

Hurry et al. (2006) undertook a Rapid Evidence Assessment that drew together empirical evidence

about interventions that focus on promoting employment for offenders. They identified 53 studies

covering a range of interventions including education, vocational training and prison work. The majority

came from the US and only nine studies came from the UK. Most studies looked at outcomes for a

single group of participants who had attended a programme. Only a third of studies included a

comparison group. Prison work and vocational training are identified as effective. Education does not

emerge as having an impact on employment but Hurry et al note that the evidence base is too small to

draw firm conclusions and state that the effectiveness of education is “still an open question” (p. 9).

They also note that there is very little description of what ‘education’ involves in the studies they

reviewed.

Jenson and Reed (2008) review six studies of correctional education on recidivism, several of which are

themselves meta-analyses. They make use of the Maryland scale including only those that are at levels

3, 4 and 5 on the scale. Categorising programmes evaluated in the studies (using the Maryland scale)

into ‘What works’, ‘What is promising’, ‘What does not work’ and ‘What is unknown’. For what works

and what does not work, there must be at least two level 3 evaluations of a programme indicating it

does (or does not) reduce recidivism before it can be categorised as working or otherwise. Five out of

6 studies considering high school and GED education indicate education is effective at reducing

recidivism including the three most methodologically rigorous. The conclusions for vocational

education are less firm; although there is indication of reduced recidivism, the studies were less

methodologically rigorous. College programming is demonstrated to have a significant effect on

recidivism and is deemed to be part of the ‘what works’ programming. Life-skills programming is

categorised as ‘what is unknown’.

Davis et al (2013) undertook a systematic review and meta-analysis of programmes that provide

education to incarcerated adults. Their systematic review included a literature search for published and

unpublished studies released between 1980 and 2011 that examined the relationship between prison

education participation and prisoner outcomes in the United States. Prison education was defined as

academic and vocational training programs and thus excluded life skills training, re-entry programs and

work placement programs. The review concentrated on recidivism, but also examined whether

participating in a prison education program was associated with an increase in labour force

participation and whether participating in a prison education program with a computer-assisted

instructional component was associated with gains in achievement test scores. Fifty-eight studies were

identified that met their criteria. Meta-analysis supported the hypothesis that receiving education

while incarcerated reduces an individual’s risk of recidivating after release. Analysis restricted to

higher-quality research studies found that, on average, prisoners who participated in correctional

education programs had 43 percent lower odds of recidivating than prisoners who did not. These results

15

were consistent even when lower-quality studies were included in the analysis. High school/general

education development (GED) programmes were the most common approach to educating prisoners

in the studies Davis et al. examined. They found that prisoners who participated in high school/GED

programs had 30 percent lower odds of recidivating than those who had not. In general, studies that

included adult basic education (ABE), high school/GED, postsecondary education, and/or vocational

training programs showed a reduction in recidivism. The meta-analysis also found that the odds of

obtaining employment post release among prisoners who participated in correctional education (either

academic or vocational programs) was 13 percent higher than the odds for those who had not

participated. However, only one study fell into the higher-quality category. Individuals who participated

in vocational training programs had odds of obtaining post-release employment that were 28 percent

higher than individuals who had not participated. In comparison, individuals who participated in

academic programs had only 8 percent higher odds of obtaining post-release employment than those

individuals who had not participated in academic programs. However, there was no statistically

significant difference between the odds ratios for the two types of programs, because the number of

vocational training studies was relatively small. Based on four available studies Davis et al. estimated

students exposed to computer-assisted instruction relative to traditional instruction learned very

slightly more in reading in the same amount of instructional time and substantially more in

mathematics. However, there was no statistically significant difference in test scores between the

different methods of instruction.

Hall (2015) conducted an ‘assessment’ of the association between correctional education programming

and recidivism. She reviews 10 research studies between 1995 and 2010 based on two key criteria:

that the article contained a primary empirical study of an accredited educational course and that the

study used recidivism as the outcome measure. She provides an overview of each of the studies

highlighting their limitations and outlining their methodologies. She also differentiates between type

of programme that has a particular affect. She makes a strong conclusion stating that education does

work as a “reduction tool for recidivism” (p. 24) and that this is true for all forms of correctional

education. However, she argues that high-level education including degrees is particularly effective at

reducing recidivism.

These previous meta-analyses indicate a general positive impact of education in prison. On the whole, research

indicates that education reduces recidivism and increases employability. Some studies question the impact of

vocational education on recidivism whilst others imply higher-level qualifications (such as university degrees)

have the largest impact (Davis et al.’s study is an exception to this). With respect to employment, there is some

evidence that there is an impact. However, studies of employment are less common and the evidence less

robust. Measuring employment post-incarceration is complex and the lack of research found reflects this.

16

The current REA was deemed appropriate because many of the identified reviews are now dated with most

having been undertaken in or before 2006. Of the two more recent studies, one is not a systematic review (Hall

2015) and the other only includes studies from the US (Davis et al. 2013).

17

3. OUTLINING THE SCOPE OF THE REIVEW

3.1 REVIEW QUESTIONS

The primary research question is:

1. What impact does education in prison have on rates of recidivism and employment?

To investigate this further we address the following questions:

2. What research is currently in existence around the link between education in prison and

recidivism/employment?

3. Is there a distinction between outcomes for different measures of recidivism/employment?

4. Does the level of education impact the outcome of engaging?

3.2 DEFINING THE INTERVENTION, TARGET GROUP AND OUTCOMES

EDUCATION AND VOCATIONAL TRAINING

We have included only programmes that are predominantly educational and delivered in prison. These include

vocational, academic and basic skills courses that are both accredited and unaccredited. We have excluded any

courses delivered outside of a traditional prison such as halfway houses, community centres or ‘camps’ for

offenders. We have also excluded life skills programmes, work place training (unless specifically towards a

qualification) and psychologically based thinking skills programming.

OFFENDERS

Initially, the aim was to develop an understanding of the impact of education with respect to distinct cohorts of

prisoners. However, the majority of studies that met the criteria looked at prisoners in general rather than

specific programmes or type of prisoner. This review pertains to adult education in particular, however, many

of the studies did not specify the age of the prisoners in their study instead just stating ‘prisoners’ or ‘adults’ or

stating a mean age. Further to this, many papers (ten) included in the review did not specify the gender of the

participants with three studies specifying only male offenders, one only female, and the remaining papers

included both male and female prisoners. Ethnicity, offence category and sentence length were either not

discussed or provided as a series of descriptive statics. As such, no study included in this report pertains to

education targeted at a specific subset of offenders. Findings here therefore relate to prisoners in general.

OUTCOMES

Recidivism outcomes are measured in a number of different ways (re-arrest, re-conviction and re-incarceration)

18

over different time-periods. Due to the small number of papers suitable for the meta-analysis stage, we have

combined all recidivism measures together. This is far from ideal as inclusion of re-incarceration due to parole

violations for example could skew the data. Employment measures are equally variable with studies measuring

for example, hours employed, wages earned or type of job. As with recidivism rates, we have had to combine

all measures of employment resulting in a very general outcome measure. This is discussed in more detail below.

19

4. METHODS USED IN THE REVIEW

In this section we describe, briefly what a Rapid Evidence Assessment is before going on to describe the

methodology we adopted to complete the Rapid Evidence Assessment.

4.1 SYSTEMATIC REVIEWS AND RAPID EVIDENCE ASSESSMENTS

SYSTEMATIC REVIEWS

The Campbell Collaboration describes a systematic review as follows:

“A systematic review summarizes the best available evidence on a specific question using transparent

procedures to locate, evaluate, and integrate the findings of relevant research.” (The Campbell

Collaboration, 2015, p. 6)

The key characteristics of a Systematic Review (based partly on Higgens & Green, 2011) are:

a clearly stated set of objectives with pre-defined eligibility criteria for the inclusion of studies;

an explicit, replicable methodology;

comprehensive searching of print, electronic, and unpublished sources;

an assessment of the validity of the findings in a way that makes the quality of the evidence

upon which the review is based transparent;

a systematic presentation, and synthesis, of the characteristics and findings of the included

studies.

Systematic reviews overcome three distinct problems. The first is the sheer volume of information that is

available; the second is that individual studies can reach conflicting results; and the third is the risk of bias in

reviews of the available information.

RAPID EVIDENCE ASSESSMENTS

Systematic reviews can take a significant period of time to complete. Rapid Evidence Assessments (REAs) are

appraisals of existing evidence that sit somewhere between the equivalent of Health Technology Assessments

(HTAs) and fully developed systematic reviews in the field of health care (Government Social Research Unit

2007b).

Rapid Evidence Assessments follow a similar process to Systematic Reviews: collating descriptive outlines of the

available evidence on a topic; critically appraising the studies identified; sifting out studies of poor quality: and

providing an overview of what that evidence tells us, and what is missing from it. Key distinctions between the

systematic review and a Rapid Evidence Assessment are that:

20

REAs include a fairly comprehensive electronic search of appropriate databases, and some searching of

other easily accessible sources, but the search is not exhaustive;

REAs will provide a summary of the evidence;

REAs can be completed in approximately 3 months as opposed to between six months and a year for a

full systematic review;

A key advantage of a REA is its speed. However, the trade-off is the potential loss of some relevant studies,

particularly unpublished ones, raising the prospect that some publication bias will be present in the results of a

REA.

4.2 CHARACTERISTICS OF THE STUDIES RELEVANT TO THE OBJECTIVES OF THE REVIEW

To address the research questions identified for this REA, we searched for reports of studies which evaluated

the effect on re-offending and post release employment of educational and vocational training interventions

delivered in a custodial setting to adult prisoners.2

We have based our definition of ‘education and vocational training delivered in a custodial setting’ on the

definition developed by Davis et al. (2013). We define an educational or vocational programme as one that

includes a basic skills (e.g. literacy, numeracy, language or ICT), academic or vocational curriculum taught by an

instructor and designed to lead to the attainment of a certificate, license, degree or other academic or vocational

qualification widely recognized in the country in which the programme is delivered. Not all of the prison learners

included in the studies necessarily completed the given qualification, but the education they engaged in had the

potential to lead to a recognised qualification. As is discussed below, some of the studies only included those

who gained a qualification whilst others included prisoners who had attended education at some point.

Educational programming could be part of a larger set of services administered to prisoners or it could be a

stand-alone programme. However, it should have an explicit academic or vocational curriculum in place with an

instructional component. Therefore, prison work programs and job placement programs lacking a structured

training component under the supervision of an instructor are not included. Educational programmes can be

delivered either face-to-face or at distance (i.e. through an on-line learning environment) or by a mix of the two.

Although the programme could include post-release (through-the-gate) elements, it must be primarily

administered while the prisoner is held in a correctional setting. Programs administered solely post-release are

excluded. Programmes that are partly delivered to prisoners on day release (e.g. Release on a Temporary License

– ROTL) are excluded.

Studies were excluded if they:

2 We initially stated explicit criteria as being for ‘over-18s’. However the lack of detail and specifics regarding the definition of an ‘adult’ in the majority of studies meant it was necessary to alter our criteria, sifting out only those who specifically focus on under 18s.

21

were not published in the English language;

were published in or before 1994;

did not evaluate the impact or outcomes of education or vocational training delivered in a custodial

sentence.

did not explicitly address academic or vocational skill - for example drug rehabilitation or anger

management programmes were excluded.

were not administered primarily in a custodial setting.

did not have reducing re-offending or promoting employment as one of their stated outcomes;

explicitly concerned offenders under 19 years old – in cases of mixed populations, i.e. non-offenders

and offenders and under 19s as well as over 19s, the researchers had to be confident that the majority

of the sample was comprised of the target group (over 19s).

4.3 SEARCH STRATEGY

A key part of the methodological process for this report involved searching for and identifying the correct papers

for inclusion. As such, the search strategy and refining process followed a rigorous methodology drawing

primarily on Davis et al (2013) and Hurry et al (2006). The first stage in this process involved locating appropriate

studies. To do this we used four methods of study identification:

Searches of electronic databases using key words;

Bibliographic searching of key systematic reviews;

Hand searching key journals;

An Internet searches of key websites likely to contain relevant ‘grey literature’.

After identifying all of the studies we could find that related to the criteria, we reviewed their abstracts to

determine if they were suitable to be taken forward for a meta-analysis. Having refined the list on the basis of

abstracts we then located and downloaded the 53 studies that met the criteria (see below for details of this).

These papers we reviewed in detail. In total, 28 studies are included in this report 18 of which provided sufficient

information to be included in the statistical analysis. This section provides a detailed overview of the steps we

took to locate and refine the papers that are included in this analysis. The purpose of this section is to provide

a clear outline of how we refined the studies to ensure we are transparent in how we came to choose the studies

included here.

SEARCHING ELECTRONIC DATABASES

Specific terms were used to systematically search electronic databases. These were

(prison* OR correctional OR Probation) AND (educat* OR Vocation* OR skills*) AND NOT “correctional

nurse” OR NOT “correctional nursing”

22

The following electronic databases were searched:

British Education Index;

Web of Knowledge (Social Science Citation Index);

ASSIA;

ERIC;

PsycINFO;

National Criminal Justice Reference Service Abstracts;

ProQuest criminal justice YES [might be same as ASSIA?];

C2-SPECTR (Campbell Collaboration);

Education Abstracts [might return same as British Education Index?]; and

Criminal Justice Abstracts

BIBLIOGRAPHIC SEARCHES OF KEY SYSTEMATIC REVIEWS

The following Systematic Reviews were searched:

Wilson, Gallagher, & MacKenzie (2000) A Meta-Analysis of Corrections-Based Education, Vocation, and

Work Programs for Adult Offenders.

Aos, Miller, & and Drake (2006) Evidence-Based Adult Corrections Programs: What Works and What

Does Not. Olympia: Washington State Institute for Public Policy.

Hurry, Brazier, Parker, & Wilson (2006) Rapid Evidence Assessment of Interventions that Promote

Employment for Offenders.

Davis, Bozick, Steele, Saunders, & Miles (2013) Evaluating the Effectiveness of Correctional Education:

A Meta-Analysis of Programs that Provide Education to Incarcerated Adults.

Hall (2015) Correctional Education and Recidivism: Toward a Tool for Reduction.

Jenson & Reed (2008) Adult correctional education programs: An update on current status based on

recent studies.

Phipps, Korinek, Aos and Lieb (1999). Research findings on adult corrections' programmes: A review.

Bouffard, Mackenzie, Layton and Hickman (2000) Effectiveness of Vocational Education and

Employment Programs for Adult Offenders: A Methodology.

HAND SEARCHING OF JOURNALS

Due to the time constraints of the project, we limited the number of journals to be searched by hand to six key

publications. These journals were chosen because they were thought likely to publish relevant studies. Where

electronic versions were available searches were from 1994 to the end of 2015. The journals hand-searched

were:

Journal of Offender Rehabilitation

23

Prison Service Journal

The Prison Journal

The Probation Journal

Journal of Correctional Education

Correction Review

GREY LITERATURE

An Internet search of the following websites was undertaken between December 2015 and January 2016. These

sites were selected because they were thought likely to carry relevant studies:

UK Ministry of Justice

Government websites are all located under www.gov.uk. A search for relevant publications was undertaken.

Publications only go back to 2010. We first went to the publications page for the UK government:

https://www.gov.uk/government/publications. From here we selected ‘Research and Analysis’ and ‘Ministry of

Justice’. The search returned 124 publications. No relevant publications were identified. We then selected

‘Research and Analysis’ and ‘National Offender Management Service’. The search returned 22 publications. No

relevant publications were identified.

Department for Business, Innovation and Skills

Department for Business Innovation & Skills is the department that funds OLASS

https://www.gov.uk/government/organisations/department-for-business-innovation-skills. From here we

selected ‘Publications’ and ‘all publication types’ and policy area ‘law and justice system’. There were 771 results.

We filtered using the key words ‘prison’, ‘correctional’ ‘Probation’ and ‘educat*’ ‘Vocation*’ and ‘skills*.’ No


Prisoners Education Trust

Prisoners Education Trust http://www.prisonerseducation.org.uk

From here we selected ‘Resource Library’. The search returned 61 publications. One report was found: Ministry

of Justice, (2015). Justice Data Lab Re-offending Analysis: Prisoners Education Trust.

CLINKS

Clinks http://www.clinks.org. From here we selected ‘Resources’ and ‘Reports’. All publications listed on

http://www.clinks.org/reports were screened. No relevant publications were identified.

US Department of Education

US Department of Education http://www.ed.gov. From here we selected ‘Data and Research’ and ‘What Works

http://www.gov.uk/

https://www.gov.uk/government/publications

https://www.gov.uk/government/organisations/department-for-business-innovation-skills

http://www.prisonerseducation.org.uk/

http://www.clinks.org/

http://www.clinks.org/reports

http://www.ed.gov/

24

Clearing house’ and ‘intervention reports’. There were 547 results and no relevant publications were identified.

We also selected ‘data and research’ and ‘single study reports’. There were 118 results and no relevant

publications were identified. We searched ‘all topics’ using the key words ‘prison’, ‘correctional’ ‘Probation’ and

‘educat*’ ‘Vocation*’ and ‘skills*.’

Washington State Institute for Public Policy

Washington State Institute for Public Policy published reports on adult corrections are located at

http://www.wsipp.wa.gov/Reports/PolicySubarea/2. There are 126 reports that have been published since

2000. No reports were founds.

Justice Policy Center at the Urban Institute

Justice Policy Center at the Urban Institutehttp://www.urban.org/policy-centers/justice-policy-center. All

publications listed on http://www.urban.org/policy-centers/justice-policy-center/publications were screened.

RAND Corporation

The RAND Corporation http://www.rand.org. From here we selected ‘educational programme evaluation’

http://www.rand.org/pubs.html?pub-date=20150202%3A&tag=Educational+Program+Evaluation&page=1.

There were 26 results and no relevant publications were identified.

We filtered using the key words ‘prison’, ‘correctional’ ‘Probation’ and ‘educat*’ ‘Vocation*’ and ‘skills*.’ No


IDENTIFYING STUDIES FOR INCLUSION

From the initial bibliographic database key word searches 4,304 (abstracts) studies were identified. Lists from

each search were imported into EndNote for de-duplication and then exported and organised in a Microsoft

Access database. Two reviewers then reviewed the abstracts of these titles and marked them as ‘met’, ‘partially

met’ or ‘not met’. Both reviewers checked 2,682 (62%) of these abstracts allowing for a check of inter-rater

reliability. 184 studies were categorised as ‘met’ or ‘partially met’. Two different reviewers then sifted through

these abstracts bringing the list down to 29 papers. A further 59 were categorised as partially met. These 59

papers were then reviewed more closely and five were included in the ‘met’ category. A further five were not

locatable and are therefore excluded from this study. From these searches, there were now 34 studies included

in our list and entered onto the data capture template.

Grey literature searches revealed a further seven studies taking the total to 41 studies. The review of key

systematic reviews revealed a range of studies the majority of which was grey literature from state-specific

websites. They revealed a further 38 relevant studies. However, six of these were excluded using the initial

criteria. There were 15 that could not be located. This is a reasonable number of potentially relevant studies

that cannot be included which could be a cause for concern. However, these studies are primarily ‘grey’

http://www.wsipp.wa.gov/Reports/PolicySubarea/2

http://www.urban.org/policy-centers/justice-policy-center

http://www.urban.org/policy-centers/justice-policy-center/publications

http://www.rand.org/

http://www.rand.org/pubs.html?pub-date=20150202%3A&tag=Educational+Program+Evaluation&page=1

25

literature. Our inability to locate them is due to the references provided being websites that are no longer

available. A further six did not match our criteria. In total, a further 11 studies were added from reviewing other

systematic reviews, the majority of which were grey literature. This meant the total number of papers identified

from abstracts was 52.

All of these papers were added to the data capture template (see Appendix A) for a more in-depth review. At

this stage, 15 were excluded as closer inspection revealed they did not meet the criteria; 6 were themselves a

review of previous literature and have been included in the review of previous meta-analyses; and 3 could not

be located. As such, the total number of studies included in this review is 28. All of these studies have been

reviewed closely and are included in the section on the characteristics of the papers.

The final stage of exclusion came at the stage of statistical analysis. Ten papers were excluded at this stage so

that in total 18 studies were suitable for meta-analysis. Of those excluded, two were first authored by Duguid.

The reason for exclusion is that in the whole sample of 28 studies, three studies were by Stephen Duguid and

colleagues all of which related to the same cohort and reported the same outcomes over the same follow-up

period. As such, only one of these studies has been included in the meta-analysis. A further five studies were

excluded due to a lack of control or comparison group; two due to unclear data presented and one due to a

weak design.

Figure 1: Paper Search Process

DATA COLLECTION AND ANALYSIS

26

A total of 52 studies were entered into the data capture tool for a more in-depth review. As discuss above, 24

of these studies were excluded at this stage of the process. The remaining 28 studies were then reviewed for

the following information;

- Specific outcomes measures of recidivism and employment investigated using statistical techniques.

- Inclusion of sufficient information to calculate individual odds ratios for the meta-analysis. In particular,

inclusion of sample sizes and numbers of recidivists in both the intervention and control.

- Follow-up periods, year of study, definitions of recidivism, type of education, country of origin and study

design.

Each paper was also rated using the Maryland Scale. Papers were excluded from the meta-analysis according to

the following methodological criteria:

- They lack sufficient information to calculate an odds ratio

- They were rated as a ‘1’ or a ‘2’ on the Maryland scale (i.e. they did not include a control group).

The data capture tool was initially modelled on Hurry et al. (2006) and Davies et al. (2013) studies. Utilising their

terminology and questions, 12 papers were reviewed to test the utility of the tool. The data capture codes were

then reviewed and refined by one reviewer. This involved either altering the categories for the codes to suit UK

terminology and the aims of the research; adding in a new question or removing questions not relevant to this

study. The redeveloped data capture tool was then piloted with two reviewers who entered data for 7 and 11

papers respectively, some duplicated. Following this, during a meeting of three reviewers the data capture tool

was finalised. Two reviewers then went through all 52 papers entering the data required. Of the 28 papers

remaining in the review process, over half were reviewed by both researchers. Halfway through the process,

their entries were compared for inter-rater reliability and any discrepancies were discussed and clarified to

ensure data was being captured in a uniform manner.

There are two stages of analysis for these studies. Following Hurry et al (2006), one is a larger descriptive analysis

providing an overview of the nature of the studies, the methodologies most often used and the outcomes that

they focussed on. Data capture tool was reviewed to allow for comparisons across studies. This allowed

reviewers to form a descriptive account of the characteristics of the studies included in this review including

research design, country of origin and choice of outcome measures.

The second stage involves a quantitative analysis of the data provided in these studies using meta-analytic

techniques. A subset of 18 out of the 28 studies reviewed provide sufficient information for statistical analysis.

Of these papers, reviewers extracted data regarding the employment and recidivism rates of prison education

participants and non-participants. This allowed for calculation of odds ratios by study followed by calculation of

pooled odds ratios to look for overall trends. Section 7.4 provides more details of the statistical analysis

employed.

27

5. RESULTS

This section provides an overview of the findings of this review of literature. As discussed above, 28 studies met

the criteria for inclusion. The characteristics of these studies are discussed in section 5.1 outlining their

structure, content and focus. These findings provide an overview of the nature of the research currently

available in this area.

Following this, 18 of the 28 studies contained sufficient information to take forward to a statistical analysis.

These 18 studies are reviewed more closely in section 5.1. Section 5.3 presents a statistical representation of

some of the characteristics of the 18 studies that are relevant to the meta-analytic procedure. The section

concludes with an outline of the findings from the statistical analysis and a discussion of what the findings mean.

5.1 CHARACTERISTICS OF INCLUDED STUDIES

In total, 28 studies are included in this review. This section will provide an overview of these studies, their

content and the information that they use.

TYPES OF REPORT

In general, there are two types of quantitative investigation into the impact of prison education on post-release

outcomes. The first is what we have termed a ‘general’ investigation whilst the other is a ‘specific’ investigation.

‘General’ investigations take the form of looking at a cohort of released prisoners from a particular place (usually

a state in the USA), categorised them as educational ‘participants’ and educational ‘non-participants’ and then

compared their recidivism rate. The term ‘general’ has been used because these reports are looking at the

general impact of engaging in education, of some sort, whilst in prison. 15 out of the 28 studies are of this type.

‘Specific’ investigations have focussed on a particular programme, or set of programmes, that have been run in

a one or more prisons. Many of these studies took a similar form to the above however they focus on a specific

course such as the impact of obtaining a GED (General Educational Development test used in the USA). 11 out

of the 28 studies take this form.

A further two studies were coded as ‘other’. One compares outcomes of two education departments to state-

wide recidivism rates whilst the other is a report of an earlier paper. Attempts were made to source the original

paper. However, we were unable to locate it and, as such, chose to include the summary report.

COUNTRY OF ORIGIN

Nearly all of the studies included in this review are from the United States of America (23/28). Three are from

Canada. However, these papers have the same first author (Duguid) and discuss the same cohort. There is also

one from Australia and one from the UK.

28

There are several potential explanations for this. In the UK, the focus of many studies has been on the ‘deficits’

of offenders thereby providing evidence for the need for education among prisoners. The argument being,

prisoners have low literacy and numeracy levels, they will therefore struggle to find a job upon release which

will affect re-offending rates. There has been little to no focus from the other perspective; does educating

prisoners actually work?

In the USA, scholars have focussed on this question more. There are two events that shed some light as to why

this is. The first was the Martinson report (1974) that infamously concluded that ‘Nothing Works’ in prison

reform. After this point, a host of academics engaged in research in an attempt to demonstrate that this wasn’t

true giving rise to the ‘what works’ agenda and new and innovative methods of evaluation that ask not ‘what

works’ but ‘how does it work and for whom’ (see for example Pawson, 2002).

The other event that is mentioned regularly throughout the papers reviewed here is the removal of PELL grants

for prisoners in 1994. PELL grants are a system of providing grants for poorer students for college level education

in the United States. Since their instatement in 1965 they were available to those incarcerated allowing for a

variety of colleges and universities across America to deliver college level education in prisons. In 1994, the

Violent Crime and Control and Law Enforcement Act revoked Pell Grant funding for any individual incarcerated

(Spearlt, 2016). Several of the papers in this study pertain to ‘postsecondary’ (what we would call university-

level, degree programmes) education in US prisons. Arguably, these papers were in reaction to the revocation

of Pell Grants with universities attempting to demonstrate the impact of this type of programming as a means

of securing funding for prison-college courses.

The lack of available evidence from the UK is surprising. The only study included in this report from the is a

report commissioned by the Prison Education Trust from the Justice Data Lab. As such, it did not provide

sufficient information for inclusion in the report. Reviewing the 25 that were excluded at the data capture stage,

several were from the UK. However, one of these reports was an opinion piece for the Justice Select Committee,

another was primarily qualitative and the third could not be located. The authors are aware of a variety of more

in-depth qualitatively based studies of prison education that is based in the UK. However, these studies do not

measure recidivism and employment which is the focus of this report.

YEAR OF STUDY

Several of the studies reviewed used data that is arguably out of date. In describing their participants,

studies either defined participants by their year of release or by the years in which they participated in prison

education. In reviewing papers for the study, it became apparent that many of the cohorts being examined were

released from prison a significant amount of time before the publication date. For example, the three Duguid

et al studies are all based on 673 prisoners who engaged in prison college programming between 1973 and 1993.

A further study considers individuals released between 1985 and 1998 whilst another looks at released offenders

between 1979-1994. All of these studies were considered for exclusion as their findings could be considered to

be out of date.

29

The information presented demonstrates the varied nature of prison education evaluation. Despite several

other meta-analyses and reviews of literature coming before this one, such variation raised questions regarding

the suitability of meta-analytic techniques in this field. As such, the reviewers reflected on whether further

criteria should be applied (such as the time of the participants’ prison education experience). We chose to

include all papers regardless of when the cohort was released. There were two key reasons for this decision.

First, having a ‘cut off’ date for engaging in education (as opposed to publication date) would go against the

normative processes of meta-analysis and complicate the inclusion/exclusion criteria. Second, early applications

of the odds ratio method for meta-analysis employed the technique to retrospectively analyse the link between

a behaviour and a disease (e.g. lung cancer and smoking). In Mantel and Haentszl’s (1959) paper the example

retrospectively analyses the likelihood of contracting lung cancer whether a person has smoked, or not smoked,

and, if the odds are sufficiently different then the statistics indicate a relationship. In their process, the

relationship holds regardless of how long a person smoked for, how many cigarettes they smoked a day, how

long ago they stopped and, importantly, if they develop lung cancer in the future. Some of the people maybe in

their 40s and others in their 70s. Still, the relationship holds. Applying this logic to the work here, odds ratios

are applicable because they provide evidence of a relationship regardless of how long prisoners were in

education, how much education they had, how long ago they came out of prison and irrespective of if they will

recidivate in the future. As such, inclusion of papers with such varying measures was deemed appropriate for

this method.

However, due to the small number of papers and the overlapping years of the studies, it was agreed that a cut

off point for publication was more suitable than a cut off point for year of study. As such, out of the 28 studies

reviewed, the participants of 12 studies were released from prison or engaged in education whilst in prison in

or before 1995. One study looks at individuals released between 1985 and 1999, another in 1996 and another

in 1997-1998. A further six studies were not clear on the years in which their participants were either prison

learners or released from prison. The remaining seven studies relate to prisoners released in or after 1999.

FOLLOW-UP PERIOD AND DEFINITIONS OF RECIDIVISM

Another issue for a meta-analysis is the varying follow-up periods employed in the studies. Studies looked at

recidivism rates after only one year (including the Justice Data Lab report from the UK) whilst another followed

participants up for 13.5 years. The majority (12/28) followed-up for 3 years. However, four were unclear what

the follow-up period was. Within each study, there was also variation in how long each participant was followed-

up for. For example, some were followed for ‘up to’ 5 years, or a ‘maximum’ of 60 months. This makes

comparisons of different studies very difficult.

In discussing the follow-up period, the outcomes measured also varied. Due to the inclusion criteria, all studies

looked at employment and/or recidivism as post-release outcome measures. However, there are various ways

30

in which both of these can be measured. Recidivism can be measured as re-arrest, re-conviction (for a new

offence) or re-incarceration (which can include parole violation). The stark measure of recidivism does not

account for severity of crime committed.

All of the studies included in this report focus either on recidivism or employment or both as their outcome

measure. Recidivism is measured in a variety of ways. Several studies use several measures of recidivism with

six using rearrests, 16 measuring reincarceration, and four use reconviction. Five are unclear on their measure

of recidivism whilst four use other measures of recidivism including parole violation.

Seven of the studies measure employment as an outcome. However, measuring employment is complex and

the varying methods employed reflect this. Some studies (3/7) looked at wages earned which was measured

differently in each study; hourly, monthly and quarterly. One considered employment as being a minimum of

30 hours per week whilst another looked for verification from a parole officer to confirm employment. Two

examined a range of factors including type of work, hourly wages and total hours worked.

DESIGN OF STUDY

The designs of the studies vary. The majority (14/28) had a non-random control vs. treatment design. Many of

these (8/28) looked at the ‘general’ impact of prison education. For example, the Batiuk et al (2005) report looks

at all prisoner released in the state of Ohio, USA, released between 1989 and 1992. The researchers then

compared the recidivism rates of those who had engaged in prison education compared to those who had not.

Several other studies took this approach. Other studies (4/28) used a ‘matched’ control group where the

intervention group was compared to a control group that were matched on a range of criteria (such as age or

sentence length). Three studies compared the recidivism rates of prison education participants to their expected

or predicted recidivism rates. However, three of these studies are the Duguid et al studies that encompass the

whole cohort.

Two studies used survival analysis as part of their method. This involves tracking a cohort of prison education

participants and measuring how many ‘survive’ without returning to prison within a given time period. None of

the studies used a random allocation methodology. One study used randomisation to select their intervention

and control groups from already existing pool of prisoners but did not randomly assign participants.

TYPE OF EDUCATION

The studies included in this report look at the relationship between attending educational courses and recidivism

and/or employment. The type of education considered varies. As nearly all of the studies included are from

North America, they consider the impact of Adult Basic Education (ABE), General Education Diploma (GED)

attainment, Post-secondary education and/or vocational education. In the UK ABE is equivalent to basic skills

(literacy and numeracy) whilst GED attainment is sufficient to gain a college placement and is (broadly)

31

equivalent to degree-entry qualifications (such as A-Levels). Post-secondary education refers to degree-level

qualifications.

Of the 28 papers reviewed, 16 papers focus on a single type of education whilst 9 cover a range of educational

levels. A further three studies are unclear about the nature of the education they are evaluating. Out of the 16

that focus on a single type of education, 10 focus on post-secondary, or degree-level, qualifications and three

on GED preparation or achievement. Only one focuses on basic skills education and two on vocational

education. The nine that cover a range of qualifications include a combination of ABE, GED, Post-secondary and

vocational qualifications. In total, four include basic skills, eight GED, 3 post-secondary and 6 vocational

education.

In general, the studies included in this review were poor in providing information regarding the nature and

content of the education delivered. None of the studies discussed the number of people in the course or how

the course was delivered. The Cronin study and the three Duguid et al studies specified that college or university

faculty members as opposed to teachers in the prison delivered the educational courses and the Batuik study

specified on-the-job training. These were however exceptions with the rest of the studies providing no

information regarding the content or method of delivery of the educational programme under discussion.

PAPERS NOT INCLUDED IN THE META-ANALYSIS

Before going on to outline the findings of the meta-analysis, this section provides an overview of the 10 papers

excluded at this stage.

- Batuik (1997) found that those who completed college-level qualifications had reduced rates

of recidivism.

- Brewster & Sharp (2002) used survival analysis to analyse length of time out of prison. They

found GED attainment is associated with longer survival whilst vocational education has the

opposite effect.

- Callan and Gardner (2005) found that participants in vocational education whilst in prison

had a recidivism rate of 23 percent compared to a rate of 32 percent.

- Cho & Tyler (2010) found that Adult Basic Education was associated with higher earning and

employment, particularly among ethnic minorities and those that engage in other

educational opportunities as well. However, they found no relationship between ABE and

recidivism.

- Duguid (1997) and Duguid et al (1998) both indicate a reduced rate of recidivism among

offenders who engaged in college-level education.

- Jenkins, Steurer, & Pendry (1995) showed that, in general, education reduces recidivism.

However, their findings indicate that GED and vocational education have greater success with

32

post-secondary or college-level education having the greatest impact on recidivism. They

state ABE is the least effective at reducing recidivism.

- Johnson (2009) found that specially dedicated centres for prisoners focussing on

employment and education results in increased employment and reduced recidivism.

- Kelso (2000) found education is most effective at reducing recidivism when it leads to

diplomas and certificates.

- Lockwood, Nally, Ho, and Knutson (2012) data indicates college education reduces recidivism.

Overall, these papers indicate a general trend that education is associated with reduced rates of recidivism.

There is some suggestion that higher levels of education have a greater impact on recidivism.

5.2 REVIEW OF STUDIES INCLUDED IN THE META-ANALYSIS

As discussed above, of the 28 studies considered to have met the criteria for the REA, only 18 used a research

method and provided sufficient information for a meta-analysis.

This section provides a brief summary of each of the studies used for the meta-analysis.

Batiuk, Lahm, McKeever, Wilcox, & Wilcox, (2005)

Reference Country of Study Education Level Measure of Recidivism

Batiuk, M. E., Lahm, K. F.,

McKeever, M., Wilcox, N.

and Wilcox, P. (2005).

Disentangling the effects

of correctional education:

are current policies

misguided? An event

history analysis. Criminal

Justice, 5(1), 55-74.

USA Secondary education; GED

preparation; post-

secondary education; and,

vocational education.

recidivism

Description of Programme

Comparison of prisoners engaged in education in Ohio Department of Corrections educational programmes

and those not. Analysis of recidivism rates with distinction between levels of education.

33

Overview of Research Design & Analysis

(Non-random) treatment vs. control follow up

Treatment Group Comparison Group Maryland Scale

GED (n=321; 45.8%

recidivism)

Control (n= 292;

50% recidivism)

3 (could be 2 due to non-

random)

Burke & Vivian (2001)


Burke, L. O. and Vivian, J.

E. (2001). The effect of

college programming on

recidivism rates at the

Hampden County house

of correction. Journal of

Correctional Education,

52(4), 160-162.

USA Post-secondary education

(Associate Degree

Programme).

reincarceration


A five-year comparison of the recidivism of participants completing at least one 3-credit college course at

the Hampden County Correctional Centre, compared to a matched control group.




College (n=32; 46.8%

reincarceration)

Non-College (n=32;

68.7%

reincarceration)

3

Cronin (2011)


Cronin (2011) The Path to

Successful Re-entry: The

Relationship Between

Correctional Education,

Employment and

Recidivism, Columbia,

Mo.: University of

USA returned to prison within

2 years of release

34

Missouri Institute of Public

Policy, Report 15-2011.


Tests three hypothesis - relationships between education in prison and after prison employment;

relationship between education and recidivism; relationship between employment and recidivism




Cohort 3: Prisoners who

earned their GED in prison

(n=2898; 40.09%)

Cohort 1: Prisoners

who came into

prison without a

GED and made no

progress (n=7449;

53.7%)

3

Duguid (1998)


Duguid, S. and Pawson, R.

(1998). Education, change,

and transformation, the

prison experience.

Evaluation Review, 2(4),

470-495.

Canada Liberal arts, prison based

University programme.

Leave


Comparison of SIR predicted and actual recidivism of a group of participants who had undertaken a liberal

arts University programme in custody, and analysis of sub-groups according to educational attainment.


3


n=654; 25% recidivated in

the three years following

n= 654; 42.2%

recidivism rate

(predicted % based

3

35

on Statistical

Indicator on

Recidivism (SIR) risk

predictor)

Duwe & Clark (2014)


Duwe, G. and Clark, V.

(2014). The effects of

prison-based educational

programming on

recidivism and

employment. The Prison

Service Journal, 94(4),

454-478.

USA Secondary; and, post

secondary.

Reincarceration up to 4

years


Analysis of the impact of secondary and postsecondary degree attainment in custody on recidivism and

post-release employment outcomes among offenders released from Minnesota prisons between 2007 and

2008.


1 Retrospective Quasi-experimental design. Comparing those who gained an educational qualification to

those that did not. Survival analysis model for recidivism. Ordinary least squares regression on

employment data.


Achieving a Secondary

degree (n=1,212; 17.3%

reincarceration)

Control Group

(n=8,182; 21%)

3

Gordon & Weldon (2003)


Gordon, H. R. D. and

Weldon, B. (2003). The

impact of career and

technical education

programmes on adult

offenders: learning behind

bars. Journal of

USA GED preparation; and,


Recidivism

36

correctional education,

54(4), 200-209.


Analysis of the recidivism of participants of vocational education, with or without GED preparation at

Huntsville Correctional Centre, compared to a control group who did not participate in educational

programmes.


Analysis of recidivism rates for vocational completers, GED prisoners and non-participating prisoners.

Differentiated between different types of recidivism.


Vocational quantification

(n=169; 8.75%)

Control Group

(n=193; 26%)

2

Harer (1995)


Harer, M.D. (1995). Prison

education programme

participation and

recidivism: a test of the

normalisation hypothesis.

Washington, D.C.: Federal

Bureau of Prisons, Office

of Research and

Evaluation 1995.

USA Adult basic education;

GED preparation; post-

secondary education;

social skills; and, adult

continuing education.

3 year follow-up


Analysis of the recidivism of participants of education backgrounds, controlling for several background and

post-release measures.




Those completing at least

0.5 courses during each 6

Control Group

(n=586; 44.5%)

3

37

months of their prison

term (n=619; 30.1%)

Hull, Forrester, Brown, & McCullen (2000)


Hull, K.A., Forrester, S.,

Brown, J. J. D. and

McCullen, C. (2000).

Analysis of recidivism

rates for participants of

the academic / vocational

/ transition education

programmes offered by

the Virginia Department of

Correctional Education.

Journal of Correctional

Education, 51(2), 256-261.


GED preparation; and,


Reincarcerated


Participants of academic programmes and participants of vocational programmes in the Virginia

Department of Corrections establishments were compared with participants who did not enrol or

participate in education or vocational programmes. Rates of reincarceration and rates of employment were

compared.




Academic completers

(n=451; 19.1%)

No educational

involvement during

incarceration

(n=1,307; 49.1%)

3

Kim & Clark (2013)


38

Kim, R.H. and Clark, D.

(2013). The effect of

prison-based college

education programs on

recidivism: Propensity

Score Matching approach.

Journal of Criminal Justice,

41(3), 196-204.

USA Post-secondary education Rearrest within 3 years


Recidivism comparison of prison based college educational programme participants, using a treatment and

matched comparison group.


Propensity Score Matching, comparison of recidivism rate between the treatment and matched comparison

group.


Intervention (after PSM

matching) (n=680; 9.4%)

Control (after PSM

matching)

(n=30838; 17.1%)

3

Lichtenberger, 2011


Lichtenberger, E. (2011).

Measuring the Effects of

the Level of Participation

in Prison-Based Career

and Technical Education

Programs on Recidivism.

Working Paper, 2011.

USA Vocational training Leave


Comparison of the recidivism of prisoners who engaged in vocational training (termed technical training)

whilst incarcerated at correctional facilities in Virginia using propensity score matching.


F13.2 Propensity score matching from a pool of controls (two groups, treatment and control)


Full Completers (n=693;

22.22%)

Comparison Group

(n=693; 28.72%)

3

39

Ministry of Justice (2015)


Ministry of Justice, (2015).

Justice Data Lab Re-

offending Analysis:

Prisoners Education Trust.

UK Post-secondary education one year proven

reoffending rate


Report on impact of receiving PET funding on recidivism compared to matched control group.


Propensity score matching from a pool of controls (two groups, treatment and control,) (Non-random)

treatment vs. control follow up


Intervention (n= 3085;

19%)

PSM Comparison

Group (203,543;

26%)

3

Nally, Lockwood, Ho, & Knutson (2011)


Nally, J. M. Lockwood, S.

R. Ho, T. Knutson, K.

(2011). The Effect of

Correctional Education on

Post release Employment

and Recidivism: A 5-Year

Follow-Up Study in the

State of Indiana. Working

Paper.

USA GED preparation;

secondary education;

post-secondary education;

and, vocational education.

Return to Custody up to 3

years


Comparison of the recidivism of participants in three prison's postsecondary education programmes (three

treatment groups) compared to three matched control groups.




Study Group: High School

(n=881; 29.7%)

Comparison Group:

High School (n=846;

63.8%)

3

40

Nuttall, Hollmen, & Staley (2003)


Nuttall, J., Hollmen, L. and

Staley, E. M. (2003). The

effects of earning a GED

on recidivism rates.

Journal of Correctional

Education, 54(3), 90-94.

USA GED preparation. Returned to custody in 3

years


Comparison of the recidivism of participants of Indiana Department of corrections programmes, compared

to comparison group.


2 Propensity score matching to individually match offenders who earned a GED or high school diploma in

prison with a comparison group released without a secondary degree. Then again for those who had or

hadn't earned a post-secondary degree.


Earned GED at DOCS

(Aged 21 or over) (n=1956;

30.2%)

No Degree (8,422;

34.6%)

3

Steurer, Smith, & Tracy (2001)


Steurer, S. J., Smith, L. G.

Tracy, A. Education

Reduces Crime: Three-

State Recidivism Study,

Lanham, Md.: Correctional

Education Association.


GED preparation; and,


Re-incarceration (3 year

follow-up)


Comparison of the recidivism of prisoners participating in Maryland, Minnesota and Ohio correctional

education programmes and a matched control group of non-participants.




Participants (n=3,099;

2.9%)

Non-Participants

(n=3,099; 29.3%)

4 (multiple states)

41

Torre & Fine (2005)


Torre, M. E. and Fine, M.

(2005). Bar none:

extending affirmative

action to higher education

in prison. Journal of Social

Issues, 61(3), 569–594.

USA Post-secondary education Returned to custody in 3

years


Comparison of the recidivism of prisoners engaged in education in Bedford Hills Correctional Facility College

and non-participants.




Intervention (n= 274;

7.7%)

Control (n=2,031;

29.9%)

3

Wilkinson & Stickrath (1995)


Wilkinson, R. A. and

Stickrath, T. J. (1995).

Evaluation of the Impact

of Correctional Education

Programs on Recidivism.

Ohio Department of

rehabilitation and

Correction

USA Secondary education; GED

preparation; post-

secondary education; and,


Recidivism Rate, 2 year

follow-up


Comparison of the recidivism of prisoners engaged in education in Ohio State prison system and non-

participants.




Qualification Achievement

(n= 1,840; 27.9%)

Comparison Group

(n=14,099; 30.4%)

3

42

Winterfield, Coggeshall, Burke-Storer, & Tidd (2009)


Winterfield, L., Coggeshall,

M., Burke-Storer, V. C. and

Tidd, S. (2009). The Effects

of Postsecondary

Correctional Education:

Final Report, Washington,

D.C.: Urban Institute.

USA Post-secondary education Returned to Prison within

1 years of release (only

Indiana*)


Comparison of the recidivism of participants in three prison's postsecondary education programmes (three

treatment groups) compared to three matched control groups.




PSE (n= 328; 2.44%) Non-PSE (n=1078;

5.29%)

4 (weighted multiple

states)

Zgoba, Haugebrook, & Jenkins (2008)


Kristen Zgoba, Sabrina

Haugebrook, and Jenkins,

K. (2008). The influence of

GED obtainment on

prisoner release

outcomes. Criminal Justice

and Behaviour, 35(3), 375

- 387.

USA GED preparation. Re-arrests (7 year follow-

up)


Comparison of the recidivism of prisoners participating in New Jersey Department of Corrections GED

programme and a matched control group of non-participants.




43

Intervention (n=250; 51%) Comparison Group

(n=153; 64%)

3

44

5.3 CONTEXT OF STUDIES INCLUDED IN THE META-ANALYSIS

As discussed above, measures of employment and recidivism varied between studies – and in some cases within.

A further point of concern for a meta-analysis is the varying follow-up periods employed in different studies (see

figure 5.1) making it difficult to compare the recidivism rates in different studies due to the variation in follow-

up period and methods of recording recidivism. However, reflecting the work of Davis et al (2013) and Hurry et

al (2006) we have combined the different measures in the 18 studies suitable for meta-analysis. However, we

caution that the recidivism rates used are based on varying follow-up periods and study periods (see figure 5.1).

Figure 5.1 Varying study periods and follow-up periods

45

5.4 META-ANALYSIS

SYNTHESIS PROCEDURES AND STATISTICAL ANALYSIS

To understand the relationship between prison education and recidivism, each study effect size was calculated

using the odds ratio (OR). This allows us to look at the number of people who have recidivated in the

intervention group against the number who have recidivated in the control group. In other words, the ratio of

recidivism of those who have engaged in education compared to those who have not.

OR = Odds Ratio

NIa is the size of the intervention group

NCa is the size of the Control group

Ia is the number of events in the intervention group (always stands that Ia<NIa) – number of people who have

recidivated that have engaged in education whilst in prison.

Ca is the number of events in the control group (always stands that Ca<NCa) – number of people who have

recidivated that have engaged in education whilst in prison.

These calculations were cross-validated using an on-line odds ratio calculator

https://www.medcalc.org/calc/odds_ratio.php and R (which was also used to produce Forest Plots).

Figure 5.2 presents a forest plot of the odds ratio and 95% confidence interval for each study. The black square

point represents the effect size estimate for each study using the odds ratio (OR). The whiskers (thin black lines)

represent the 95% confidence intervals which tells us the range of values between which we can be 95%

confident that the value actually lies. Therefore, the smaller the ‘whiskers’ the more confident we can be that

the findings are a true reflection of impact.

The plot illustrates that all studies favour education, meaning delivering education in prison settings have a

positive impact of recidivism (various forms as described above) when compared to the outcomes of a

comparison or control group. All studies examined demonstrated this effect; however, two studies – Batuik

(2005) and Burke (2001) – have confidence intervals that breach the positive effect. This means that two studies

show that education participation could have a negative effect on recidivism. However, the pooled Odds Ratio

https://www.medcalc.org/calc/odds_ratio.php

46

effect is 0.64 (confidence interval between 0.62 and 0.67). This is an overall positive effect of education on

recidivism. Importantly, the confidence interval for this is very narrow and we can therefore be confident that

the true impact of education on recidivism is close to that which has been found here.

The reviewers also ran tests to look at whether there was a statistical difference between odds ratios based on

different educational levels. Several studies indicate that the higher the level of education the greater the

impact on recidivism. For example, Wilkinson’s study indicates that completion of higher-level qualifications

such as college-level degrees results in lower recidivism rates than achieving qualifications in lover levels.

However, they failed to supply n-values for the control groups to allow for inclusion of this study in this analysis.

Therefore, statistical analysis regarding educational levels is inconclusive with further investigation required.

Figure 5.2 Forest Plot of the Odds Ratio for each of the 18 recidivism studies

47

PRISON EDUCATION ON EMPLOYMENT POST RELEASE

From the 18 studies identified for the recidivism meta-analysis, five studies have examined the effect of prison

education interventions on employment outcomes. As discussed above, there are different measures of

employment outcomes making comparisons between studies difficult to make. Cronin (2011) stated that the

employment rate was defined as individuals having a full-time job as reported by their parole officer within 2

years of release. 58.6% of prisoners who earned their GED in prison had a full-time job, compared with 45.7%

of prisoners who came into prison without a GED and made no progress. Duwe et al (2014) stated that 59.5% of

offenders achieving a Secondary degree in prison found employment within 2 years, compared with 49.8% in

the control group. Similarly, 71 percent of prisoners achieving a post-secondary qualification found employment

First Author (Year) Odds Ratio (95% CI)

Batuik (2005) 0.84 [ 0.61 , 1.16 ]

Burke (2001) 0.40 [ 0.14 , 1.11 ]

Cronin (2011) 0.57 [ 0.52 , 0.62 ]

Duguid (1998) 0.45 [ 0.36 , 0.57 ]

Duwe (2014) 0.78 [ 0.67 , 0.92 ]

Gordon (2003) 0.27 [ 0.14 , 0.51 ]

Harer (1995) 0.53 [ 0.42 , 0.67 ]

Hull (2000) 0.24 [ 0.18 , 0.31 ]

Kim (2013) 0.50 [ 0.38 , 0.65 ]

Lichtenberger (2011) 0.70 [ 0.55 , 0.90 ]

MOJ (2014) 0.66 [ 0.60 , 0.73 ]

Nally (2012) 0.28 [ 0.23 , 0.35 ]

Nuttall (2003) 0.81 [ 0.73 , 0.91 ]

Steurer (2001) 0.71 [ 0.63 , 0.79 ]

Torre (2005) 0.19 [ 0.12 , 0.30 ]

Wilkinson (1995) 0.88 [ 0.79 , 0.98 ]

Winterfield (2009) 0.44 [ 0.21 , 0.94 ]

Zgoba (2008) 0.35 [ 0.22 , 0.55 ]

Pooled Effect 0.64 [ 0.62 , 0.67 ]

Favours Intervention Favours Comparison

Odds Ratio

0.1 1.0 10.0

48

compared with 68.3% in the control group. Hull et al. (2000) identified that of the 68 of prisoners who completed

programming in prison, 77.9 percent were employed for a period exceeding 90 days, compared with 42 percent

(the paper states 54.6%, but the calculation is incorrect) of the 183 prisoners with no educational programming.

Nally et al. (2012) study identified that 28.1 percent of the 1,077 offenders in the study group were employed

for at least one quarter in the 2008 Q1 to 2009 Q2 study period. This was lower than the comparison group

(37.1%). However, an examination of employment sustainability indicates that 17.5% of offenders from the

study group remained in employment for three or more quarters (9 months or more) compared with 14.7

percent from the comparison group. Finally, Steurer et al. (2001) identified of the 3,099 participants in prison

education, 77.3 percent were employed at some point during the following three years (‘ever’ employed). This

is lower than the non-participants (81.4%)

The Forest Plot (figure 5.3) illustrates the odds ratio for each study. As the desire outcome is an increase in

employment (as opposed to a decrease in recidivism) the box and whiskers that sit to the right of the line show

a positive effect of education. Those that sit to the left, indicate a negative effect. The first three studies depicted

in figure 5.3 (Cronin, Duwe and Hull) have a positive effect of education in prisons on employment outcomes.

The final two studies (Nally and Steurer) illustrate a negative effect. However, as in the case of recidivism, we

have used the ‘headline’ figure of recidivism within each report to generate this graph. The benefit of this

method is that we are able to compare similar measures. However, individual studies often look at the situation

more closely and Nally et al report that if we look at longer-term employment (or sustainable employment) the

effect of education is in fact positive; it is only when we combine it with reports of those in short-term

employment that a negative effect is detected. This illustrates the complexity of link between education and

employment. Despite the inclusion of some reports that indicate a negative impact of education on

employment, the overall Pooled Effect of prison education on employment outcomes is positive with a pooled

odds ratio of 1.24 (confidence intervals between 1.16 and 1.31).

Figure 5.3 Forest Plot of the Odds Ratio for each of the five employment studies

49

INTERPRETING ODDS RATIOS

Odds ratios are calculated using the probability of recidivism/employment of prison education participants

against non-participants. Finding a pooled ratio of 0.64 means that participating in prison education reduces

the likelihood of recidivism by approximately one third. Note this is distinct from saying that the results show

prison education reduces recidivism by a third. Instead, the odds ratio says that a person who has engaged in

prison education is two thirds as likely (0.64=64%~2/3) as someone who has not engaged in prison education to

recidivate.

Interpreting the findings from the employment studies should be done with more caution as two out of the

three studies sit to the left of the forest plot above (figure 5.3). However, the overall pooled odds ratio stands

at 1.24. Again, we can interpret this as those who have engaged in prison education are 24% more likely to find

employment than those who have not. This does not mean that the rate of employment is increased by 24%,

but that the likelihood of finding employment is increased. In other words, if rates of employment are very low,

the impact of education may be very small.

First Author (Year) Odds Ratio (95% CI)

Cronin (2011) 1.68 [1.54, 1.83]

Duwe (2014) 1.48 [1.23, 1.78]

Hull (2000) 4.86 [2.55, 9.26]

Nally (2012) 0.66 [0.55, 0.80]

Steurer (2001) 0.78 [0.69, 0.88]

Pooled Effect 1.24 [1.16, 1.31]

Favours Intervention Favours Comparison

Odds Ratio

0.1 1.0 10.0

50

6. KEY FINDINGS

There is very little research relating to the impact of education in prisons in England and Wales. Research

reviewed here primarily comes from the United States and often relates to degree-level qualifications. Studies

lack consensus on the definition of recidivism and the appropriate measure for employment. Furthermore,

many of the studies lack sufficient details regarding the nature and content of the intervention to be able to

make specific conclusions regarding ‘what works’.

Many of the studies included in this report fail to supply sufficient information regarding the cohort under study.

In analysing the studies available that investigate the link between education and recidivism, it became apparent

that the small number that are available are often of low quality. Few studies gave details regarding the type of

education they referred to, the nature of delivery, or the population they referred to. There is no way to know

if education is more or less effective for particular individuals and instead come to the vague conclusion that

education ‘works’. However, it is not possible to discuss what type of education is most effective nor at what

stage in a sentence or for what age group it is most relevant. Possibly most importantly, it is not possible to

discern whether education acts as a catalyst for change or an enabler for change. With respect to employment,

we do not know if it is the skills gained, the qualifications the individual now has or whether it is a show of

confidence and articulacy gained through education that secured work.

Overall, the REA and meta-analytic process has been useful in understanding the landscape of information

available in this area. It has revealed an overall trend that education in prisons is related to lower recidivism

rates and increased employment, although the latter is stated with less certainty. It has revealed the lack of

extant literature related to the situation in England and Wales in particular and highlighted the complexities of

comparing findings when there is no universal agreement on the measure of recidivism or employment. Finally,

it has highlighted the need to combine methods to look beyond the term ‘education’ and consider types of

education and methods of delivery as well as to consider how and why education ‘works’.

51

7. REFERENCES

Papers with one asterisk were one of the 28 papers reviewed in detail.

Papers with two asterisks were one of the 28 papers reviewed in detail and also one of the 18 papers included

in the meta-analysis.

Aos, S., Miller, M., & and Drake, E. (2006). Evidence-Based Adult Corrections Programs: What Works and What

Does Not. . Washington State Institute for Public Policy. Olympia.

Batuik, M. E. (1997). The State of Post-Secondary Correctional Education in Ohio . Journal of Correctional

Education , Vol.48(2) pp.70-72.

**Batiuk, M. E., Lahm, K. F., McKeever, M., Wilcox, N., & Wilcox, P. (2005). Dientangling the effects of

correctional education: are current policies misguided? An event history analysis. Criminal Justice , Vol.5(1)

pp.55-74

Bouffard, J., Mackenzie, D. L., & Hickman, L. J. (2000). Effectiveness of Vocational Education and Employment

Programs for Adult Offenders. Journal of Offender Rehabilitation , Vol. 31(1-2) pp.1-41.

*Brewster, D. R., & Sharp, S. F. (2002). Educational Programs and Recidivism in Oklahoma: Another look. The

Prison Journal , Vol.82(3) pp.314-334.

Brunton-Smith, I., & Hopkins, K. (2014). The impact of experience in prison on the employment status of longer-

sentenced prisoners after release Results from the Surveying Prisoner Crime Reduction (SPCR) longitudinal cohort

study of prisoners. London: Ministry of Justice.

**Burke, L. O., & Vivian, J. E. (2001). The effect of college programming on recidivism rates at the Hampden

County house of correction. Journal of Correctional Education , Vol.52(4) pp.160-182.

*Callan, V., & Gardner. (2005). Vocational education and training provision and recidivism in Queensland

correctional institutions. National Centre for Vocational Education Research .

*Cho, R. M., & Tyler, J. H. (2010). Does Prison-Based Adult Basic Education Improve Postrelease Outcomes for

Male Prisoners in Florida? . Crime and Delinquency, Vol XX(X) pp.1-31.

Coates, S. (2016) Unlocking Potential: A review of education in prison, London: Ministry of Justice

Creese, B. (2015). An assessment of the English and Maths skills levels of prisoners in England. Centre for

Education in the Criminal Justice System. London: UCL Institute of Education.

**Cronin. (2011). The Path to Successful Reentry: The Relationship Between Correctional Education, Employment

and Recidivism, . University of Missouri Institute of Public Policy, Report 15, Columbia, Mo.:

Duwe, G., & Clark, V. (2014). The effects of prison-based educational programming on recidivism and

employment. . The Prison Service Journal , 94 (4), 454-478.

Davis, L. M., Bozick, R., Steele, J. L., Saunders, J., & Miles, J. N. (2013). Evaluating the Effectiveness of Correctional

Education: A Meta-Analysis of Programs that Provide Education to Incarcerated Adults,RAND Corporation: .

Retrieved Feb 24, 2016 from: http://www.rand.org/pubs/research_reports/RR266.html

Duguid, S. (1997). Confronting Worst Case Scenarios: Education and High Risk Offenders . Journal of Correctional

Education , 48 (4), 153-159.

http://www.rand.org/pubs/research_reports/RR266.html

52

**Duguid, S. and Pawson, R. (1998). Education, change, and transformation, the prison experience. Evaluation

Review , Vol.2(4) pp.470-495.

Duguid, S., Hawkey, C., & Knights, W. (1998). Measuring the Impact of Post-Secondary Education in Prison .

Journal of Offender Rehabilitation , 27, 87-106.

Education and Skills Committee (2005) Prison Education, London: The Stationary Office Ltd

Gillis, C. A. (2000). Reconceptualizing offender employment. Retreived from http://www.csc-

scc.gc.ca/research/forum/e122/e122h-eng.shtml on 1st March 2016

**Gordon, H. R. D. and Weldon, B. (2003). The impact of career and technical reducation programmes on adult

offenders: learning behind bars. Journal of Correctional Education, Vol.54(4) pp.200-209.

Gove, M. (2015). Announcement of a review of education in adult prisons, Written Statement to Parliament:

Education in Prison, 8th September. Retrieved 2016 4-Mar from www.gov.uk:

https://www.gov.uk/government/speeches/education-in-prison

Government Social Research Unit (2007b) Background paper 2 – What do we already know? Harnessing existing

research. London: Cabinet Office.

Hall, J. (2015). Correctional education and recidivism: Toward a tool for reduction. The Journal of Correctional

Education , Vol.66(2) pp.4-29.

**Harer, M.D. (1995). Prison education programme participation and recidivism: a test of the normalisation

hypothesis. Washington, D.C.: Federal Bureau of Prisons, Office of Research and Evaluation 1995.

Harper, G. and Chitty, C. (Eds.) (2005) ‘The Impact of corrections on re-offending: a review of ‘what works’’,

Home Office Research Study 291, London: Home Office

Hopkins, K. (2012). The pre-custody employment, training and education status of newly sentenced prisoners

Results from the Surveying Prisoner Crime Reduction (SPCR) longitudinal cohort study of prisoners. London:

Ministry of Justice.

Hull, K., Forrester, S., Brown, J. J., & McCullen, C. (2000). Analysis of recidivism rates for participants of the

academic / vocational / transition education programmes offered by the Virginia Department of Correctional

Education. Journal of Correctional Education , Vol.51(2) pp.256-261.

Hurry, J., Brazier, L., Parker, M., & Wilson, A. (2006). Rapid Evidence Assessment of Interventions that Promote

Employment for Offenders: Research Report No 747. London DFES.

Jenkins, H. D., Steurer, S. J., & Pendry, J. (1995). A Post Release Follow-Up of Correctional Education Program

Completers Released in 1990-1991 . Journal of Correctional Education , Vol.46(1) pp.20-24.

Jenson, E. L., & Reed, G. E. (2008). Adult correctional education programs: An update on current status based on

recent studies. Journal of Offender Rehabilitation , Vol. 44 (1) pp.81-98.

*Johnson, J. A. (2009). Lifetech Institute: Leading Change through Transitional Centers. Community College

Journal of Research and Practice, Vol.33(11) pp.942-944.

*Kelso, J. C. (2000). Recidivism Rates for Two Education Programs' Graduates Compared to Overall Washington

State Rates. Journal of Correctional Education, Vol.51(2) pp.233-236.

http://www.csc-scc.gc.ca/research/forum/e122/e122h-eng.shtml

http://www.csc-scc.gc.ca/research/forum/e122/e122h-eng.shtml

53

**Kim, R.H. and Clark, D. (2013). The effect of prison-based college education programs on recidivism:

Propensity Score Matching approach. Journal of Criminal Justice, Vol.41 (3) pp.196-204.

**Lichtenberger, E. (2011). Measuring the Effects of the Level of Participation in Prison-Based Career and

Technical Education Programs on Recidivism. Working Paper, 2011.

*Lockwood, S., Nally, J. M., Ho, T., & Knutson, K. (2012). The Effect of Correctional Education on Postrelease

Employment and Recidivism: A 5-Year Follow-Up Study in the State of Indiana . Crime & Delinquency, Vol.58(3)

pp.380-396 .

MacKenzie, D. L. (2006). What works in corrections: Reducing the criminal activities of offenders and delinquents.

New York: Cambridge University Press.

Martinson (1974). ‘What works? Questions and answers about prison reform’, The Public Interest, 35. 22-54.

Ministry of Justice. (2010). Compendium of reoffending statistics and analysis Executive Summary: Ministry of

Justice Statistics bulletin. London: Ministry of Justice.

Ministry of Justice. (2013). Transforming Rehabilitation: A summary of evidence on reducing re-offending .

Ministry of Justice.

Minstry of Justice. (2014). Surveying Prisoner Crime Reduction (SPCR). Retrieved 2016 14-Mar from www.gov.uk:

https://www.gov.uk/government/collections/surveying-prisoner-crime-reduction-spcr

**Ministry of Justice, (2015). Justice Data Lab Re-offending Analysis: Prisoners Education Trust.

**Nally, J. M., Lockwood, S. R., Ho, T., & Knutson, K. (2011). The Effect of Correctional Education on Postrelease

Employment and Recidivism: A 5-Year Follow-Up Study in the State of Indiana . Working Paper.

**Nuttall, J., Hollmen, L. and Staley, E. M. (2003). The effects of earning a GED on recidivism rates. Journal of

Correctional Education, Vol.54 (3) pp.90-94

Pearson, F. S., & Lipton, D. S. (1999). The effectiveness of educational and vocational programs: CDATE meta-

analyses. Presented at Annual Meeting American Society Criminolgy. Toronto.

Sherman, L. W., Gottfredson, D., MacKenzie, D., Eck, J., Reuter, P. and Bushway, S. (1998) ‘Preventing Crime:

What Works, What Doesn’t, What’s Promising’, National Institute of Justice Research in Brief, Washington D. C.:

National Institute of Justice

Social Exclusion Unit (2002). Reducing Re-offending by ex-prisoners: Summary of the Social Exclusion Unit report.

Retrieved Feb 2016, 24 from

http://webarchive.nationalarchives.gov.uk/+/http:/www.cabinetoffice.gov.uk/media/cabinetoffice/social_excl

usion_task_force

Spearlt. (2016). Keeping in REAL: Why congress must act to restore Pell Grant funding for prisoners. University

of Massecheusetts Law Review , 11

**Steurer, S. J., Smith, L. G., & Tracy, A. (2001). Education Reduces Crime: Three- State Recidivism Study,.

Lanham, Md: Correctional Education Association.

**Torre, M. E. and Fine, M. (2005). Bar none: extending affirmative action to higher education in prison. Journal

of Social Issues, Vol.61 (3) pp.569–594.

https://www.gov.uk/government/collections/surveying-prisoner-crime-reduction-spcr

54

Viechtbauer (2010). Conducting meta-analysis in R with the metafor package. Journal of Statistical Software Vol.

36(3)

Visher, C., Winterfield, L., & Coggeshall, M. (2005). Ex-offender employment programs and recidivism: a meta-

analysis. . Journal of Experimental Criminology Vol.1(3) pp.295–315

**Wilkinson, R. A. and Stickrath, T. J. (1995). Evaluation of the Impact of Correctional Education Programs on

Recidivism. Ohio Department of rehabilitation and Correction

Wilson, D. B., Gallagher, C. A., & MacKenzie, D. L. (2000). A Meta-Analysis of Corrections-Based Education,

Vocation, and Work Programs for Adult Offenders. Journal of Research in Crime and Delinquency , Vol. 37(4)

pp.347-368

**Winterfield, L., Coggeshall, M., Burke-Storer, V. C. and Tidd, S. (2009). The Effects of Postsecondary

Correctional Education: Final Report, Washington, D.C.: Urban Institute

**Zgoba, K., Haugebrook, S. and Jenkins, K. (2008). The influence of GED obtainment on prisoner release

outcomes. Criminal Justice and Behaviour, Vol.35 (3) pp.375 - 387.

55

8. APPENDICIES

APPENDIX A: DATA CAPTURED

F0 Reference

F1 Author

F2 Year

F3 Title

F4 Brief description

F5 Why partially met has now become met?

F6 Country of origin

F7 Type of report

F8 Programme Name

F9 Age

F10 Gender

F11 Ethnicity

F12 Outcomes measured (tick all that apply)

F13 Design of Study (please give details where appropriate.

Maryland Scale

F14 Sample size F15 Are there any 'sub outcomes'?

F16 What year is the data collection/follow up period? (NB, note how long the follow up period is)

F17 Does the study contain contextual information about the labour market?

F18 Does the study contain any contextual information about the relationship between the offender and the

supervisor?

F19 Does the study include any qualitative data?

F20 What is the NATURE of the intervention? (Check all that apply.)

56

F21 Type of intervention

F22 What is the CONTENT of the intervention?

F23 Duration of programme? (Should we replace with a text box? (e.g. number of hours, days, weeks, etc.)?)

F24 Intensity of programme (e.g. number of hours, days, weeks, etc.)?)

F25 Is the education targeted at a specific type of offender?

F26 WHO delivers the intervention to treated participants? (Note all that apply.)

F27 HOW is the intervention delivered? (Check all that apply.)

F28 How are the individuals who deliver the intervention employed? (Check all that apply.)

F29 Does the intervention include any post‐release follow-up programming?

F30 Was there evidence variation within the conditions (i.e., actual dosage) experienced by the TREATED group?

F31 Was there evidence of variation within the conditions (i.e., actual dosage) experienced by the COMPARISON

group?

F32 Is there evidence that researchers were independent of the intervention development and implementation?

(Check one. Note that involvement in research design is not a threat to researcher independence.)

F33 Is there evidence as stated by the author of possible contamination in the intervention implementation i.e.

possible spill-over between treatment and comparison conditions, including partial treatments? (Check one.)

F34 Does the report contain information regarding mechanisms?

F35 Is the data presented appropriate for a meta-analysis?

30 Study settings

Descriptive for Meta-analysis information intervention

Descriptive for Meta-analysis information control

rapid eviden e assessment (rea) of …...at the abstract review stage. using this tool, we reviewed...

Documents