Skip navigation

BJA - How Effective is Correctional Education

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

C O R P O R AT I O N

How Effective Is
Correctional Education,
and Where Do
We Go from Here?
The Results of a Comprehensive Evaluation

Lois M. Davis, Jennifer L. Steele, Robert Bozick, Malcolm V. Williams,
Susan Turner, Jeremy N. V. Miles, Jessica Saunders, Paul S. Steinberg

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

The research described in this report was sponsored by the Bureau of Justice
Assistance and was conducted in the Safety and Justice Program, within
RAND Justice, Infrastructure, and Environment.
This project was supported by Grant No. 2010-RQ-BX-001 awarded by the Bureau of Justice
Assistance to the RAND Corporation. The Bureau of Justice Assistance is a component of the
United States Department of Justice, Office of Justice Programs. Points of view or opinions in
this document are those of the authors and do not necessarily represent the official position or
policies of the U.S. Department of Justice.

Library of Congress Cataloging-in-Publication Data is available for this
publication.
ISBN: 978-0-8330-8493-4

The RAND Corporation is a nonprofit institution that helps improve policy
and decisionmaking through research and analysis. RAND’s publications do
not necessarily reflect the opinions of its research clients and sponsors.

Support RAND —make a tax-deductible charitable contribution at
www.rand.org/giving/contribute.html

R® is a registered trademark.
Cover images via Fuse/Thinkstock (graduate) and viperagp/Fotolia (prison background)

© Copyright 2014 RAND Corporation

This document and trademark(s) contained herein are protected by law. This
representation of RAND intellectual property is provided for noncommercial
use only. Unauthorized posting of R AND documents to a non-R AND
website is prohibited. RAND documents are protected under copyright law.
Permission is given to duplicate this document for personal use only, as long as
it is unaltered and complete. Permission is required from RAND to reproduce,
or reuse in another form, any of our research documents for commercial use.
For information on reprint and linking permissions, please see the RAND
permissions page (www.rand.org/pubs/permissions.html).
RAND OFFICES
SANTA MONICA, CA • WASHINGTON, DC
PITTSBURGH, PA • NEW ORLEANS, LA • JACKSON, MS • BOSTON, MA
CAMBRIDGE, UK • BRUSSELS, BE
www.rand.org

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Foreword

Each year, more than 2 million adults leave the nation’s prisons and jails and return to their
families and communities. While many successfully reintegrate into their communities, find jobs,
and become productive members of society, others may commit new crimes and return to jail or
prison. For juveniles involved in the juvenile justice system, the rate of youth incarceration in the
United States is more than three times the highest rates in other developed nations. Although
many factors account for why some formerly incarcerated adults and youth succeed and some
don’t, lack of education and skills is one key reason. This is why correctional education
programs—both academic and vocational—are provided in correctional facilities across the
nation. But do such correctional education programs actually work? We care about the answer
because we want ex-prisoners to successfully reenter communities and because we have a
responsibility to use taxpayer dollars judiciously to support programs that are backed by
evidence of their effectiveness—especially during difficult budgetary times like these. Across
this Administration, we are committed to investing in evidence-based programming,
investigating promising practices, and making science a priority.
Fortunately, the passage of the Second Chance Act of 2007 gave us a chance to get at this
fundamental question because it included a specific provision to improve education in U.S.
prisons and jails. The Office of Justice Programs’ Bureau of Justice Assistance within the U.S.
Department of Justice, with input from the Office of Vocational and Adult Education, U.S.
Department of Education, competitively awarded a project to the RAND Corporation in 2010 to
comprehensively examine the current state of correctional education for incarcerated adults and
juveniles, current and emerging trends in the field, and what can be done to improve the field
moving forward. The study conducted a meta-analysis and systematic review to measure the
effectiveness of correctional education for incarcerated adults and juveniles, respectively, and a
survey of states’ correctional education directors to understand concerns and emerging trends.
The results of the meta-analysis are truly encouraging. Confirming the results of previous
meta-analyses—while using more (and more recent) studies and an even more rigorous approach
to selecting and evaluating them than in the past—the study show that correctional education for
incarcerated adults reduces the risk of post-release reincarceration (by 13 percentage points) and
does so cost-effectively (a savings of five dollars on reincarceration costs for every dollar spent
on correctional education). And when it comes to post-release employment for adults—another
outcome key to successful reentry—researchers find that correctional education may increase
such employment.
Because juvenile offenders have a right to a public education, all programs for incarcerated
youth include a correctional education component. As such, effectiveness here has to focus on
describing the balance of evidence favoring the types of interventions examined. Interventions,
iii

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

methods, and outcomes of interest varied a great deal across the systematic evaluation, with
studies ultimately falling into six categories: Corrective Reading (a packaged intervention);
computer-assisted instruction (comprising three other packaged reading interventions);
personalized instruction; other remedial instruction; vocational education; and GED completion.
Though each category included only a few studies, and though the quality and size of these
studies was very limited, two interventions showed particular promise: Read 180 (for reading
improvement) and a personalized and intensive approach piloted at the Avon Park Youth
Academy in Florida (for diploma completion and post-release employment).
Because the landscape in which correctional education occurs has been altered by the recent
recession, researchers surveyed state correctional education directors to help get a pulse on what
is going on and what concerns states face. Some key insights from the survey include the
recognition that the 2008 recession and its long aftermath have had dramatic and negative effects
on correctional education spending; that there is a growing emphasis on providing vocational
education programming that will lead to industry or nationally recognized certifications; that the
importance of computer technology in correctional education is growing but use of technology is
mixed and access to the internet by incarcerated students is very limited; that states have
significant concerns about how ready they are to implement the new 2014 GED exam and
computer-based testing; and that while a large number of states are providing post-secondary
education, most is paid for by inmates or their families, not by states or the Federal government.
Overall, this study shows that the debate should no longer be about whether correctional
education is effective or cost-effective but rather on where the gaps in our knowledge are and
opportunities to move the field forward. In that vein, the study argues for a need to fund research
that both improves the evidence base that the study shows is lacking and that gets inside the
“black box” of interventions to answer questions about the dosage associated with effective
programs, the most effective models of instruction and curriculum in a correctional setting, and
who benefits most from different types of correctional education programs. Having such
knowledge is key to telling us which programs should be developed and funded—which
programs will provide the greatest return on taxpayer dollars. The study also shows the field is
ripe for larger-scale randomized trials and natural experiments that look at the impacts of
correctional education provided to juveniles and that can shed much-needed light on what works
in these settings.
And there is a need for research that investigates the impact of broader trends in corrections
for correctional education in particular. These include the increasing use of privately operated
facilities for adults and juveniles (particularly for juveniles), efforts by states to reduce the size of
their state prison population at the “front end” (for example, by reducing prison admissions) and
the “back end” (such as by parolee revocations), and the trend of keeping youth in the
community if at all possible instead of placing them in correctional institutions and of keeping
them at the local versus the state level.

iv

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

While much still remains to be done, the results are encouraging, and the findings and
recommendations in this study are intended to ensure that moving forward we understand how to
best to leverage academic and vocational education programs to improve the reentry outcomes of
incarcerated adults and juveniles. We are pleased to have been able to work cooperatively with
the RAND staff to offer this important information.
Denise E. O’Donnell, J.D.
Director, Bureau of Justice Assistance
Office of Justice Programs
U.S. Department of Justice

v

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Preface

The Second Chance Act of 2007 (Pub. L. 110-199) is a historic piece of legislation designed to
improve outcomes for and provide a comprehensive response to the increasing number of
individuals who are released from prisons, jails, and juvenile residential facilities and returning
to communities upon release. The Second Chance Act’s grant programs are funded and
administered by the Office of Justice Programs within the U.S. Department of Justice. In 2010,
for the first time, funding was set aside for a comprehensive study of correctional education. The
Office of Justice Programs’ Bureau of Justice Assistance awarded the RAND Corporation a
cooperative agreement to undertake a comprehensive examination of the current state of
correctional education for incarcerated adults and juveniles, where it is headed, which
correctional education programs are effective, and how effective programs can be implemented
across different settings. Our first report, published in 2013—Evaluating the Effectiveness of
Correctional Education: A Meta-Analysis of Programs That Provide Education to Incarcerated
Adults—presented a comprehensive review of the scientific literature and a meta-analysis that
synthesized the findings from multiple studies on the effectiveness of correctional education
programs in helping to reduce recidivism and improve postrelease employment outcomes for
incarcerated adults (Davis et al., 2013).
This final report to the U.S. Attorney General first presents a summary of the findings from
our earlier literature review and meta-analysis on the effectiveness of correctional education
programs for incarcerated adults. It also provides three new sections. The first of these is a
systematic review of correctional education programs for juveniles. The second is the results of a
national survey of state correctional education directors, which provide an up-to-date picture of
what the field of correctional education looks like today in the United States and explores the use
of computer technology, preparations for the new 2014 GED exam, and the impact of the 2008
recession. We conclude with a set of recommendations, as part of our original charge for this
study, on improvements needed to further the field of correctional education.
These results will be of interest to federal and state policymakers; administrators of state
departments of corrections, public safety, and education; correctional educators and college
educators; career technical training providers; and other organizations that provide educational
services and training to the currently incarcerated or formerly incarcerated. These results will
also be of interest to those in the U.S. Departments of Justice and Education who are committed
to ensuring the availability and quality of correctional education programs for incarcerated adults
and juveniles.

vi

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

The RAND Safety and Justice Program
The research reported here was conducted in the RAND Safety and Justice Program, which
addresses all aspects of public safety and the criminal justice system, including violence,
policing, corrections, courts and criminal law, substance abuse, occupational safety, and public
integrity. Program research is supported by government agencies, foundations, and the private
sector.
This program is part of RAND Justice, Infrastructure, and Environment, a division of the
RAND Corporation dedicated to improving policy and decisionmaking in a wide range of policy
domains, including civil and criminal justice, infrastructure protection and homeland security,
transportation and energy policy, and environmental and natural resource policy.
Questions or comments about this report should be sent to the project leaders, Lois M. Davis,
Ph.D. (Lois_Davis@rand.org) and Robert Bozick, Ph.D. (Robert_Bozick@rand.org). For more
information about the Safety and Justice Program, see http://www.rand.org/safety-justice or
contact the director at sj@rand.org

vii

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Contents

Foreword ........................................................................................................................................ iii	
  
Preface............................................................................................................................................ vi	
  
Figures............................................................................................................................................ xi	
  
Tables ............................................................................................................................................ xii	
  
Summary ...................................................................................................................................... xiv	
  
Acknowledgments..................................................................................................................... xxvii	
  
Abbreviations ............................................................................................................................. xxix	
  
1. Introduction ................................................................................................................................. 1
Overview .................................................................................................................................................. 1	
  
The Evolution of Correctional Education in the United States ................................................................ 4	
  
Organization of This Report ..................................................................................................................... 6	
  

2. How Effective Is Correctional Education for Incarcerated Adults? ........................................... 8

Introduction .............................................................................................................................................. 8	
  
Previous Meta-Analyses on Correctional Education ................................................................................ 8	
  
Lipton, Martinson, and Wilks (1975) ................................................................................................... 9	
  
Wilson, Gallagher, and MacKenzie (2000) ........................................................................................ 10	
  
MacKenzie (2006) .............................................................................................................................. 10	
  
Aos, Miller, and Drake (2006) ........................................................................................................... 11	
  
Approach ................................................................................................................................................ 11	
  
Approach to Conducting the Systematic Review of the Literature .................................................... 12	
  
Approach to Conducting the Meta-Analysis ...................................................................................... 14	
  
Meta-Analysis Findings .......................................................................................................................... 15	
  
The Relationship Between Correctional Education and Recidivism ................................................. 16	
  
The Relationship Between Correctional Education and Postrelease Employment ............................ 17	
  
The Relationship Between Correctional Education and Test Scores ................................................. 18	
  
Results of Comparing Correctional Education and Reincarceration Costs ............................................ 19	
  
Summary ................................................................................................................................................. 21	
  

3. A Systematic Review of Correctional Education Programs for Incarcerated Juveniles ........... 23

Introduction ............................................................................................................................................ 23	
  
Purpose and Organization of the Chapter ........................................................................................... 25	
  
Approach ................................................................................................................................................ 26	
  
Document Identification ..................................................................................................................... 27	
  
Eligibility Assessment ........................................................................................................................ 27	
  
Scientific Review ............................................................................................................................... 32	
  
Synthesis of Eligible Studies .............................................................................................................. 33	
  

viii

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Results .................................................................................................................................................... 35	
  
Corrective Reading ............................................................................................................................. 36	
  
Computer-Assisted Instruction ........................................................................................................... 40	
  
Personalized and Intensive Instruction ............................................................................................... 46	
  
Other Remedial Instruction Programs ................................................................................................ 51	
  
Vocational /Career Technical Education ............................................................................................ 54	
  
GED Completion ................................................................................................................................ 57	
  
Discussion ............................................................................................................................................... 60	
  
Limitations of Our Approach ............................................................................................................. 60	
  
Key Insights for the Research Community ........................................................................................ 61	
  
Key Insights for Policymakers and Practitioners ............................................................................... 62	
  

4. RAND Correctional Education Survey ..................................................................................... 64
Introduction ............................................................................................................................................ 64	
  
Approach ............................................................................................................................................ 64	
  
Results .................................................................................................................................................... 66	
  
Overview of Correctional Education Programs Today ...................................................................... 67	
  
Funding of Correctional Education and Impact of the 2008 Recession ............................................. 69	
  
Postsecondary Education .................................................................................................................... 75	
  
Use of Technology and Preparedness for Implementation of the 2014 GED Exam .......................... 76	
  
Outcome Indicators and Postrelease Measures of Success ................................................................ 82	
  
Discussion ............................................................................................................................................... 86	
  
Variation in Correctional Education Programming Across the States ............................................... 86	
  
Impact of the 2008 Recession............................................................................................................. 86	
  
Use of Information Technology ......................................................................................................... 87	
  
Readiness for the 2014 GED Exam and Computer-Based Testing .................................................... 88	
  
Postsecondary Education .................................................................................................................... 90	
  
5. Conclusions and Recommendations ......................................................................................... 91
Introduction ............................................................................................................................................ 91	
  
Correctional Education for Adults .......................................................................................................... 91	
  
Correctional Education for Juveniles ..................................................................................................... 94	
  
Improving the Evidence Base for Adult and Juvenile Correctional Education ...................................... 96	
  
Apply Stronger Research Designs ...................................................................................................... 96	
  
Measure Program Dosage................................................................................................................... 97	
  
Identify Program Characteristics ........................................................................................................ 97	
  
Examine More-Proximal Indicators of Program Efficacy.................................................................. 98	
  
Implications of Broader Trends in Corrections for Correctional Education .......................................... 98	
  
Concluding Thoughts ........................................................................................................................... 101	
  
Appendix A. Summary of Studies for the Juvenile Correctional Education Review ................. 102	
  
Appendix B. RAND Correctional Education Survey Questionnaire .......................................... 115	
  

ix

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

References ................................................................................................................................... 131	
  

x

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Figures

Figure 3.1. Eligibility Assessment of Potential Documents for Inclusion in the Systematic
Review .................................................................................................................................. 32	
  

xi

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Tables

Table 2.1. Summary of Meta-Analysis Findings by Outcome ..................................................... 15	
  
Table 3.1. Operational Definitions of Evidence Rating Categories on the What Works
Clearinghouse and Maryland Scientific Methods Scales for the Juvenile Systematic
Review .................................................................................................................................. 30	
  
Table 3.2. Summary of Findings for Reading Interventions ........................................................ 36	
  
Table 3.3. Summary of Findings for Other Juvenile Correctional Interventions ......................... 45	
  
Table 4.1. Number of States Offering Educational Programs to Adult State Prisoners,
by Type of Program .............................................................................................................. 67	
  
Table 4.2. Number of States Offering Nationally or Industry-Recognized Certifications ........... 68	
  
Table 4.3. Degree to Which Participation in Correctional Education Programs Is Mandatory.... 69	
  
Table 4.4. Mean Change in State Correctional Education Budgets Between FY2009 and
FY2012 ................................................................................................................................. 69	
  
Table 4.5. Mean Change in States’ Correctional Education Budgets Per Student Between
FY2009 and FY2012 ............................................................................................................ 70	
  
Table 4.6. Number of States Reporting Decrease in Funding Between FY2009 and FY2012 .... 71	
  
Table 4.7. Of Those States That Reported A Decrease in Funding Between FY2009 and
FY2012, Changes Made in Staffing Levels and Capacity in Response to Budget Cuts ...... 71	
  
Table 4.8. Impact on Mean Number of Students Enrolled in Academic Programs and
Vocational Education/CTE Programs, FYs 2009 and 2012 ................................................. 73	
  
Table 4.9. Mean Change in the Number of Employee or Contract Teachers by
Type of Program, FYs 2009 and 2012 .................................................................................. 74	
  
Table 4.10. Funding Sources Used to Pay for Adult Inmates’ Postsecondary Education or
College Courses .................................................................................................................... 76	
  
Table 4.11. Types of Technology Hardware and Networks Used in States’ Correctional
Education Systems ................................................................................................................ 77	
  
Table 4.12. Instruction Methods for Academic Programs or Vocational Education/CTE
Courses.................................................................................................................................. 78	
  
Table 4.13. Number of States Offering Access to the Internet for Correctional Education
Programs ............................................................................................................................... 79	
  
Table 4.14. Professional Development Training for Teachers/Instructors to Prepare Them
to Teach the New 2014 GED Exam ...................................................................................... 80	
  
Table 4.15. Concerns About Forthcoming Changes to the 2014 GED Exam and the Move
to Computer-Based Testing .................................................................................................. 81	
  
Table 4.16. Outcome Indicators Tracked by States’ Correctional Education Systems ................ 82	
  
xii

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Table 4.17. Postrelease Indicators States Consider to Be Important Outcome Measures for
Correctional Education ......................................................................................................... 83	
  
Table 4.18. Federal, State, or Private Grant Programs States’ Correctional Education
Systems Participate in ........................................................................................................... 84	
  
Table 4.20. Amount of Funding States’ Correctional Education Programs Received in
2012 from the Workforce Investment Act, Perkins Act, and States’ Higher
Educational/Aid Resources ................................................................................................... 85	
  
Table A.1. Summary of Studies, Samples, and Effects in the Systematic Review .................... 103	
  

xiii

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Summary

Introduction
Each year, more than 700,000 incarcerated individuals leave federal and state prisons; within
three years of release, 40 percent will have committed new crimes or violate the terms of their
release and be reincarcerated. Although a number of factors impede the ability of ex-offenders to
successfully reintegrate into communities and, thus, affect recidivism rates, one key factor is that
many ex-offenders do not have the knowledge, training, and skills to support a successful return
to their communities. Research, for example, shows that ex-offenders, on average, are less
educated than the general population: 37 percent of individuals in state prisons had attained less
than a high school education in 2004, compared with 19 percent of the general U.S. population
age 16 and over; 16.5 percent of state prisoners had just a high school diploma, compared with
26 percent of the general population; and 14.4 percent of state prison inmates had at least some
postsecondary education, compared with 51 percent of the general U.S. adult population.
Moreover, literacy levels for the prison population also tend to be lower than that of the general
U.S. population.
This lower level of educational attainment represents a significant challenge for ex-offenders
returning to local communities, because it impedes their ability to find employment. A lack of
vocational skills and a steady history of employment also have an impact, with research showing
that incarceration impacts unemployment and earnings in a number of ways, including higher
unemployment rates for ex-offenders and lower hourly wages when they are employed. Also,
individuals being released to the community face a very different set of job market needs than
ever before, given the growing role of computer technology and the need for at least basic
computer skills.
Given these gaps in educational attainment and vocational skills and the impact they have on
ex-offenders, one strategy is to provide education to inmates while they are incarcerated, so that
they have the skills to support a successful return to their communities. Historically, support for
educational programs within correctional settings has waxed and waned over time as the nation’s
philosophy of punishment has shifted from rehabilitation to crime control.
Although there is general consensus today that education is an important component of
rehabilitation, the question remains: How effective is it in helping to reduce recidivism and
improve postrelease employment outcomes? The question is especially salient as the nation as a
whole and states in particular have struggled with the need to make spending cuts to all social
programs due to the recession of 2008 and its long aftermath. With funding from the Second
Chance Act of 2007 (Pub. L. 110-199), the U.S. Justice Department’s Bureau of Justice

xiv

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Assistance awarded RAND a cooperative agreement in 2010 to comprehensively examine the
current state of correctional education for incarcerated adults and juveniles, where it is headed,
which correctional education programs are effective, and how effective programs can be
implemented across different settings.
The study was designed to address the following key questions of importance to the field of
correctional education:
1. What is known about the effectiveness of correctional education programs for
incarcerated adults?
2. What is known about the effectiveness of correctional education programs for juvenile
offenders?
3. What does the current landscape of correctional education look like in the United States,
and what are some emerging issues and trends to consider?
4. What recommendations emerge from the study for the U.S. Department of Justice and
other federal departments to further the field of correction education, and where are there
gaps in our knowledge? What promising practices, if any, emerge from this review and
evaluation?
To address these questions, we used a mixed-methods approach. This report first presents a
summary of the prior systematic literature review and meta-analysis of adult correctional
education programs (Davis et al., 2013) which included studies completed between 1980 and
2011. It then presents three new sections: a systematic literature review of primary studies of
correctional education programs for juveniles, and a nationwide web-based survey of state
correctional education directors. We conclude with a set of recommendations for moving the
field forward. For purposes of our study, we defined correctional education for incarcerated
adults as including the following:
•
•
•
•

Adult basic education: basic skills instruction in arithmetic, reading, writing, and, if
needed, English as a second language (ESL)
Adult secondary education: instruction to complete high school or prepare for a
certificate of high school equivalency, such as the General Education Development
(GED) certificate
Vocational education or career technical education (CTE): training in general
employment skills and in skills for specific jobs or industries
Postsecondary education: college-level instruction that enables an individual to earn
college credit that may be applied toward a two- or four-year postsecondary degree.

To meet our definition of correctional education, the program had to be administered at least
partly within a correctional facility. Programs that also included a postrelease transition
component remained eligible as long as part of the program was administered in a correctional
setting.
For the juvenile program systematic review, we define incarcerated youth as individuals
under age 21 who are legally assigned to correctional facilities as a result of arrest, detainment
for court proceedings, adjudication by a juvenile court, or conviction in an adult criminal court

xv

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

(Office of Juvenile Justice and Delinquency Prevention, 2013). We define correctional
education as any academic or vocational education/CTE program provided within the
correctional facility setting, regardless of jurisdiction. As with our adult review, we permitted
eligible interventions for juveniles to include an aftercare (postrelease) component, but the
interventions had to be delivered primarily in the correctional facility.

How Effective Are Correctional Education Programs for Incarcerated
Adults?
In terms of the effectiveness of correctional education programs for incarcerated adults, early
reviews in this area found inconclusive evidence to support their efficacy—a finding that
contributed to the popular belief that “nothing works” in prisoner rehabilitation; however, this
conclusion may have been premature, given that appropriate analytic techniques had not been
developed. More recent reviews, using meta-analysis techniques, question the conclusions of the
earlier work, finding evidence of a relationship between correctional education program
participation before release and lower odds of recidivating after release (Wilson, Gallagher, and
MacKenzie, 2000; MacKenzie, 2006; Aos, Miller, and Drake, 2006). However, the most recent
meta-analyses (Aos, Miller, and Drake, 2006; MacKenzie, 2006) did not consider employment
outcomes; thus, whether program participation is associated with postrelease success in the labor
market remained unclear.
These earlier reviews provide the context for the current systematic review and metaanalysis. Our systematic review scanned the universe of potential documents to compile all
available empirical research studies that examine the effect of correctional education programs
on the three outcomes of interest—recidivism, postrelease employment, and reading and math
scores. This search yielded 1,112 documents, of which 267 were identified as primary empirical
studies. To be in our meta-analysis, the study needed to meet three eligibility criteria: (1)
evaluate an eligible intervention, defined here as an educational program administered in a jail or
prison in the United States published (or released) between January 1, 1980, and December 31,
2011; (2) measure the effectiveness of the program using an eligible outcome measure, which for
our meta-analysis included recidivism, postrelease employment, and achievement test scores;
and (3) have an eligible research design, which, for our purposes, is one where there is a
treatment group comprising inmates who participated in or completed the correctional education
program and a comparison group of inmates who did not.
Of the 267 primary empirical studies, 58 met all three eligibility criteria.1 With respect to
recidivism, based on the higher-quality research studies, we found that, on average, inmates who
participated in correctional education programs had a 43 percent lower odds of recidivating
1

Our recidivism analysis is based on 71 effect sizes from 50 studies, our employment analysis is based on 22 effect
sizes from 18 studies, and our test score analysis is based on nine effect sizes from four studies.

xvi

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

than inmates who did not, thus indicating that correctional education is an effective strategy for
reducing recidivism.2 This estimate is based only on 9 effect sizes from studies that met higher
levels of rigor (i.e., earned 4s or 5s on the Maryland Scientific Methods Scale), but the results
were very similar even when the lower-quality studies were included in the analysis. This
translates to a reduction in the risk of recidivating of 13 percentage points for those who
participated in correctional education programs versus those who did not.
When aggregating across 18 studies that used employment as an outcome, we found that the
odds of obtaining employment postrelease among inmates who participated in correctional
education (either academic or vocational/CTE programs) were 13 percent higher than the odds
for those who did not. However, the findings are only suggestive about whether correctional
education is an effective strategy in improving postrelease employment outcomes because only
one of the 18 studies were of higher quality (level 4 or higher), thus limiting our ability to make a
more definitive statement.
When aggregating across four studies that used achievement test scores as an outcome, we
found that learning gains in both reading and in math among inmates exposed to computerassisted instruction were similar to learning gains made by inmates taught through traditional
(face-to-face) instruction methods.3
Although doing a formal cost-effectiveness analysis was beyond the scope of this study, we
wanted to provide some context for what the meta-analysis findings mean. Focusing on the
outcome of recidivism and using a hypothetical pool of 100 inmates, we compared the direct
costs of correctional education programs and of incarceration itself. We found that the direct
costs of providing correctional education far outweighed the direct costs of reincarceration.
More specifically, for a correctional education program to be cost-effective—or break-even—we
estimated that it would need to reduce the three-year reincarceration rate by between 1.9
percentage points and 2.6 percentage points. Given that our findings indicate that participation in
correctional education programs is associated with a 13-percentage-point reduction in the risk of
reincarceration three years following release, correctional education programs appear to far
exceed the break-even point in reducing the risk of reincarceration. We also note that the results
are likely to be conservative, because they do not include the indirect costs of reincarceration.

How Effective Are Correctional Education Programs for Juvenile
Offenders?
When it comes to assessing correctional education programs for juvenile offenders, we face a
fundamental difference between juvenile and adult correctional policy: Juveniles in the United
2

We define recidivism was defined a number of ways in the individual studies reviewed, including re-offending, rearrest, re-conviction, re-incarceration, technical parole violation, and successful completion of parole. In our pool of
50 studies that had recidivism outcomes, the majority used re-incarceration as the outcome measure (n = 34).
3

Three of the four studies, representing seven effect sizes, were of higher quality.

xvii

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

States have a right to a public education. Therefore, all programs for incarcerated youth include
an educational component. This means that the question facing policymakers is not whether to
provide education services for juveniles in correctional facilities, but which types of programs
are most effective. The meta-analytic approach in our adult analysis included many types of
correctional education, each of which was compared with a no-correctional-education scenario.
However, that approach is less well suited to studying the effectiveness of juvenile correctional
education programs, because correctional education programs are typically present in all juvenile
facilities. Instead, our approach to synthesizing research on juvenile correctional education was
to undertake a systematic review, in which we screened and evaluated articles using the same
criteria as we used in our adult meta-analysis. But rather than aggregating estimated effect sizes
across studies that are testing widely different hypotheses for treatment versus comparison
groups, we focus on describing the balance of evidence favoring the types of interventions
examined in the literature we reviewed.
Altogether, the document search process resulted in 1,150 citations for title-and-abstract
screening, which, in turn, led to 157 manuscripts eligible for full-text screening. Of those, 18
studies were deemed eligible for the systematic review by having met three criteria: (1) be an
eligible intervention, defined as any academic or vocational education/CTE intervention
program, with an eligible population, defined as consisting primarily of individuals age 20 or
below, in an eligible setting, defined as any facility regardless of jurisdiction (i.e., local, state) to
which juveniles are confined because of arrest, court proceedings, or adjudication/conviction; (2)
use eligible outcome measures, defined as any measure of recidivism (e.g., re-arrest,
reconviction, or re-incarceration), postrelease employment, academic attainment (e.g., GED or
high school completion), and academic performance (e.g., test scores in reading and
mathematics); and (3) have an eligible research design. This includes a comparison-group
design in which a group of incarcerated juveniles who received an intervention is compared with
a group of incarcerated juveniles that did not, or who received a different version of the
treatment. We also included a class of approaches called single-case designs, which involve
systematically introducing an intervention with one or a few students in an effort to demonstrate
causal effects on outcomes such as participant behavior or learning. These studies typically
include a large number of pre- and post-intervention outcome measurements, allowing students
to function as their own controls. Because these designs usually focus on only a handful of
participants, they typically preclude traditional hypothesis testing. This means that their findings
cannot be generalized to a larger, hypothetical population of interest.
Our systematic review reveals great heterogeneity in terms of interventions, methods, and
outcomes of interest. Among the 18 eligible studies we identified (16 comparison-group studies
and two well-executed single-case design ones), we classified the interventions into six
categories—Corrective Reading (a commercially packaged curriculum), computer-assisted
instruction, personalized instruction, other remedial education, vocational education, and GED
completion. Studies in the first two categories focused on packaged and branded reading
xviii

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

interventions (Corrective Reading, Read 180, Fast ForWord, and Tune in to Reading) and
focused on reading performance as the dependent variables of interest. Studies in the latter three
categories focused on a broader set of outcomes, including not only reading and mathematics
performance but also measures such as diploma completion, postrelease employment, and
postrelease recidivism.
Given the small number of studies in each category, we cannot easily extrapolate the effects
of differential dosages or implementation approaches. However, taken in conjunction with the
broader research literature on each of the interventions examined, our systematic review does
identify two interventions that show particular promise: Read 180 (for reading improvement)
and Florida’s Avon Park Youth Academy (for diploma completion and postrelease employment).
Both of these interventions are supported by a large and rigorous study within juvenile
correctional settings, and the effectiveness of Read 180 is further substantiated by several large
and well-executed studies outside of correctional facilities. Beyond these compelling studies, we
find that evidence for two other packaged interventions, Corrective Reading and Tune in to
Reading, is positive, but the underlying studies are too small to warrant generalization. Evidence
concerning vocational education/CTE and GED completion is also positive, but the underlying
research designs are vulnerable to selection bias. This limits the quality of conclusions that can
be drawn about these programs.

What Is the Current Correctional Education Landscape and What Trends
Are Important?
When we began the correctional education study, we recognized early on that the 2008 recession
had a substantial effect on the field of correctional education, with many states reporting cuts in
funding for programs and changes to their delivery models for educating incarcerated adults.
This means that today correctional education in the United States likely looks very different from
correctional education during the time that many of the studies in the meta-analysis and adult
systematic review were undertaken. Understanding these differences helps us to put our review
results in context and provides the basis for forward-looking policy recommendations.
In July 2013, we fielded the RAND Correctional Education Survey to help fill a critical void
in our understanding of the organization and delivery of academic and vocational education/CTE
to incarcerated adults. This web-based survey of correctional education directors in all 50 states
provides us with insights into how states dealt with the recession of 2008, how correctional
education is currently provided to incarcerated adults in the United States, what technology is
being used, and how states fund correctional education. We also gathered information on
preparations for the new 2014 GED exam. We classified the size of state by the adult prison
population in 2012 and considered small states to have an adult prison population in the range
(1–24,999); medium states, in the range (25,000–49,999); and large states, 50,000 or more adult
prisoners. The overall response rate was 46 out of 50 states, or 92 percent. Of these 46, 42
xix

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

completed the entire questionnaire, and four provided only partial responses to the survey. Forty
of the respondents had responsibility for both adult correctional education and vocational
training in their state; five respondents for academic education only; and one respondent for
vocational training only.
Variation in Correctional Education Programming Across the States
In 2013, most states offered adult basic education, GED courses, and vocational education/CTE
programs, and most reported having special education courses available. Higher-level
educational programming such as adult secondary education and postsecondary education was
offered in 32 of the states, although smaller states were less likely to do so. Postsecondary
education courses today in 28 states are primarily paid for by the individual inmate or by family
finances; in 16 states, state funding is used to cover the costs of postsecondary education, and 12
states reported using college or university funds.
Participation in correctional education programs is mandatory in 24 states for adult inmates
without a high school diploma or GED, and in 15 states it is mandatory for adults below a certain
grade level, with smaller states less likely to require mandatory participation.
An emerging trend is a growing emphasis on providing vocational education/CTE
programming that will lead to industry or nationally recognized certifications. Smaller states
were more likely to emphasize vocational education/CTE training for state prisoners than
medium-sized or large states.
Impact of the 2008 Recession
The effect of the 2008 recession was an overall 6 percent decrease on average in states’
correctional education budgets between fiscal years (FYs) 2009 and 2012. The largest impact on
budgets was felt by medium-sized and large states (on average, a 20 percent and 10 percent
decrease, respectively). Overall, the mean dollars spent per student for correctional education
was $3,479 in FY 2009, compared with $3,370 in FY 2012—this represented a 5 percent
decrease on average in the dollars spent per student.
The result was a contraction in the capacity of academic education programs and an overall
decrease of 4 percent on average in the number of adult students who participated in these
programs, with medium-sized and large states experiencing somewhat larger decreases (10
percent and 8 percent, respectively, compared with a 1 percent decrease for small states). In
addition, 20 states reduced the number of course offerings for academic programs during this
time period.
The effect of the staffing and capacity cost-cutting measures on teachers for academic
programs was particularly felt in medium-sized and large states. Overall, there was, on average, a
4 percent decrease in the number of academic teachers who were employees. The largest
decrease occurred in medium-sized and large states (on average, 44 percent and 20 percent,
respectively, compared with a 5 percent decrease for small states).
xx

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Vocational education/CTE programs seemed to have fared somewhat better during the
recession than academic programs in terms of reductions in the number of students enrolled in
vocational education/CTE programs and in the number of instructors. On average, there was a 1
percent increase in the number of students enrolled in vocational/CTE programs between
FYs2009 and 2012. However, this appears to be largely driven by an increase, on average, of 7
percent within the smaller states. In comparison, the medium-sized and large states experienced a
reduction in the number of students in vocational education/CTE programs, on average, of 4
percent and 11 percent, respectively. There also appears to have been a modest expansion of
vocational education/CTE programs in small and medium-sized states during this time period, as
evidenced by a modest increase between FYs 2009 and 2012 in the number of vocational
education/CTE instructors who were employees (on average 8 percent and 24 percent,
respectively for small and medium-sized states). Still, 38 percent of small states and 50 percent
of medium-sized states reported that they had reduced the number of course offerings for
vocational education/CTE programs in response to budget cuts.
Use of Information Technology
One of the major trends that will shape the future of work in the 21st century is the growing role
of information technology in society, with technological change resulting in an increased
demand for a skilled workforce (Karoly, 2013). Further, distance learning and online instruction
are growing trends in the United States, with increasingly more educational courses being
offered online either by colleges or virtual high schools. Computer-assisted instruction is also
appealing in offering the opportunity to tailor instruction and coursework to the needs of the
individual student.
The importance of computing skills for today’s job market is recognized by state correctional
education directors, as reflected by the fact that 24 states reported offering a Microsoft Office
certification as part of their vocational education/CTE programs. However, our survey results
indicate that the role of computer technology in correctional education is a mixed story. We
found that the use of computers for instructional purposes is common, with 39 states reporting
the use of desktop computers (either standalone or networked) and 17 states reporting the use of
laptops. However, access to the Internet and the use of Internet-based instruction (one-way or
interactive) is reported to be limited in states’ correctional facilities. Thirty states reported that
only teachers and instructors have access to live Internet technology. In 26 states, inmate
students lack access to any Internet technology, and in only 16 states do inmate students have
access to simulated Internet programs. In terms of instructional methods that use some type of
technology, only ten states reported that they had closed-circuit television, and only a few states
reported using it to provide one-way or interactive video/satellite instruction.

xxi

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Readiness for the 2014 GED Exam and Computer-Based Testing
The GED is the predominant way that inmates earn their high school equivalency diplomas
(Harlow, 2003), and GED completion is often a prerequisite for many vocational training
programs. The 2014 GED exam not only represents a more rigorous test, being aligned with the
Common Core State Standards (CSS), but it will also rely on a new test delivery model—
namely, computer-based testing to replace the old paper-and-pencil exam (Lockwood et al.,
2013). This represents a profound change to states and one that presents some key challenges.
GED completion rates were seen as an important outcome indicator to track by 40 states that
took part in our survey. Yet, of the 31 states planning to implement the 2014 GED exam, 14
states expected that the more rigorous GED exam and the use of computer-based testing may
have a negative effect on the number of adult inmates who will be prepared to take the new
exam, and 16 states expected a negative effect on GED completion rates. This was particularly
true for the medium-sized and large states.
All but two of the 31 states planning to implement the 2014 GED exam expressed concerns
about the new exam and computer-based testing. Nineteen states were concerned about their
teachers being adequately prepared to teach the new GED exam, and 24 of the states were
concerned about the length of time it may take to prepare students for the more rigorous exam. In
addition, 12 of the states reported concerns that limited access to computers may preclude some
students from taking the new GED exam. Also, responding directors in 14 of the states reported
concerns that their teachers may not be adequately prepared to implement computer-based
testing. Other concerns expressed were the cost to the individual student and the cost of the new
GED exam to their institutions, with some states considering the adoption of alternative high
school equivalency exams. In general, smaller states expressed fewer concerns; however, our
survey results suggest that states with the majority of the prison population (i.e., medium-sized
and large states) expect to encounter a number of challenges in implementing the new GED
exam and test delivery system.

What Are Some Key Recommendations for Moving Forward?
This study’s key finding is that correctional education is effective in reducing recidivism for
incarcerated adults and that there is some evidence that it also is effective, especially vocational
education/CTE programs, in improving individuals’ likelihood of postrelease employment. Also,
our cost analysis showed that correctional education is highly cost-effective for incarcerated
adults: For every dollar spent on correctional education, five dollars are saved on three-year
reincarceration costs. As is the case for adults, our report also provides the most comprehensive
systematic review we are aware of on what works in correctional education for incarcerated
juveniles. Thus, the debate should no longer be about whether correctional education is effective
or cost-effective; rather, the debate should focus on where the gaps in our knowledge are and
opportunities to move the field forward.
xxii

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

We offer some recommendations and next steps—that are drawn from our evaluation results;
while this report is to the U.S. Attorney General, these recommendations will also be of interest
to other federal departments and agencies focused on reentry. These recommendations are
intended to provide a roadmap for building on the gains made to-date in educating incarcerated
individuals to improve their chances of success upon release and reentry into local communities.
Correctional Education for Adults
Our survey results provide solid evidence about the dramatic impact the 2008 recession had on
correctional education in the United States. The recession and its long aftermath led to a
reduction in correctional education spending and a decrease in the number of incarcerated adults
who participate in these programs. This raises the question of whether the trade-offs we are
making in terms of cost savings today with reductions in educational programming are
worthwhile, considering the future costs of reincarceration and the effect that such lost
opportunities may have on individuals’ chances of finding employment and being successful in
reintegrating back into society. State corrections directors want to know how they can modify
their models of education to trim their budgets while still maintaining the effectiveness of their
programs. The results of our meta-analysis (Davis et al., 2013) show that correctional education
programs are dramatically effective in reducing recidivism, and there is some evidence of
improvements to postrelease employment outcomes. We also showed that correctional education
programs are highly cost-effective for incarcerated adults. But because of limitations in quality
of the evidence base, we cannot answer the other critical questions needed to inform discussions
about modifications to educational programming in a resource-constrained environment. We
concur with MacKenzie’s (2008) assessment that we still are unable to get at what is inside the
“black box” of what works in correctional education, to answer such questions as:
•
•
•
•

What dosage is associated with effective programs, and how does it vary for different
types of academic programs and students?
What models of instruction and curriculum delivery (e.g., one-on-one, traditional
classroom lectures, computer-based learning) are most effective in a correctional
environment?
Who benefits most from different types of correctional education programs?
What principles from adult education and learning may be applicable to correctional
education?

Thus, we recommend the following to help address these concerns:
•
•

Focus research and evaluation efforts at the federal and state levels to address these
questions so that policymakers and state correctional education directors can make
informed trade-offs in budget discussions.
Have federal and state governments and philanthropy fund (1) evaluations of programs
that illustrate different educational instructional models, with the goal of getting inside
the black box; (2) evaluations of programs that are trying innovative strategies to
implement technology and leverage distance learning in the classroom; and (3) an
xxiii

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

•

•
•

analysis of what lessons from the larger literature on adult education may be applied to
correctional education.
Have the federal government monitor and assess the impact of the new GED and
computer-based testing on correctional education implementation and outcomes.
Consider opportunities to provide technical assistance to states in helping educators teach
the material for the more-rigorous content in the new GED. In juvenile correctional
settings, technical assistance for implementing the new Common Core State Standards,
which have influenced the move toward a more-rigorous GED, is also likely to be
needed.
Conduct new research on instructional quality in correctional education settings, and on
ways to leverage computer technology to enhance instruction in correctional settings.
Given the changes in the U.S. economy and the shifting needs of the 21st century
workforce, conduct an assessment at the federal and state levels about what such changes
mean for the criminal justice–involved population. Consider a summit at the state and
federal levels with private industry about what opportunities are available to formerly
incarcerated individuals and what skills will be needed in the future.

Correctional Education for Juveniles
Based on our systematic review of the literature on education provided to juveniles in
institutional setting, we believe that the field is ripe for larger-scale randomized trials. Two of the
studies we reviewed, Loadman et al.’s (2011) Read 180 study and the National Council on Crime
and Delinquency’s (2009) Avon Park study, suggest that such studies, though challenging to
undertake, are feasible. The literature is also ripe for rigorous evaluations of natural experiments,
such as Aizer and Doyle’s (2013) study of the effects of juvenile incarceration using naturally
occurring random assignment to harsh judges. Studies that take advantage of rigorous causal
methods in juvenile settings can shed much-needed light on what works in these settings. Several
of the smaller randomized trials we include in our review have noted the difficulties of high
student turnover in correctional facilities and of simply gaining permission to undertake research
in these facilities (Shippen et al., 2012, Calderone et al., 2009). As such, we recommend that
the focus be on developing larger-scale randomized trials and rigorous evaluations of
natural experiments. Such research efforts will clearly take time to develop and execute. They
will ideally be realized through long-term partnerships between researchers and correctional
facilities. Because such partnerships take time to establish, there may also be a federal role in
galvanizing them. The U.S. Department of Education Institute of Education Science’s recent
grant program for supporting research partnerships between school systems and researchers
offers one potential model. Informed by such partnerships, facilities can make increasingly
evidence-based decisions that not only improve their students’ prospects but also reduce the
social incidence of crime and delinquency.

xxiv

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Improving the Evidence Base
In our meta-analytic report (Davis et al., 2013), we laid out a number of recommendations to
improve the evidence base for adult correctional education. Those recommendations also pertain
to juvenile correctional education and merit summarizing here. We recommend that the federal
and state governments and philanthropy invest in well-designed evaluations of correctional
education programs and use funding and grant mechanisms to encourage improvements in
four areas to further develop the evidence base for correctional education:
•

•

•
•

Apply stronger research designs to help establish a causal relationship between
correctional education participation and successful outcomes for inmates to help rule out
the possibility of selection bias. In this context, identifying the appropriate comparison
groups is important, as is establishing a study registry to help sort out the different effect
sizes found across studies.
Measure program dosage to help put the findings from individual studies in their proper
contexts. The lack of dosage information means that there is little to no empirical
evidence that can help inform policymakers on “how much” correctional education is
necessary to produce a change in the desired outcomes.
Identify program characteristics to help policymakers identify promising or evidencebased programs that could be potentially replicated in other settings and specific
exemplary programs.
Examine more proximal indicators of program efficacy to help better refine the
process through which correctional education helps shape how former inmates reintegrate into the community. This includes understanding how improving the skills and
abilities of inmates (i.e., “human capital” in economics parlance) could, in turn, improve
former inmates’ chances of continuing education/training upon release and then finding
gainful employment.

In addition, a study registry of correctional education evaluations would further aid in
developing the evidence-base in this field to help inform policy and programmatic
decisionmaking.
Implications of Broader Trends in Corrections for Correctional Education
Several trends in the field of adult and juvenile corrections have important implications for
correctional education that merit further consideration. First, many states are undertaking
measures to reduce the size of their state prison population using a variety of means. This
includes both “front-end” strategies—such as reducing prison admissions, diverting offenders to
county- rather than state-level institutions, or changing felonies to misdemeanors—and “backend” strategies—such as reducing sentence lengths through earned credits or good time and
revocations for probationers and parolees. All these changes in the correctional landscape have
implications for how we think about providing academic education and vocational
education/CTE to incarcerated adults. For example, there is California’s Public Safety
Realignment, where county jails now have some inmates serving sentences of two, three, or

xxv

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

more years instead of the typical length of stay of 2–3 months. However, county jails are not set
up to provide rehabilitative services over the long-term, including academic programs and
vocational/CTE programs. The result of various states’ and localities implementing strategies to
keep low-level offenders at the local level is that, in some instances, we may end up with a twotiered system of education where, ironically, more serious offenders who serve their sentence in
state prison may have better access to correctional education programs than low-level offenders
who serve their sentences in county jails. Such policy changes also raise other questions: Are
there differences in access to academic education and vocational education/CTE programs
depending on the setting where one serves one’s sentence? Are there differences in educational
and employment outcomes for offenders who serve their time at the local level compared with
offenders serving their sentence in state prison systems?
Second, a long-term trend in the field of juvenile corrections is to keep youth in the
community if at all possible instead of placing them in correctional institutions, and to keep them
at the local versus the state level. Our systematic review focused on what works with
incarcerated youth in part because the broader literature on educational interventions for juvenile
offenders outside of correctional facilities is even more nebulous. An important direction for
future research is to identify interventions that improve juveniles’ educational, employment, and
recidivism outcomes in less-restrictive settings, such as alternative schools or traditional schools.
To guide policy improvements, stronger federal reporting requirements about local correctional
education practices could help facilitate improved state and local comparisons of program
effects. Whether collected federally or privately, a central repository of such information (e.g.,
staffing levels and expertise, curriculum used, hours of instruction provided, types of programs
offered) would provide a valuable tool to policymakers and researchers alike.
We recommend that policymakers seek to assess and understand the implications of
these trends in the field of corrections with respect to their impact on correctional
education.

Concluding Thoughts
There are more than 2 million adults incarcerated in the United States. This study demonstrates
that education programs can help adults get back on their feet upon release from prison and help
juveniles involved with the juvenile justice system to continue with their education. Education
programs are also highly cost-effective in helping to reduce recidivism. States will continue to
operate in a reduced funding environment for the near future. The findings and recommendations
we present here are intended to ensure that, moving forward, we understand how best to leverage
academic education and vocational education/CTE programs to improve the reentry outcomes of
incarcerated adults and juveniles.

xxvi

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Acknowledgments

We are particularly grateful for the guidance and feedback provided throughout this project by
our Bureau of Justice Assistance project officers, Gary Dennis, senior policy advisor for
corrections, and Thurston Bryant, policy advisor. We are also grateful for the valuable input and
feedback provided by Brenda Dann-Messier, Assistant secretary for vocational and adult
education, and John Linton, director, Office of Correctional Education, Office of Vocational and
Adult Education, U.S. Department of Education. We also appreciate the support and insights
provided by Steve Steurer, executive director of the Correctional Education Association.
The overall direction of the project was guided in part by a steering committee that included
John Dowdell (director of the Gill Center for Business and Economic Education at Ashland
University and co-editor of the Journal of Correctional Education), William Sondervan
(professor and director of Criminal Justice, Investigative Forensics, and Legal Studies at the
University of Maryland University College), Stephen Steurer (executive director of the
Correctional Education Association), and Susan Turner (professor of criminology, law, and
society at the University of California–Irvine).
We wish to express our appreciation to the state correctional education directors who
throughout this project gave of their time and provided thoughtful input in numerous discussions
as well as through their participation in the RAND Survey of State Correctional Education
Directors. Their survey responses provided valuable information and insights on the field of
correctional education today. We also wish to thank the Correctional Education Association’s
State Council of Directors and especially Tammy Barstad for their assistance with the survey.
John Linton, John Dowdell, Steve Steurer, and Jerry McGlone reviewed several drafts of the
questionnaire.
In addition, a number of individuals within and outside RAND contributed to various aspects
of the project. The Scientific Review Team members helped guide the selection of intervention
characteristics to be abstracted and served as independent reviewers in abstracting the study
information that were inputs for the adult meta-analysis. They included Cathryn Chappell
(Ashland University), John Dowdell (Ashland University), Joseph Gagnon (University of
Florida), Paul Hirschfield (Rutgers University), Michael Holosko (University of Georgia), David
Houchins (Georgia State University), Kristine Jolivette (George State University), Larry
Nackerud (University of Georgia), Ed Risler (University of Georgia), and Margaret Shippen
(Auburn University). RAND members of the Scientific Review Team included Ramya Chari,
Sarah Greathouse, Lisa Sontag-Padilla, and Vivian Towe. Before manuscripts were sent to the
Scientific Review Team, they were screened for preliminary eligibility by Pardee RAND
Graduate School students, including Nono Ayivi-Guedehoussou, Stephanie Chan, Megan
Clifford, Russell Lundberg, Shannon Maloney, Christopher McLaren, and Nicole Schmidt. One
xxvii

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

screener, Lopamudra Das, also provided invaluable assistance in the abstraction of juvenile
outcome data. Within RAND, the web-based survey was programmed and fielded by Julie
Newell and Diana Malouf of RAND’s Multimode Interviewing Capability (MMIC), and
RAND’s Correctional Education website was developed by Sue Phillips. Dionne Barnes-Proby
provided project management and research assistance, Jamie Greenberg summarized statistics on
justice-involved youth, and Judy Bearer provided administrative support. The project benefited
from suggestions by Paul Heaton, Suanne Hempel, Becky Kilburn, and Brian Jackson at RAND,
and from the editing and publications production support provided by James Torr and Jocelyn
Lofstrom.
Finally, we appreciate the insights provided by our technical reviewers, Lynn Karoly, a
senior economist at RAND, and Gerald Gaes, visiting faculty, Florida State University.

xxviii

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Abbreviations

BJA

Bureau of Justice Assistance

GED
Maryland Scale
RD
RYEF
SCA
WWC

General Education Development
Maryland Scientific Methods Scale
regression discontinuity
Regional Youth Educational Facility
Second Chance Act
U.S. Department of Education’s What Works Clearinghouse

xxix

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

1. Introduction

Overview
Each year, more than 700,000 incarcerated individuals leave federal and state prisons; within
three years of release, 40 percent will have committed new crimes or violated the terms of their
release and be reincarcerated. Although a number of factors impede the ability of ex-offenders to
successfully reintegrate into communities and, thus, affect recidivism rates, one key factor is that
ex-offenders do not have the knowledge, training, and skills to support a successful return to
their communities. Research, for example, shows that ex-offenders, on average, are less educated
than the general population:
•
•
•

37 percent of individuals in state prisons had attained less than a high school education in
2004, compared with 19 percent of the general U.S. population age 16 and over.
16.5 percent of state prisoners had a high school diploma, compared with 26 percent of
the general population.
14.4 percent of state prison inmates had at least some postsecondary education, compared
with 51 percent of the general U.S. adult population (Crayton and Neusteter, 2008).

Moreover, literacy levels for the prison population also tend to be lower than that of the
general U.S. population. The 2003 National Assessment of Adult Literacy (NAAL) assessed the
English literacy of 1,200 inmates (ages 16 and older) in state and federal prisons and a sample of
18,000 adults (ages 16 and older) living in U.S. households. On average, adult inmates had lower
scores on all three literacy scales—prose, document, and quantitative—than the general U.S.
population (Greenberg, Dunleavy, and Kutner, 2007).
This lower level of educational attainment represents a significant challenge for ex-offenders
returning to local communities, because it impedes their ability to find employment. A lack of
vocational skills and a steady history of employment also have an impact, with research showing
that incarceration impacts unemployment and earnings in a number of ways, including higher
unemployment rates for ex-offenders and lower hourly wages when they are employed.
The dynamics of prison entry and reentry make it difficult for this population to accumulate
meaningful, sustained employment experience (Raphael, 2007–2008). For example, using data
from the Fragile Families and Child Wellbeing Study, an analysis of the effects of incarceration
on the earnings and employment in a sample of poor fathers found that the employment rates of
formerly incarcerated men were about 6 percentage points lower than those for a similar group of
men who had not been incarcerated (Gellar, Garfinkel, and Western, 2006).
The stigma of having a felony conviction on one’s record also is often a key barrier to
postrelease employment (Pager, 2003). Holzer, Raphael, and Stoll (2003) conducted a series of
surveys of employers in four major U.S. cities and found that employers were much more averse

1

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

to hiring ex-offenders than they were in hiring any other disadvantaged group. Willingness to
hire ex-offenders was greater for jobs in construction or manufacturing than for those in the retail
trade and service sectors; employers’ reluctance was greatest for violent offenders as opposed to
nonviolent drug offenders.
Finally, individuals being released to the community face a very different set of demands for
skill sets in today’s job market than ever before, with the growing importance of information
technology and the need for basic computer skills (Karoly, 2013).
Given these gaps in educational attainment and vocational skills and the impact they have on
ex-offenders, one strategy is to provide education to inmates while they are incarcerated so that
they have the skills to support a successful return to their communities. Prisoner education (or
correctional education) in the United States dates back more than 200 years. Support for
educational programs within correctional settings has waxed and waned over time as the nation’s
philosophy of punishment has shifted back and forth from rehabilitation to crime control.
Although the general consensus today is that education is a useful component of the
rehabilitation process, the question remains, how useful is it? This question has been particularly
salient as the nation as a whole and states have struggled with the need to make spending cuts to
all social programs throughout the recession of 2008 and its long aftermath.
On April 9, 2008, the Second Chance Act (SCA) (Pub. L. 110-199) was signed into law. This
important piece of legislation was designed to improve outcomes for individuals who are
incarcerated, most of whom will ultimately return to communities upon release. The SCA’s grant
programs are funded and administered by the Office of Justice Programs within the U.S.
Department of Justice. In 2010, funding was set aside under the SCA to, for the first time,
conduct a comprehensive study of correctional education. The Office of Justice Programs’
Bureau of Justice Assistance (BJA) awarded the RAND Corporation a cooperative agreement to
comprehensively examine the current state of correctional education for incarcerated adults and
juveniles, where it is headed, which correctional education programs are effective, and how
effective programs can be implemented across different settings.
Our study was designed to address the following key questions of importance to the field of
correctional education:
1. What is known about the effectiveness of correctional education programs for
incarcerated adults?
2. What is known about the effectiveness of correctional education programs for juvenile
offenders?
3. What does the current landscape of correctional education look like in the United States,
and what are some emerging issues and trends to consider?
4. What recommendations emerge from the study for the U.S. Department of Justice and
other federal departments to further the field of correction education and where are there
gaps in our knowledge? What promising practices, if any, emerge from this review and
evaluation?

2

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

To address these questions, we used a mixed-methods approach. This included systematically
identifying, screening, and reviewing available evaluations of correctional education programs
for incarcerated adults or juveniles completed in the United States between 1980 and 2011.
Using eligible studies of correctional education programs for adults, we conducted a metaanalysis to synthesize the estimated effects of correctional education programs on three kinds of
outcomes: recidivism, employment, and academic skills. We published the results of the
literature review and meta-analysis of adult correctional education programs in an earlier report
(Davis et al., 2013), and we summarize them in this report. This report also presents three new
sections. First, for incarcerated juveniles, we systematically summarized the literature about the
estimated effects of six types of programs on academic, employment, and recidivism outcomes,
but we did not have enough studies testing common hypotheses to conduct a formal metaanalysis. Second, We also conducted a nationwide survey of state correctional education
directors to gather information on how correctional education is provided today, and about the
impact that the global financial recession has had on correctional education programs in the
United States. Finally, this report presents our recommendations.
For this study, we define correctional education to include the following:
•
•
•
•

Adult basic education: basic skills instruction in arithmetic, reading, writing, and, if
needed, English as a second language (ESL)
Adult secondary education: instruction to complete high school or prepare for a
certificate of high school equivalency, such as the General Education Development
(GED)
Vocational education or career technical education (CTE): training in general
employment skills and in skills for specific jobs or industries4
Postsecondary education: college-level instruction that enables an individual to earn
college credit that may be applied toward a two- or four-year postsecondary degree.

To meet our definition of correctional education, the program had to be administered at least
partly within a correctional facility. Programs that also included a postrelease transition
component remained eligible as long as part of the program was administered in a correctional
setting. Although some may consider life skills programs a part of correctional education, in
conjunction with BJA, we agreed to focus specifically on the four types of academic and
vocational education/CTE programs summarized above. We also restricted our scope to focus on
correctional education programs provided in the institutional setting, as opposed to postrelease or
community-based programs.
For the juvenile program systematic review, we define incarcerated youth as individuals
under age 21 who are legally assigned to correctional facilities as a result of arrest, detainment
for court proceedings, adjudication by a juvenile court, or conviction in an adult criminal court
4

Vocational education is now commonly called career technical education. “Vocational education” is the term we
used in our adult meta-analysis (Davis et al., 2013), and it used in most of the studies we reviewed. We use the term
“vocational education/CTE” throughout this report.

3

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

(Office of Juvenile Justice and Delinquency Prevention, 2013). We define correctional
education to include any academic or vocational education/CTE program provided within the
correctional facility setting, regardless of jurisdiction. As with our adult review, we permitted
eligible interventions for juveniles to include an aftercare (postrelease) component, but the
interventions had to be delivered primarily within the correctional facility.
Finally, our focus is on correctional education programs provided to incarcerated adults in
state prison and to juveniles in the institutional setting at the state and local levels. These foci
enable us to address the question of what is known about the effectiveness of correctional
education—specifically, academic programs and vocational education/CTE programs—for
incarcerated adults and juveniles in the United States.
In the remainder of this chapter, we provide an overview of the evolution of the field of
correctional education in the United States and describe a roadmap for the remaining chapters.

The Evolution of Correctional Education in the United States
To understand where correctional education stands today, it is useful to briefly consider the
history of correctional education in the United States. Educational programs for incarcerated
individuals were first introduced in in the United States when clergyman William Rogers began
offering instruction at Philadelphia’s Walnut Street Jail in 1789 (Chlup, 2005). The first schools
in prisons were known as “Sabbath schools,” created by the Boston Prison Discipline Society in
1833 and focused on moral and religious instruction (Gehring, 1997).
Support for educational programs within correctional settings has waxed and waned over
time as the nation’s philosophy of punishment has shifted from rehabilitation to crime control
and then back again. For example, during Andrew Jackson’s presidency (1824–1837),
Americans believed that crime was posing a fundamental threat to the stability and order of
society (Chlup, 2005), and the general belief was that a primary purpose of punishment was
rehabilitation to change an individual’s behaviors (F. Allen, 1981; Chlup, 2005). The period
between 1901 and 1920 (known as the Progressive Era) was a period of social activism and
political reform in the United States, including a focus on prison reform and an emphasis on
educating prisoners (Chlup, 2005).
In 1965, the Survey for the President’s Commission on Law Enforcement and Administration
of Justice, carried out by the National Council on Crime and Delinquency, found that many
institutions did little to prepare prisoners for reentry (Chlup, 2005), that a high number of
offenders were “severely handicapped educationally,” and that many had dropped out of school
(President’s Commission on Law Enforcement and Administration of Justice, 1967). In addition,
the survey found that offenders tended to have unstable work records and lacked a vocational
skill. The commission recommended that correctional institutions upgrade educational and
vocational training programs, extending them to all inmates who could profit from them, and that

4

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

states should, with federal support, establish programs to recruit and train academic and
vocational instructors to work in correctional institutions.
One large-scale program, started in 1965 and continuing through the 1970s with funding
support from the U.S. Department of Education, facilitated the development of plans and
curricula for adult basic education in prisons in 45 states as well as the training of prison
personnel on in the implementation and evaluation of correctional education (Ryan and McCabe,
1994).
In 1971, the inmate uprising at the Attica Correctional Facility in Attica, New York, resulted
in the deaths of 11 prison employees and 32 unarmed prisoners (Chlup, 2005). The prisoners’
demands for political rights and better living conditions included the provision of rehabilitative
programming and access to educational programs.
The 1970s are often considered the “Golden Age” of correctional education (Ryan and
McCabe 1994, p. 451). During this period, education was regarded as the most important tool for
successful rehabilitation. Adult basic education and GED programs were being provided,
vocational training programs were being given a high priority, and postsecondary education
programs were being offered through prison release and correspondence courses (Ryan and
McCabe, 1994). However, by the 1980s support among the public and policymakers for
correctional education once again waned, and funding for education in prison suffered dramatic
cuts (Lillis, 1994). Approximately half of correctional systems made cuts in inmate education
programs, especially in vocational and technical training, with corrections officials citing state
budget cuts as the main reason for most program reductions (Lillis, 1994).
In the 1980s, the Federal Bureau of Prisons (BOP) initiated mandatory education starting
with the establishment of its first mandatory adult basic education program in 1982 and followed
by a requirement in 1983 that each institution have its own qualified reading specialist or special
educator instructor (Ryan and McCabe, 1994). By 1986, the standard for mandatory education in
the BOP system was an eighth-grade achievement level. Subsequently, a number of states
followed the BOP’s example in requiring that inmates attend school for a minimum number of
months if they did not have a specified reading level or had not received a high school diploma
or GED (Steurer et al., 2010).
The Higher Education Act, passed in 1965, provided student loans via Pell college tuition
grants to any qualified students (including incarcerated individuals) to help pay for their
postsecondary education (Crayton and Neusteter, 2008). However, in the mid-1990s the Violent
Crime Control and Law Enforcement Act of 1994 (Pub. L. 103-332) eliminated federal and state
inmate eligibility for Pell college tuition grants, thus, affecting many college programs for
inmates (Taylor, 2005). Additionally, limitations were placed on the amount of federal adult
education and vocational education/CTE funds available for correctional education programs
(Coley and Barton, 2006). Within one year of eliminating Pell Grant access to prisoners,
participation in postsecondary correctional education programs dropped 44 percent (Marks,
1997).
5

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

In 2007, the Second Chance Act (SCA) (Pub. L. 110-199) was enacted to improve outcomes
for and provide a comprehensive response to the increasing number of individuals who were
released from prisons, jails, and juvenile residential facilities and returning to communities upon
release. The SCA is a historic piece of legislation focusing on reentry programs as an interagency
priority at the federal level. In 2010, the SCA for the first time set aside dedicated funding for
correctional education, which led to this RAND study.
The 2008 recession deeply affected correctional systems. States were forced to curtail
spending of all kinds, with correctional expenditures especially attractive targets for state belttightening. During FY2010, 31 of the 50 state departments of corrections had mid-year cuts,
totaling $806 million (NGA/NASBO, 2010). Strategies to reduce correctional expenditures and
achieve operational efficiencies included closing prisons, reducing staff, and curtailing services
and programming. Correctional education (and other rehabilitative) programs experienced deep
budget cuts in a number of states, resulting in some dramatic reductions in the number of
programs offered, the size of classes, and in the number of inmates who participate in these
programs, as well as changes to the models of delivery.
In 2012, anecdotal evidence suggests an uptick in funding for correctional education in some
states based on informal reports from state correctional education directors citing either no
further funding cuts or even some minor increases in funding—a situation that has enabled them
in some cases to begin modestly rebuilding programs (authors’ personal communications with
the Correctional Education Association Leadership Forum). However, most state correctional
education directors expect to face a reduced funding environment for correctional education
programs and the continuing need to demonstrate a return on investment. In the face of
budgetary pressures, and to inform future budget decisions, state legislatures and departments of
corrections are asking fundamental questions: How effective are these programs? What return on
investment do they provide?
Researchers have attempted to answer these questions before. In 1975, Lipton, Martinson,
and Wilks published a systematic review of 231 studies of prisoner rehabilitation programs and
concluded that there was no conclusive evidence that correctional education was beneficial. This
assertion was later challenged by Wilson, Gallagher, and MacKenzie in 2000. In the next
chapter, we provide a more up-to-date and comprehensive assessment.

Organization of This Report
The remainder of this report is organized to address the major research questions listed above. In
Chapter Two, we summarize the results of our previously published meta-analysis (Davis et al.,
2013) conducted as part of our BJA cooperative agreement examining the effectiveness of
correctional education for incarcerated adults and present the results of a cost analysis. In
Chapter Three, we present the results of a systematic review of the evidence about the
effectiveness of educational interventions implemented within juvenile correctional facilities. In
6

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Chapter Four, we present the results of a national survey of state correctional education directors
that describes the current landscape of correctional education and explores the impact of the
2008 recession. A thorough explanation of the approach for the meta-analysis, systematic
review, and survey are included in the relevant results chapters. In Chapter Five, we summarize
our study’s key findings and discuss their policy implications and directions for future research.

7

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

2. How Effective Is Correctional Education for Incarcerated
Adults?

Introduction
As mentioned in Chapter One, the centerpiece of RAND’s evaluation of correctional education
for BJA entailed determining how effective correctional education is in improving outcomes for
incarcerated adults. We focused on three outcomes of interest: reducing recidivism, promoting
postrelease employment, and improving learning in reading and in math among adults. To
measure effectiveness, we carried out a comprehensive systematic review of existing literature to
identify relevant studies of correctional education effectiveness, followed by a meta-analysis of
the relevant studies identified—a statistical method that synthesizes findings across multiple
studies. Also, to put the effectiveness results in some context, we performed a basic costeffectiveness analysis based on the assessed effectiveness of correctional education for reducing
recidivism.
This chapter provides an overview of our review and synthesis of 58 identified and relevant
studies published during the past three decades to assess what the existing research base has to
say about the effectiveness of correctional education relative to the three outcomes. The full set
of findings appears in our earlier report, Evaluating the Effectiveness of Correctional Education:
A Meta-Analysis of Programs that Provide Education to Incarcerated Adults (Davis et al., 2013),
which is available on the RAND website at www.rand.org/t/RR266.
In this chapter, we briefly describe the history of meta-analyses in correctional education to
help readers understand how our work builds on previous efforts. Then, we discuss our approach
to identifying and evaluating individual studies, followed by a summary of the main findings
from the meta-analysis. We conclude with the results of our basic cost-effectiveness analysis.
The earlier meta-analytic report (Davis et al., 2013) contains extensive technical information on
the search process and the statistical underpinnings of the meta-analysis. Readers interested in
those details are encouraged to read the earlier report.

Previous Meta-Analyses on Correctional Education
Understanding the role that correctional education plays in rehabilitating incarcerated adults and
improving their reentry into society is a key goal of our study and of the meta-analysis we
conducted. Given the long history of correctional education in the United States, previous studies
have examined its effectiveness through meta-analyses of available evidence. As a backdrop to
our study, we first synthesized findings from previous meta-analyses of correctional education
programs in the United States. In keeping with our study goals, we synthesized only meta8

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

analyses that explicitly focused on education programs administered primarily to adult offenders
in correctional facilities. According to our review, three major published meta-analyses meet
these criteria: Wilson, Gallagher, and MacKenzie (2000); MacKenzie (2006); and Aos, Miller,
and Drake (2006).5 These studies differ in their parameters, methods, and conclusions. We
review the findings from each in turn, focusing first on the systematic review of correctional
education programs conducted by Lipton, Martinson, and Wilks (1975) that set the stage for the
current policy discourse and research direction in the field.6
Lipton, Martinson, and Wilks (1975)
In 1975, Douglas Lipton, Robert Martinson, and Judith Wilks published a systematic review of
231 studies of prisoner rehabilitation programs spanning the years 1945 to 1967—a review that
provided the first major effort to take stock of the potential efficacy of correctional education.
Within their sample of 231 programs, Lipton and his team identified a subset of “skilldevelopment programs,” which consisted of academic and/or vocational training. They
summarized comparisons of program participants and nonparticipants in studies that used
recidivism and employment as outcomes. In their review, they discussed differences in
methodological quality, highlighting (where appropriate) studies with carefully or poorly
selected comparison groups. However, this variation in research design did not factor into how
they tallied statistically significant program effects. As long as the study had a group exposed to
correctional education (a treatment group) and a group that was not exposed to correctional
education (a comparison group), it was included in their review. None of the studies reviewed
employed random assignment.
Across eight studies that assessed recidivism, three showed significantly lower rates of
recidivism among program participants, and one showed significantly higher rates of recidivism
among program participants. The other four studies showed no differences between the treatment
and comparison groups. In two studies that examined employment as an outcome, offenders who
participated in vocational training programs fared worse than nonparticipants after being
released. Overall, their review found no conclusive evidence that correctional education was
beneficial and that, in some cases, it might even be harmful. Lipton et al.’s systematic review is
5

The studies included in these meta-analyses are largely based on studies of correctional education programs in the
United States. However, a handful of international studies are also included.
6

Since the publication of the landmark Lipton, Martinson, and Wilks study in 1975, there have been other
systematic reviews of adult correctional education that do not apply meta-analytic methods (e.g., Gaes, 2008), and
there have been meta-analyses of correctional education programs administered to juvenile offender populations
(e.g., Lipsey, 2009). With the exception of the Lipton, Martinson, and Wilks study, which is important to
acknowledge because of its seminal role in the field, we discuss only meta-analyses of adult correctional education
programs, because their methods, findings, and conclusions are most relevant for providing context to our study.
Additionally, readers should note that we are aware of two dissertations (Chappell, 2003; Wells, 2000) that have
used meta-analytic techniques to assess the relationship between correctional education and recidivism. We do not
review their analyses in depth here, but their findings, by and large, accord with those of Wilson, Gallagher, and
MacKenzie (2000); MacKenzie (2006); and Aos, Miller, and Drake (2006).

9

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

notable, in part, because it set the tone for future research and policy discourse in the field—
establishing the belief that “nothing works” in prisoner rehabilitation. However, the three more
recent meta-analyses have turned that belief around.
Wilson, Gallagher, and MacKenzie (2000)
Twenty-five years after the Lipton et al.’s work, in 2000, David Wilson, Catherine Gallagher,
and Doris MacKenzie (2000) at the University of Maryland revisited Lipton et al.’s work,
conducting a meta-analysis that included 33 studies of correctional education programs
administered to adults published after 1975—a time period that broadly covered the time since
the Lipton et al. study was released. Wilson and his team sought to address some limitations in
Lipton et al.’s work, in particular by using formal meta-analytic techniques (techniques that were
not yet developed when the Lipton et al. study was conducted), which average findings of
multiple studies into a single parameter of program or “treatment group” efficacy. Additionally,
they rated each study using a scale that they and their colleagues at the University of Maryland
developed specifically for systematic reviews of correctional programs (Sherman et al., 1997).
This scale, referred to as the Maryland Scientific Methods Scale ranges from 1 to 5, and accords
the highest rating (5) to well-executed randomized control trials and the lowest ranking (1) to
studies that lack a comparison group. Wilson and his colleagues dropped all studies that lacked a
comparison group and used the Maryland Scale rating as a statistical control in their analysis.
Only three of the 33 studies included employed random assignment. Whereas the Lipton et al.
study documented mostly mixed results, the Wilson et al. study found that correctional programs
were beneficial, by and large. In their meta-analysis, they showed that participation in academic
programs—including adult basic education, GED, and postsecondary education programs—was
associated with an average reduction in recidivism of about 11 percentage points. Thus, Wilson
and his team’s findings, based on more recent programs and more rigorous methods of analysis,
questioned the claim that “nothing works.”7
MacKenzie (2006)
A few years later in 2006, Doris MacKenzie, a co-author of the Wilson study, updated the
Wilson et al. study meta-analysis to include a handful of newer studies, to limit the sample to
only those studies published after 1980, and to those studies that had stronger study designs.
Specifically, to be included in MacKenzie’s re-analysis, the study needed to employ one of the
following three research designs: a well-executed randomized, controlled trial; a quasiexperimental design with very similar treatment and comparison groups; or a quasi-experimental
design with somewhat dissimilar treatment and comparison groups, but reasonable controls for
7

Since the publication of the Lipton, Martinson, and Wilks (1975) study, a number of criminologists and
policymakers questioned the claim that “nothing works.” However, it was not until the Wilson, Gallagher, and
MacKenzie (2000) study’s meta-analysis that a comprehensive evaluation of the literature was synthesized in a
systematic way to directly challenge the conclusion of the Lipton, Martinson, and Wilks (1975) study.

10

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

differences. Only one of the 13 studies included employed random assignment. In her reanalysis, she again found that academic program participation appeared beneficial: The odds of
not recidivating were 16 percent higher among academic program participants than
nonparticipants. However, with the new sample parameters in place, she now found that
vocational program participation was also associated with a reduction in recidivism: The odds of
recidivating were 24 percent lower among vocational program participants than nonparticipants.
She did not update the analysis of employment.
Aos, Miller, and Drake (2006)
Finally, also in 2006, Steve Aos, Marna Miller, and Elizabeth Drake of the Washington State
Institute for Public Policy conducted a meta-analysis of 571 offender rehabilitation programs for
adults and for juveniles, ranging from counseling to boot camps to education. They limited their
sample to studies conducted from 1970 onward and, like MacKenzie’s meta-analysis published
the same year, included only studies that met the same design criteria as MacKenzie (listed
above). The number of studies reviewed that employed random assignment is not clearly stated
in Aos et al.’s documentation. In analyzing 17 studies of academic education programs and four
studies of vocational education/CTE programs administered to adults, they found results that
largely agreed with MacKenzie’s: On average, participants had lower rates of recidivism than
their nonparticipant peers. Specifically, they found that academic program participation was
associated with a 7 percent reduction in recidivism, and vocational program participation was
associated with a 9 percent reduction in recidivism.
In sum, then, early reviews of correctional education programs administered to adults found
inconclusive evidence to support their efficacy—a finding that contributed to the popular belief
that “nothing works” in prisoner rehabilitation; however, this conclusion may have been
premature, given that appropriate analysis techniques had not been developed (Slavin, 1984).
More recent reviews using meta-analysis techniques question the conclusions of the earlier work,
finding evidence of a relationship between correctional education program participation before
release and lower odds of recidivating after release. However, the most recent meta-analyses
(Aos, Miller, and Drake, 2006; MacKenzie, 2006) did not consider employment outcomes; thus,
whether program participation is associated with postrelease success in the labor market
remained unclear.

Approach
In this section, we discuss the two-part approach we took to conducting our evaluation, starting
the systematic review of the literature to identify and narrow down the relevant research from the
published and grey literature and then turning to the meta-analysis of the relevant research
results.

11

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Approach to Conducting the Systematic Review of the Literature
As discussed in Chapter One, correctional education is a highly varied approach to rehabilitating
inmates, the administration of which depends on state and federal resources, the infrastructure
and staffing in place at the facility, and the skills and abilities of the inmate population (e.g.,
some inmates require basic literacy and quantitative skills, while some are able to take collegelevel courses). Given all these variables, practitioners take a variety of programmatic approaches
to rehabilitation, which, in turn, means that researchers who seek to empirically study how
effective correctional education is vary in their research designs and in their study foci.
To accommodate this variation in both programming and in research methods, we took as
inclusive an approach as possible to evaluating the existing literature base so that our evaluation
findings would generalize to the broadest set of programs. Thus, our analysis was intended to
provide an overall empirical examination of correctional education’s effectiveness based on the
array of programs and systems in place during the past three decades, and was not a program-byprogram evaluation. To be as informative as possible to policymakers and practitioners, we
restricted our focus to correctional education programs administered to adults in the United
States. There were not enough studies of correctional education administered to juveniles that
met our methodological standards to produce a sound meta-analysis. Hence, we instead provide a
systematic review of the juvenile correctional education literature, presented in Chapter Three of
this report.
We began our search by scanning the universe of potential documents to compile all
available empirical research studies that examine the effect of correctional education programs
on the three outcomes of interest—recidivism, postrelease employment, and reading and math
scores. This included a search of relevant research databases; a “grey literature” search of online
repositories maintained by research organizations, think tanks, and universities; and a
bibliographic scan of all major literature reviews, systematic reviews, thematic policy overviews,
and existing meta-analyses of inmate rehabilitation programs. This search yielded 1,112
documents, of which 267 were identified as primary empirical studies. We define a primary
empirical study as one in which the authors were directly responsible for the research design,
data analysis, and the reporting of the findings.
To be included in our meta-analysis, the study needed to meet three eligibility criteria. First,
it needed to evaluate an eligible intervention, defined here as an educational program
administered in a jail or prison in the United States published (or released) between January 1,
1980, and December 31, 2011. We define an educational program as one that includes an
academic or vocational curriculum taught by an instructor, designed to lead to the attainment of a
degree, license, or certification. The instruction needed to occur while the participant was
incarcerated; thus, postrelease/parolee-focused programs were not eligible. Also, instructional
programs that did not explicitly address academic or vocational skills—for instance, life skill
programs or cognitive-behavioral programs—were not eligible.

12

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Second, the study needed to measure the effectiveness of the program using an eligible
outcome measure, which for our meta-analysis included recidivism, postrelease employment,
and achievement test scores.
Finally, the study needed to have an eligible research design, which, for our purposes, is one
where there is a treatment group comprising inmates who participated in or completed the
correctional education program and a comparison group of inmates who did not. Two doctoral
students from the Pardee RAND Graduate School of Public Policy reviewed each study
independently to determine whether studies met each of the three criteria. A project team
member (Ph.D.-level) reviewed and reconciled cases where the students did not agree on their
assessments.
Of the 267 primary empirical studies, 58 met all three eligibility criteria and were then
subjected to a full scientific review conducted by a team of ten faculty members from various
academic departments across the country who had substantive expertise in correctional
education, criminal justice, and/or social services for at-risk populations. Each team member
independently reviewed each study and extracted key pieces of information about the program
being evaluated and about the study’s setting, participants, and design. Two project team
members reviewed and reconciled cases where the review team did not agree on their
assessments. For additional quality control, in addition to the aforementioned external review, all
outcome metrics were independently assessed and verified by two Ph.D. researchers who were
members of the project team.
Once all data were abstracted and verified, each study was rated in terms of its scientific
rigor. A key metric of scientific rigor is the degree to which the researchers who conducted the
studies effectively mitigated the threat of selection bias. This form of bias occurs either when
program participation is voluntary or when inmates are selected to participate by program
officials based on objective measures of academic readiness (e.g., literacy level, Test for Adult
Basic Education (TABE) scores, pre-incarceration grade level proficiency) and/or subjective
perceptions of the inmate’s competencies. When such nonrandom selection occurs, IF inmates
receiving correctional education are more motivated, more academically prepared, and in better
health than their peers who do not participate in the program; this means that any difference
observed on outcomes between treatment and comparison groups may reflect the types of
inmates who participate in the program and not necessarily the effect of the program on the
inmate.
To mitigate this form of bias as much as possible, we partitioned our findings two ways: (1)
findings aggregated across all eligible studies, and (2) findings aggregated across only those
studies with the most rigorous research designs. The former includes all studies that met our
eligibility criteria, while the latter includes only those studies that used either a well-executed
randomized controlled trial or a quasi-experimental design with very similar treatment and

13

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

comparison groups.8 When possible, we focused our analytic attention on the latter set of studies,
because they are least likely to be affected by selection bias, and thus, best positioned to estimate
the true effect of participating in a correctional education program.
Approach to Conducting the Meta-Analysis
Meta-analysis is a statistical technique applied to data obtained from a systematic review in
which findings of multiple studies are averaged into a single parameter that measures how much
program participants (the treatment group) differ from nonparticipants (the comparison group)
on pre-selected outcomes. We constructed our treatment and comparison groups as
conservatively as possible, following an intent-to-treat approach. In an intent-to-treat approach,
every subject who was assigned to the treatment group is analyzed on the outcome of interest as
a member of the treatment group, regardless of whether or not they received the full dosage of
the treatment through completion. We conducted three separate meta-analyses corresponding to
our three outcomes of interest: recidivism, postrelease employment, and achievement test scores.
Our recidivism analysis is based on 50 studies, our employment analysis is based on 18 studies,
and our achievement test score analysis—which looks at reading and math score tests—is based
on four studies.9 These add up to more than the 58 studies that emerged from the systematic
analysis, because some studies used both recidivism and postrelease employment as outcomes
and, thus, contribute to both the recidivism and the employment meta-analysis. All four of the
studies that used achievement test scores as the outcome variable evaluated the effects of
computer-assisted instruction. Therefore, although our analyses of recidivism and postrelease
employment outcomes look at a broad range of correctional education programs, our analysis of
achievement test scores is narrowly focused on programs with computer-assisted instruction. For
more information on the details of the meta-analysis and how it was conducted, see the metaanalysis report (Davis et al., 2013).

8

To classify studies according to how much they mitigated selection bias, we used the Maryland Scientific Methods
Scale, a well-regarded rating scheme of research designs in the social sciences developed by criminologists at the
University of Maryland (Farrington et al., 2006; Sherman et al., 1997). The scale has five levels: Level 5 indicates a
well-executed randomized controlled trial with low attrition; Level 4 is a quasi-experimental design with very
similar treatment and comparison groups; Level 3 is a quasi-experimental design with somewhat dissimilar
treatment and comparison groups but reasonable controls for differences; Level 2 is a quasi-experimental design
with substantial baseline differences between the treatment and comparison groups that may not be well-controlled
for; and Level 1 is a study that does not have a separate comparison group receiving the treatment. In our
partitioning of the findings, we consider those studies as having the highest degree of rigor if they received either a
Level 4 rating or a Level 5 rating. More specific information on the properties of the Maryland Scientific Methods
Scale and how we applied it to our 58 studies is included in the full companion report (Davis et al., 2013).
9

Across the 58 studies, we were able to extract a total of 102 effect sizes. The number of effect sizes exceeds the
number of studies, because a study could contain multiple treatment and comparison groups, and thus multiple
comparisons. Our recidivism analysis is based on 71 effect sizes from 50 studies, our employment analysis is based
on 22 effect sizes from 18 studies, and our test score analysis is based on nine effect sizes from four studies.

14

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Meta-Analysis Findings
In this section, we present the results of our meta-analysis. Table 2.1 provides a summary of the
findings across the outcome domains. In the table, we interpret the difference in each outcome
between the treatment and comparison group for all studies regardless of the quality of the study
design and then only for the most rigorous studies. We note the total number of studies that
contributed to the finding (n) and whether the finding was statistically significant at conventional
levels (p < .05) or not statistically significant (n.s.). We then summarize the overall effectiveness
of correctional education for each outcome. In the remainder of this section, we discuss the
recidivism, postrelease employment, and achievement score findings in the table in more detail.
Table 2.1. Summary of Meta-Analysis Findings by Outcome
Summary of Meta-Analysis Results
Outcome

Recidivism

Post-Release
Employment

All Eligible Studies

Most Rigorous Studies

Inmates who participated in
correctional education programs
had a 36% lower odds of
recidivating than inmates who did
not

Inmates who participated in
correctional education programs
had a 43% lower odds of
recidivating than inmates who did
not.

n = 50 studies, p < .05

n = 7 studies, p < .05

The odds of obtaining employment
among inmates who participated in
correctional education programs
are 13% higher than the odds of
obtaining employment among
inmates who did not.

The odds of obtaining employment
among inmates who participated in
correctional education programs are
48% higher than the odds of
obtaining employment among
inmates who did not.

n = 18 studies, p < .05

n = 1 study, p < .05

Overall Effectiveness
Correctional education is
an effective strategy for
reducing recidivism

Correctional education is
potentially an effective
strategy for improving the
postrelease employment
prospects of inmates.

Achievement Test Scores

Reading
Achievement
Test Scores

Math
Achievement
Test Scores

The overall effect of computerNot applicable
assisted instruction relative to
traditional instruction is 0.04 grade
levels, or about 0.36 months of
learning in reading
n = 4 studies, not significant

The overall effect of computerNot applicable
assisted instruction relative to
traditional instruction is 0.33 grade
levels, or about 3 months of
learning in math.
n = 3 studies, not significant

15

Learning gains in reading
among inmates exposed
to computer-assisted
instruction are similar to
learning gains in reading
among inmates taught via
traditional instruction
methods
Learning gains in math
among inmates exposed
to computer-assisted
instruction are similar to
learning gains in math
among inmates taught via
traditional instruction
methods

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

The Relationship Between Correctional Education and Recidivism
Recidivism is one of the most commonly used measures of effective prisoner rehabilitation in
criminal justice research, because it conveniently indicates how much individuals are able to
successfully re-integrate back into their communities and desist from further criminal activity.
Thus, it is not surprising that the majority (n = 50) of our eligible studies used recidivism as the
primary outcome measure.
However, when we looked across studies, we found that recidivism was defined a number of
ways; these definitions included reoffending, rearrest, reconviction, reincarceration, technical
parole violation, and successful completion of parole. We used whatever form of recidivism the
study authors reported so that we could be as inclusive as possible. In addition to how they
defined recidivism, studies varied in the time period through which they followed the study
participants after release from prison, ranging from six months to over ten years. When there
were multiple outcomes and/or time periods reported, we gave preference to reincarceration
(because this represents the modal definition of recidivism used by the authors of the studies; n =
34) and recidivism measured within one year of release or as close as possible to one year of
release (because this represents the modal time period used by the authors of the studies; n = 13).
For our analysis of recidivism, the treatment group consisted of inmates who participated in
or completed a correctional education program and the comparison group consisted of inmates
who did not participate in or complete the correctional education program.
The top row of Table 2.1 summarizes the results of our recidivism meta-analysis. When we
aggregated across all 50 studies—which represent 32 years of empirical assessments of the
effects of correctional education and which have analyses ranging in methodological quality and
rigor—we find that inmates who participated in correctional education programs had a 36
percent lower odds of recidivating than inmates who did not. This relationship is statistically
significant at p < .05.
Because many correctional education programs select inmates to participate based on their
level of academic preparedness or motivation, it could be the case that lower recidivism rates
among treatment group members reflect preexisting differences between the treatment and
comparison groups and not the effect of program participation—a selection bias that could be
reflected in the studies with the least rigorous research designs. However, this relationship
remained robust when we restricted our analysis to the seven studies with the most rigorous
research designs: Inmates who participated in correctional education programs had a 43 percent
lower odds of recidivating than inmates who did not. This relationship is statistically significant
at p < .05. This suggests that selection bias is not driving our findings. Given that we find
significantly lower rates of recidivism among treatment group members when looking at all

16

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

eligible studies as well as when looking only at those studies with the strongest research designs,
we conclude that correctional education is an effective strategy for reducing recidivism.10
The Relationship Between Correctional Education and Postrelease Employment
Postrelease employment is an important measure of correctional education’s effectiveness
because it indicates how much the training and skills received while incarcerated prepare
individuals to enter and succeed in the formal labor market. However, unlike the case with
recidivism, fewer eligible studies used employment as the outcome (n = 18).
Like recidivism, postrelease employment was defined a number of ways, including having
ever worked part-time since release, having ever worked full-time since release, having been
employed for a specified number of weeks since release, and employment status at the time of
data collection. Additionally, studies varied in the time period through which they followed the
study participants after release from prison, ranging from three months to 20 years. When there
were multiple outcomes and/or time periods reported, we gave preference to having ever worked
full- or part-time since release (because this represents the modal definition of postrelease
employment used by the study authors; n = 9) and employment measured within one year of
release (because this represents the modal time period used by the study authors; n = 7).
For our analysis of postrelease employment, the treatment group consisted of inmates who
participated in or completed a correctional education program, and the comparison group
consisted of inmates who did not participate in or complete the correctional education program.
The second row of Table 2.1 summarizes the results of our postrelease employment metaanalysis. When aggregating across all 18 eligible studies, we find that the odds of obtaining
postrelease employment among inmates who participated in correctional education programs are
13 percent higher than the odds of obtaining postrelease employment among inmates who
did not. This relationship is statistically significant at p < .05. Unlike with our analysis of
recidivism, we cannot assess how much this finding is robust to the threat of selection bias,
because there is only one study that can be classified as having a rigorous research design.
Although this one study (which used matching to create a comparison group) did find
statistically significant higher rates of postrelease employment among treatment group members,
it assessed only one program, and, thus, it is not possible to generalize this finding to the array of
programs in operation. Given that we find significantly higher rates of postrelease employment
among treatment group members when looking at all eligible studies but cannot rule out the
possibility of selection bias, the evidence is only suggestive that correctional education is

10

In additional analyses not shown, we find that this relationship holds across academic (adult basic education, high
school/GED programs, postsecondary) programs, and vocational certification programs.

17

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

potentially an effective strategy for improving the postrelease employment prospects
of inmates.11
The Relationship Between Correctional Education and Test Scores
Test scores—reading and math—measure the immediate cognitive gains that result from
exposure to educational programming and are common metrics used in studies that seek to
measure educational effectiveness. Unfortunately, only four studies in our systematic review
used test scores as an outcome. As mentioned earlier, these four studies specifically evaluated the
effects of computer-assisted instruction; as such, they provide evidence about the effectiveness
of this particular instructional approach, one that is becoming increasingly common in
correctional facilities. Three of the four studies used rigorous research designs (specifically,
randomization), but because there are so few studies, we pool them all together and do not
disaggregate by the quality of the research design. All four studies used reading test scores as an
outcome, and three of the four studies used math test scores as an outcome, as shown in Table
2.1.
We converted the outcomes into grade equivalents, where one unit is equal to a single ninemonth academic year of learning in a particular content area. This metric typically refers to a
standard scholastic setting rather than a correctional education setting, in which students receive
approximately one hour of instruction in each of six to seven content areas for five days per
week. Unlike the studies of recidivism and employment, the treatment groups for these four
studies include those who were taught through computer-assisted instruction and the comparison
groups included those who were taught through traditional face-to-face methods.
The third row of Table 2.1 summarizes the results of our reading test score meta-analysis,
and the fourth row of Table 2.1 summarizes the results of our math test score meta-analysis. In
both sets of analyses, we find that inmates receiving computer-assisted instruction have greater
gains than inmates receiving traditional face-to-face instruction. Specifically, we find that the
overall effect of computer-assisted instruction relative to traditional instruction is equivalent to
about 0.36 months of learning in reading and three months of learning in math. However,
because neither of these gains is statistically significant, we conclude that learning gains in both
reading and in math among inmates exposed to computer-assisted instruction are similar to
learning gains made by inmates taught through traditional instruction methods.
Because computer-assisted instruction can be self-paced and can be supervised by a person
other than a licensed classroom teacher, it may be less costly to administer and could even allow
correctional facilities to expand their instructional course offerings. For these reasons, the finding
of no statistically significant difference between computer-assisted and traditional (face-to-face)
instruction suggests that, based on current evidence, computer-assisted instruction may be a
11

In additional analyses not shown, we find that this relationship holds across academic (adult basic education, high
school/GED programs, postsecondary) programs, and vocational certification programs.

18

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

reasonable alternative to traditional, face-to-face classroom instruction in correctional facilities.
It is also true that the technology underlying computer-assisted instruction has improved (and
will continue to improve) over time relative to the four studies assessed; this suggests that the
effect of computer-assisted technology relative to traditional instruction methods may be larger
than it appears in our assessment. Given that we only had four studies, the most recent of which
was published in 2000, more research is needed to understand the efficacy of this instructional
approach as it is currently being implemented.

Results of Comparing Correctional Education and Reincarceration Costs
Although our meta-analysis shows that correctional education is effective at reducing recidivism,
is it cost-effective? For example, it could be that the gains in reduced recidivism are outweighed
by the costs of providing the correctional education programs. Although a formal costeffectiveness analysis was beyond the scope of our study, to place our meta-analytic findings
into context, we undertook a basic cost analysis using estimates of the costs of correctional
education and those of incarceration.12 The cost analysis is done for a three-year window after
release from prison. The detailed methodology of how we created the cost estimates can be
found in Chapter Three of our earlier report (Davis et al., 2013).
To estimate the direct costs of providing education to inmates, we obtained three inputs.
First, we derived an estimate of the cost per year per inmate for correctional education. We used
data from Bazos and Hausman (2004), who calculated the average cost of correctional education
programs per inmate participant using information from the Three States Study, which assessed
the relationship between correctional programs and recidivism in Maryland, Minnesota, and
Ohio for approximately 3,170 inmates (Steurer, Smith, and Tracy, 2003). We also used data from
the 2007 Corrections Compendium Survey Update on Inmate Education Programs (Hill, 2008).
These two sources estimated that the average annual cost of correctional education programs per
inmate participant was $1,400 and $1,744, respectively.
Second, the reincarceration rate affects the cost-effectiveness of the intervention: The higher
the reincarceration rate, the greater the potential cost savings. We used the three-year
reincarceration rate estimates for correctional education participants and nonparticipants.
Specifically, we used the most conservative reincarceration rate estimates based on the Pew
Charitable Trust’s most recent national estimate of reincarceration based on 41 states: 43.3
percent for individuals who did not receive correctional education, and 30.4 percent for who
those did—a risk difference of 12.9 percentage points, as estimated from our meta-analysis (Pew
Center on the States, 2011).

12

Although our meta-analysis incorporated a range of indicators to construct our measure of recidivism (e.g.,
reincarceration, rearrest, parole revocation rates), here we are able to base our cost analysis on estimates of cost for
three-year reincarceration rates.

19

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Third, we used data on the average annual cost per inmate of incarceration from the Bureau
of Justice Statistics’ (Kyckelhahn, 2012) analysis of state corrections’ expenditures13 and the
Vera Institute of Justice’s study on the price of prisons (Henrichon and Delaney, 2012), which
collected cost data from 40 states using a survey; these two studies estimated the average annual
cost per inmate to be $28,323 and $31,286, respectively.14 Assuming a mean incarceration length
of stay of 2.4 years (Pastore and Maguire, 2002), we calculated the average incarceration costs as
between $67,975 and $75,086, respectively, based on the two studies. For more information on
the details of the cost analysis and how it was conducted, see the meta-analysis report (Davis et
al., 2013).
Focusing on a hypothetical pool of 100 inmates, the direct costs of correctional education
programs and of incarceration itself, and using a three-year reincarceration rate, we estimate that
the direct costs of providing education to inmates range from $140,000 to $174,400 for the pool
of 100 inmates (or $1,400 to $1,744 per inmate) using the two sources mentioned above. The
three-year reincarceration costs for those who did not receive correctional education would be
between $2.94 million and $3.25 million, versus $2.07 million and $2.28 million for those who
did. Reincarceration costs are thus $870,000 to $970,000 less for those who receive correctional
education. Thus, the direct costs of providing correctional education appear to far outweigh the
direct costs of reincarceration.
Another way to look at the cost-effectiveness of providing correctional education is to
calculate the break-even point—defined as the risk difference in the reincarceration rate required
for the cost of correctional education to be equal to the cost of incarceration. For a correctional
education program to be cost-effective—or break even—we estimated that a program would
need to reduce the three-year reincarceration rate by between 1.9 percentage points and 2.6
percentage points. In fact, our meta-analytic findings indicate that participation in correctional
education programs is associated with a 13 percentage-point reduction in the risk of
reincarceration three years following release. Thus, correctional education programs appear to
far exceed the break-even point in reducing the risk of reincarceration. Given that some
programs appear more effective than others, the exact ratio of costs to benefits will naturally
depend on the effectiveness of a particular program. Future investments in correctional education
would ideally be designed to allow for rigorous identification of effective programs’ features.

13

Expenditure data were extracted from the U.S. Census Bureau’s Annual Survey of State Government Finances. It
is available on the U.S. Census Bureau’s website at http://www.census.gov/govs.
14

Ideally, one would use marginal costs rather than average costs in this calculation, but the extent to which
marginal cost differs from average cost is likely to depend on the scale of the intervention, and there is no natural
scale here—depending on the intervention, one could apply a correctional education program to a very small number
of inmates or throughout an entire state prison system. For simplicity in what follows, we imagine an intervention of
sufficient magnitude so as to allow for shifts in quasi-fixed factors such as the amount of housing needed for inmates
or the number of prison staff, in which case average costs might offer a good approximation for marginal costs.

20

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Because the analysis above accounts only for direct costs and not for indirect costs, such as
the financial and emotional costs to crime victims and costs to the criminal justice system as a
whole, this is a conservative estimate of the broader effect correctional education could yield.

Summary
In this chapter, we provided an overview of our systematic review and meta-analysis of studies
of correctional education programs administered to adults. As part of our review, we identified
58 studies of educational programs administered in a jail or prison in the United States published
(or released) between January 1, 1980, and December 31, 2011 that used recidivism, postrelease
employment, or achievement test scores as outcome measures and that had adequate treatment
and comparison groups.
Based on the higher-quality research studies, we found that, on average, inmates who
participated in correctional education programs had 43 percent lower odds of recidivating than
inmates who did not. This translates to a reduction in the risk of recidivating of 13 percentage
points, suggesting that correctional education is an effective strategy for reducing recidivism.
When aggregating across 18 studies that used employment as an outcome, the odds of
obtaining employment postrelease among inmates who participated in correctional education
(either academic or vocational education/CTE programs) were 13 percent higher than the odds
for those who did not. However, only one of the 18 studies had a high-quality research design,
limiting our ability to draw conclusions about the effectiveness of correctional education in this
area.
Lastly, when aggregating cross four studies that used achievement test scores as an outcome,
we found that learning gains in both reading and in math among inmates exposed to computerassisted instruction are similar to learning gains made by inmates taught through traditional
(face-to-face) instruction methods.
Our meta-analysis of correctional education builds off a series of past meta-analyses. While
the first meta-analysis in the field had mixed results and led to the belief that “nothing works” in
the field of prisoner rehabilitation, it had some limitations. Three more recent meta-analysis,
including newer studies and better meta-analytic techniques, reverse those earlier results,
showing that providing correctional education to inmates does reduce recidivism and improve
other outcomes. Our meta-analysis, which uses more recent studies and an even more rigorous
approach, confirms what the past three meta-analyses have shown about the effectiveness of
correctional education.
Finally, although doing a formal cost-effectiveness analysis was beyond the scope of this
study, a basic cost comparison of the direct costs of providing correctional education to inmates
and applying those costs to a hypothetical pool of 100 inmates and the direct costs of
incarceration. We found that providing correctional education to prisoners is cost-effective

21

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

compared with the direct costs of reincarceration. We also note that the results are likely to be
conservative, because they do not include the indirect costs of reincarceration.

22

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

3. A Systematic Review of Correctional Education Programs for
Incarcerated Juveniles

Introduction
In 2011, about 61,000 individuals below age 21 were incarcerated on any given day in the United
States. This figure represents roughly a quarter of 1 percent of the population age 15–20 in the
United States (U.S. Census Bureau, 2011; Sickmund et al., 2013). By comparison, roughly 1
percent of the U.S. adult population was being held in prisons and jails in the same year (Glaze
and Parks, 2012, U.S. Census Bureau, 2011), so the rate of juvenile incarceration is markedly
lower than that for adults. In addition, the number of incarcerated youth in the United States has
declined steadily in the past decade and a half, dropping from about 105,000 in 1997 (Sickmund
et al., 2013). Despite this promising trend, incarceration rates among juveniles are still far higher
in the United States than in other developed nations. In 2002, the proportion of incarcerated
juveniles among 12 developed nations ranged from a hundred-thousandth of a percent in Japan to
about seven-hundredths of a percent in New Zealand and South Africa (Hazel, 2008).15 The rate
of youth incarceration in the United States is therefore more than three times the highest rates in
other developed nations.
In this chapter, we define incarcerated youth as individuals under age 21 who are legally
assigned to correctional facilities as a result of arrest, detainment for court proceedings,
adjudication by a juvenile court, or conviction in an adult criminal court (Office of Juvenile
Justice and Delinquency Prevention, 2013). It is important to note, however, that correctional
facilities are themselves a heterogeneous category. As of 2011, the largest share of incarcerated
youth were housed in short-term detention centers (34 percent), followed by group homes (30
percent) and long-term secure facilities (27 percent).16 A smaller proportion were confined in
ranch or wilderness camps (4 percent), shelters (2 percent), reception centers (2 percent), and
boot camps (1 percent) (Sickmund et al., 2013). Importantly, these proportions exclude a
nontrivial number of juveniles tried in adult criminal courts or confined in adult facilities. As of
2011, an estimated 10,000 youth on any given day were being held in adult jails and prisons
(National Juvenile Justice Network, 2011).
Among individuals incarcerated in juvenile correctional facilities in the United States in
2011, about 86 percent were male, and the population included a marked overrepresentation of
15

In terms of nations similar to the United States, the rate was reported to be about five-hundredths of a percent in
England and Wales and about half that in Australia. Data for Canada were not reported.
16

Long-term secure facilities included training and reform schools as well as other juvenile correctional facilities.

23

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

youth of color (Sickmund et al., 2013). Forty percent of incarcerated youth were black and 23
percent were Hispanic, as compared with about 13 percent and 17 percent, respectively, in the
U.S. population at large. White, non-Hispanic youth made up 32 percent of the incarcerated
juvenile population, as compared with 63 percent of the U.S. population (Sickmund et al., 2013,
U.S. Census Bureau, 2013). About 30 percent of youth incarcerated in juvenile correctional
facilities in 2011 were under the age of 16. Another 55 percent were ages 16 or 17, and 14
percent were ages 18 to 20 (Sickmund et al., 2013).17
Juvenile offenders hail disproportionately from challenging circumstances. According to
2003 data from the nationally representative Survey of Youth in Residential Placement, 56
percent of incarcerated youth reported that they had been living with only one parent at the time
of commitment, and 26 percent reported that they had been living with neither parent. Nine
percent of surveyed youth reported that they, themselves, were already parents (Sedlak and
McPherson, 2010). Ninety percent of the incarcerated youth were found to have some type of
emotional problem, and 71 percent had multiple emotional problems. A striking 20 percent of
surveyed males and 40 percent of females reported that they had previously attempted suicide
(Snyder and Sickmund, 2006). In addition, substance use history is much higher among
incarcerated youth than among other youth in the population. Among incarcerated youth in 2003,
84 percent reported having used marijuana in their lifetimes, and 30 percent said they had used
cocaine or crack; the corresponding numbers among non-incarcerated youth were 30 percent and
6 percent, respectively (Sedlak and McPherson, 2010).
Beyond these environmental and psychological risk factors, the problems facing juvenile
offenders are compounded by comparatively weak academic skills. The average reading ability
of incarcerated youth has been estimated at the fourth-grade level, placing them five years
behind average grade-level targets (Project READ, 1978). Though this estimate is widely cited, it
is based on data collected between 1976 and 1978; it is not clear how or whether the average
literacy skills of juvenile offenders have changed during the past three decades. Incarcerated
youth are also more likely than their counterparts to be learning disabled. In a 2003 synthesis of
the research on youth with disabilities, Mears and Aron (2003) summarized evidence suggesting
that between 30 and 50 percent of incarcerated youth have special education disabilities, as
compared with approximately 10 percent of non-incarcerated youth. They also approximated that
the proportion with emotional disabilities exceeded 50 percent, that as many as 20 may have
severe emotional disturbances, and that as many as 12 percent may be mentally retarded.
However, they noted that reporting on these percentages is quite difficult, given that screening
procedures and diagnostic criteria are applied inconsistently across contexts and that students’
disability records are not consistently transferred from their regular schools to their correctional
education programs. This is important because under the Individuals with Disabilities Education
Act (IDEA) (Pub. L. 101-476), minors with disabilities have a legal right to a free, appropriate
17

Figures for incarcerated youth are based on 2011 data; comparison data for the U.S. population come from 2012.

24

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

public education in the least restrictive environment, including appropriate accommodations
(U.S. Department of Education, 2013a). Due to the need for accommodations and regulatory
documentation, students with disabilities who qualify for services under IDEA cost
approximately 1.9 times as much to educate as those without disabilities (Chambers et al., 2004).
Given the disproportionate representation of students with disabilities in juvenile correctional
facilities and the cost of educating these students, juvenile correctional facilities often struggle to
adequately serve the special needs of their students (Leone, 1994; Pasternak et al., 1988).
Nationally, the long-term outlook for youth who commit crimes as juveniles is somewhat
unclear because states track juvenile recidivism using different metrics and different subsets of
offenders, and some states do not make such data available at all. According to a 2006 report that
used data from Florida, New York, and Virginia, the 12-month re-arrest rate among released
juvenile offenders in either the juvenile or adult system was 55 percent. Using data from eight
states, the same report estimated that 33 percent of juvenile offenders were re-adjudicated in a
juvenile court or reconvicted in an adult criminal court within 12 months after release from a
juvenile facility (Snyder and Sickmund, 2006).
One important determinant of juvenile offenders’ prognosis appears to be incarceration itself.
A recent study examined ten years of administrative data from Chicago and capitalized on
plausibly random variation in juveniles’ appearance before judges inclined to assign
incarceration (Aizer and Doyle, 2013). Adjusting for demographic and crime severity variables,
the study found that juvenile incarceration reduced offenders’ high school completion rates by 13
percentage points and increased their adult incarceration rates by 22 percentage points. The
authors suggested that expansion of less-restrictive penalties such as electronic monitoring and
home confinement may be an especially effective—as well as cost-effective—strategy for
reducing future criminality among juvenile offenders.
Purpose and Organization of the Chapter
Given that juveniles have the right to a publicly funded education, policymakers face the
question of how best to provide educational services that will lower young offenders’ risk of
future crime and increase their chance of success in the legitimate economy. This question is
particularly acute for offenders assigned to correctional facilities. Since incarcerated youth
cannot take advantage of the public education systems available in their communities,
correctional facilities must provide an alternative educational system to serve them. In educating
juveniles, correctional facilities must serve a highly transient population of students who bring a
widely varied set of educational and emotional needs (Sedlak and McPherson, 2010).
This chapter systematically reviews evidence about the effectiveness of educational
interventions implemented within juvenile correctional facilities. Though, as noted above, some
juveniles are tried and incarcerated as adults, our discussion and analysis in this chapter is
limited to education in juvenile correctional facilities. This is because our extant meta-analysis of
correctional education programs for adults has already examined the research on educational
25

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

programs for individuals held in adult facilities (Davis et al., 2013) and because juvenile
correctional facilities constitute a distinct and separate part of the correctional system with their
own sets of policy conditions and constraints (Gagnon et al., 2009).
In this chapter, we first present our methodological approach to the systematic review. We
then summarize the results for the eligible studies identified, which we classify into six
intervention types—Corrective Reading (a commercially packaged curriculum), computerassisted instruction, personalized academic instruction, remedial academic instruction, vocational
education, and GED completion. Based on the dependent variables in the eligible studies, we
summarize available evidence for five types of outcomes: reading skills, mathematics skills,
diploma completion, postrelease employment, and postrelease recidivism. For each intervention
type reviewed, we contextualize our discussion of the eligible articles in terms of the wider body
of literature that is not eligible for the systematic review (due to population, setting, methods,
etc.) but that does address the effectiveness of the intervention under consideration. We conclude
the chapter with a broad summary of findings, a discussion of methodological limitations, and
suggestions for research and policy.

Approach
A fundamental difference between correctional education for juvenile and adult populations is
that juveniles in the United States have a right to a public education. Therefore, all programs for
incarcerated youth include a correctional education component. In other words, the question
facing policymakers is not whether to provide education services for juveniles in correctional
facilities, but which types of programs are most effective. The meta-analytic approach in our
adult analysis included many types of correctional education, each of which was compared
against a no-correctional-education scenario. A meta-analysis works best when you have a
relatively homogenous intervention (i.e., prison education) tested among many arenas and
settings. For adults, we were able to conduct a meta-analysis by comparing individuals who
received some form of correctional education to those who did not. However, that approach is
less well suited to studying the effectiveness of juvenile correctional education programs, since
programs with an absence of correctional education are typically not present in the United States
context on which our study is focused. Instead, our approach to synthesizing research on juvenile
correctional education is to undertake a systematic review, in which we screen and evaluate
articles using the same criteria as we employed in the adult meta-analysis. Yet, we now
summarize the findings of the research as a literature review rather than aggregating estimated
effect sizes across studies that are testing widely different hypotheses for the treatment versus
nontreatment groups. We focus on describing the balance of evidence favoring the types of
interventions examined in the literature we reviewed.

26

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Document Identification
Our comprehensive search for articles and reports pertaining to juvenile correctional education
was nearly identical to the search process we undertook for the adult meta-analysis. We
employed the same databases in our search, which were the Education Resources Information
Center (ERIC), Education Abstracts, Criminal Justice Abstracts, National Criminal Justice
Reference Service Abstracts, Academic Search Elite, EconLit, Sociological Abstracts, and
Google Scholar. As was true for the adult meta-analysis, our juvenile search was limited to
studies conducted in the United States and released from 1980 through 2011. However, the
search terms we employed in this case were specific to correctional education for juveniles. Our
searches required one of the following descriptors from each set of terms:
•
•
•

youth or juvenile
juvenile justice, prison, jail, incarcerat* [where the asterisk serves as a wildcard, allowing
for different word endings], detention center or corrections
some indication of program type, including education, academic, diploma, GED, literacy,
math, reading, science, job skills, job training, apprentice,* vocational education, voc
tech, occupational education, career technical education, workforce [or work force]
development, workforce training, workforce preparation, or school to work.

We then supplemented this list with manuscripts cited by other literature reviews on the topic
of juvenile education. Altogether, the document search process resulted in 1,150 citations for
title-and-abstract screening, as shown in Figure 3.1, which summarizes our search and screening
process.
Eligibility Assessment
The documents collected through the database searches were then screened for eligibility by two
graduate students at the Pardee RAND Graduate School who had been trained in screening
procedures for the project. The manuscript screening procedures were identical to those
described for the adult meta-analysis report (Davis et al., 2013), except that for the juvenile
review, the manuscripts were required to describe the effects of an academic or vocational
intervention on incarcerated juveniles, where the definition of juveniles was permitted at that
stage to be defined by the manuscripts themselves or to include participants under age 21. As in
the adult meta-analysis, the manuscripts were also required to be primary, empirical studies
rather than literature reviews or opinion pieces. Each manuscript was screened independently by
two screeners, and conflicts were resolved by a senior member of the research team. The
screening process yielded 157 manuscripts eligible for full-text screening.
The full-text screening process involved an independent review of the full manuscripts—not
just their titles and abstracts—by two graduate students. As was true for the adult meta-analysis,
disagreements between the two screeners were resolved by a senior member of the research
team. To pass full-text screening and be deemed eligible for inclusion in the systematic review,
the study was required to meet three criteria:
27

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

•
•
•

evaluate an eligible intervention within an eligible population and setting
measure success of the program using an eligible outcome measure
employ an eligible research design.

For this systematic review of the juvenile literature, an eligible intervention was defined as
any academic or vocational education/CTE intervention program. An eligible population—
namely, juveniles—was defined for our purposes as consisting primarily of individuals age 20 or
below. An eligible setting was any facility, regardless of jurisdiction (state, local, etc.), to which
juveniles were confined due to arrest, court proceedings, or adjudication/conviction. Eligible
interventions were limited to academic or vocational education/CTE programs. Eligible
interventions were permitted to include an aftercare (i.e., postrelease) component, but the
interventions had to be delivered primarily within the correctional facility setting. Interventions
that did not provide instruction in academic or vocational skills—for instance, mentoring
programs, substance abuse programs, and mental health programs—were excluded from the
definition of an eligible intervention.
We define eligible outcome measures as any measure of recidivism (e.g., re-arrest,
reconviction, or re-incarceration), postrelease employment, academic attainment (e.g., GED or
high school completion), or academic performance in reading and mathematics (e.g., test scores).
Measures of academic performance and completion could be gathered during incarceration or
postrelease.
Finally, we included two types of studies in the definition of eligible research design. The
first and most common type was a comparison-group design in which a group of incarcerated
juveniles who received an intervention were compared with a group of incarcerated juveniles
who did not, or who received a different version of the intervention. The second type was a
single-case design, which is a research approach discussed later in this section.
For comparison-group designs, as in our adult meta-analysis, we rated the rigor of the
juvenile studies using two scales that closely correspond to one another—the Maryland Scientific
Methods Scale and the U.S. Department of Education’s What Works Clearinghouse (WWC)
scale. On both, assessments of rigor reflect the extent to which the designs protect against
selection bias, or unobserved differences between the treatment and comparison groups that are
correlated with the outcome of interest. Table 3.1 summarizes the standards employed for both
scales. Randomized trials with low attrition constitute the most rigorous of these types of
designs, because randomizing the two groups renders the treatment and comparison group alike
in expectation. We assigned these designs a 5 (the highest rating) on the Maryland Scale, and a
“Meets Standards” rating on the WWC scale. Studies that demonstrate very close matches
between treatment and comparison groups on relevant observable characteristics (at minimum,
age, prior offenses, baseline education level, and time to data collection) are awarded a 4 on the
Maryland Scale and a “Meets Standards with Reservations” rating on the WWC scale. Studies
that do not demonstrate strong baseline matches (within a 20th of a standard deviation for the
aforementioned variables) but that attempt to control for observed baseline differences earn a 3
28

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

on the Maryland Scale, but “Do Not Meet Standards” on the WWC scale. Studies that do not
attempt to control for observed baseline differences between the treatment and comparison
groups earn a 2 on the Maryland Scale, and do not meet WWC standards. The Maryland Scale
assigns a rating of 1 to studies that do not include a comparison group because they include no
way to estimate even roughly what would have happened to the treatment group in the absence
of treatment, and these studies are not eligible for WWC review. Consistent with our adult metaanalysis, we formally exclude Maryland level-1 studies from those juvenile studies we deem
eligible for our systematic review of evidence in this chapter. However, unlike in our adult metaanalysis, we do include brief descriptions of level-1 studies that are relevant to the interventions
under discussion, insofar as they help to contextualize findings from the eligible studies. We
make mention of relevant level-1 studies because these studies are often cited in the broader
literature on what works in juvenile correctional education (e.g., Houchins et al., 2008; Wexler et
al., 2013). Our aim in briefly mentioning level-1 studies as part of the research context while
excluding them from our discussion of eligible studies is to acknowledge the centrality of some
of these studies to the correctional education literature, while also highlighting why their designs
do not warrant even limited inferences about program impact.

29

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Table 3.1. Operational Definitions of Evidence Rating Categories on the What Works
Clearinghouse and Maryland Scientific Methods Scales for the Juvenile Systematic Review
What  Works  Clearinghouse  
(WWC)  Scale  

Maryland  (MD)  
Scientific  Methods  Scale  

Meets  Standards  

5  

Meets  Standards  with  
Reservations  

4  

Joint  Operational  Definition  
Randomized,  controlled  trial  with  
attrition  below  the  liberal  WWC  
threshold,  or  single-­‐‑case  designs  
with  well-­‐‑established  pre-­‐‑  and  
post-­‐‑intervention  trends  
Quasi-­‐‑experimental  design  (or  
high-­‐‑attrition  RCT)  in  which  the  
treatment  and  comparison  groups  
are  matched  (within  about  1/20th  
of  an  SD)  at  baseline  on  at  least  
age,  prior  offenses,  baseline  
educational  level,  and  time  to  
data  collection.  Or  single-­‐‑case  
designs  with  moderately  establish  
trends.  
Treatment  and  comparison  
groups  are  matched  on  1-­‐‑2  
variables  other  than  gender,  
and/or  there  are  statistical  
controls  for  at  least  some  baseline  
differences  between  groups  other  
than  gender  
No  random  assignment  for  
matching,  and  no  statistical  
controls  for  baseline  differences  
between  treatment  and  
comparison  groups  

3  

Does  Not  Meet  Standards  
2  

1  

No  separate  comparison  group  

We make one notable exception to the comparison-group requirement, and that is for studies
that use a class of approaches called single-case designs. Single-case designs are commonly
employed in special education research, where large samples are often unavailable for
intervention evaluation (Kratochwill et al., 2010). They involve systematically introducing an
intervention with one or a few students in an effort to demonstrate causal effects on outcomes
such as participant behavior or learning. These studies typically include a large number of preand post-intervention outcome measurements, allowing students to function as their own
controls. In this way, they do include a comparison condition, but the comparison condition is
the sustained pre-intervention state of the group that eventually receives treatment. Focusing on
one or a handful of participants, these designs typically lack statistical power for conventional
30

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

hypothesis testing. However, insofar as it is possible to establish a clear trend for student
performance in the absence of the intervention, then clear deviations from that trend in the
presence of the intervention can be causally attributed to the intervention itself. The U.S.
Department of Education’s What Works Clearinghouse (WWC) has therefore established
specific standards for the rigor of single-case design studies (Kratochwill et al., 2010). We
follow these standards when rating the single-case designs included in our systematic review.
This means that we assign a level-5 rating on the Maryland Scale to studies that receive the
highest rating (“Meets Standards”) under WWC standards for single-case designs, because these
studies demonstrate a strong basis for inferring that observe effects are causal. This is consistent
with our overall use of the Maryland Scale/WWC ratings as measures of internal validity
(freedom from selection bias) rather than external validity (generalizability to broader
populations). Still, because single-case design studies are very small and do not permit
hypothesis testing, we caveat our level-5 ratings for single-case design studies with asterisks in
Appendix Table A. This reflects the fact that the studies conform to a parallel set of internal
validity standards established by the WWC for these designs. We acknowledge that the ability to
generalize from such studies is limited by the very small samples they include.
Among the 157 studies eligible for full-text screening, nine could not be located. Another 12
were duplicates. This resulted in 136 that received full-text screening, of which 27 were deemed
eligible for detailed review by our Scientific Review Team. Figure 3.1 illustrates the winnowing
process, from the initial search through title-and-abstract screening and full-text screening. It also
summarizes the reasons for exclusion among the articles that received a full-text screening.

31

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Not Primary Empirical Research on
Juvenile Correctional Education
(n=993)

1,150 Documents
Identif ied

iPrimary Research on
Juvenile Correctional
Education (n=157)

~

i-

-!-

~
Not Able t o
Locate (n=9)

Duplicat e
Documents (n=12)

Documents Procured for
Full-Text Review (n=136)

~

-l-

~

118 Studies Excluded from
Systematic Review
Ine ligible Research Desig n (n=32)
Ine ligible Intervention (n=8)
Ine ligible Intervention and Dependent
Variables (n=3)
Ine ligible Research Desig n and
Dependent Variables (n=27)
Ine ligible Research Desig n and
Intervention (n=13)
Ine ligible Research Desig n,
Intervention, and Dependent
Variables (n=3S)

18 Studies Included in
Systematic Review

Sclentll/c Revf8w
The 2711111dies that passed full-text IICI'eeDiDg wue 1hca reviewed iDdependen1ly by two Pb.D.level reeean:heD specifically 1rained in the data ex1mction prolxK:ol, aa dcecn"bed in Chapter' Two
of tho mllta-aoalytic report (Davia et al., 2013). The !Cientific :reviL!W team for tho juvenile
S)'!temlltie nMIWI' comp!lised a subset of review team mmnben liom the adult meta.-llllalysis who
broiJ8ht specific lll!pCIItise in wnectional edncation torjllvlllilos. M such. it included fivo

faculty memb11111 iiom ac•d-ic depallmenlll acrou1ba eollllt!y, who Jeeaived. two days af
training in the data c::maction proc e1111, with a l!lllluquent online evllluation for oo:aning, and with
periodic no:rming feedback via email The data emaction protocol for the juvenile systematic
review waa modified slightly iiom the protocol fur tbe adult meta-enalysis, in 1bat a question waa

32

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

added about the standalone or integrated nature of the correctional education program, and a
checkbox option for special education was added to the intervention content item. In addition,
the scientific review team was asked only to identify the outcomes and baseline variables
collected in the study rather than collecting the actual values of the variables, which were instead
extracted by a graduate student on the project, with detailed checking and confirmation by a
senior member of the research team. A copy of the juvenile scientific review protocol, including
the main worksheet, the outcomes worksheet, the baseline characteristics worksheet, and the
glossary, is shown in Appendix C.
Synthesis of Eligible Studies
Based on the extracted data, each study was rated for rigor on the Maryland Scale and WWC
scale. Eighteen of the 27 studies that underwent scientific review were deemed eligible for
formal inclusion in the analysis. Data from the 18 eligible studies were organized and
summarized by intervention type; summaries of each are shown in Table A.1 in Appendix A.
This table includes information about each study, including its citation, a description of the
treatment and comparison conditions it examines, a listing of the demographics of the study
population and the size of the treatment and comparison groups, a brief description of
intervention duration and frequency, where reported, and quantitative summaries of the effect
sizes reported or inferable in each of the studies. In cases where we had to calculate effect sizes
on test score outcomes, we subtracted pre-post changes for the comparison group from pre-post
changes for the treatment group, and divided by the pooled standard deviation of the pretest
score. For dichotomous outcomes such as diploma completion rates, employment rates, and
recidivism rates, we define effects as the percentage for the treatment group minus the percentage
for the comparison group. The Maryland Scale rating reflects the extent to which the research
design and analysis mitigated selection bias in the effect size estimate. In studies that employed
random assignment, we report on the intent-to-treat effects, meaning the differences between
participants assigned to the treatment and control groups, regardless of their compliance and
persistence in the intervention. For randomized trials, a Maryland Scale rating of 5 means that
the attrition rates of the treatment and control groups fell below the more liberal of the two
attrition thresholds established by the WWC for a study to “Meet Standards.” We provide
additional details about the calculation we used for this threshold in our adult meta-analytic
report (Davis et al., 2013).
Summaries of the nine studies that were deemed ineligible due to research design (i.e., level1 studies) are still shown—but are shaded in blue—in Table A.1. Because of the small number of
eligible studies within each category and because the hypotheses tested within a category of
intervention types were not always uniform, our findings for each analytic category are described
qualitatively rather than aggregated quantitatively across studies, as they were in the metaanalysis for adults. In the results section that follows, we present a synthesis of findings for each
intervention category.
33

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

We begin our discussion of each intervention category by discussing the broader literature
pertaining to that intervention, even if that literature focuses on noncorrectional settings or
populations. Our discussion of the research contexts also includes a few studies (those in shaded
rows in Table A.1) that were conducted in juvenile correctional facilities but were rated a level 1
on the Maryland scientific methods scale because they lacked an adequate comparison condition.
After briefly presenting the research context for each intervention, we discuss the findings of
studies that were eligible for the systematic review. Where sufficient data are available, we
report effect sizes in pooled pretest standard deviation units for assessment results, and in
percentage point differences for recidivism and employment rates. For single-case design
studies, we calculate and report mean differences in performance before and after intervention
administration.18 We conclude each intervention subsection with a summary of what can be said
about that intervention in juvenile correctional settings, based on the preponderance of extant
evidence.
Distinctions from Other Reviews of Interventions for Juvenile Offenders

Our systematic review is intended to complement other reviews that have examined the evidence
on educational interventions for juvenile offenders. Lipsey (2009) has conducted a
comprehensive meta-analyses on interventions designed to reduce juvenile delinquency. His
study differed from ours, however, in that it included a wide array of interventions beyond
education and was not limited to studies conducted within correctional facilities. He also limited
his analysis to studies that focused on recidivism as the dependent variable of interest. Sander
and colleagues (2012) also examined the effects of a broad array of interventions for juvenile
offenders, including educational and non-educational interventions, and including studies
conducted within and outside of correctional facilities, but they diverged from Lipsey in that they
focused on academic rather than recidivism outcomes. In contrast to both of those studies, we
limit our review to studies of academic and vocational education/CTE interventions implemented
with juveniles incarcerated in correctional facilities.
Building on a small review of reading interventions in correctional settings by Krezmien and
Mulcahy (2008), Wexler et al. (2013) undertook a review that focused only on academic
interventions undertaken in juvenile correctional facilities, and that looked exclusively at
academic outcomes. Though our review partially overlaps with that of Wexler et al. (2013), it
differs in three key ways. First, similar to our meta-analysis of adult interventions, we include
both academic and vocational education/CTE interventions rather than academic interventions
only. Second, we consider not only academic achievement outcomes but also employment and
18

Kratochwill et al. (2010) acknowledge the difficulty of presenting summary statistics for single-case design
studies, where determinations of evidence are based on visual examination of trends rather than conventional
hypothesis testing. They discuss reporting the percentage of non-overlapping data points as a way of describing
effects, but this approach still says little about the substantive magnitude of effects. We report mean gains to help
clarify the magnitudes in a way that is comparable to our reporting for comparison-group studies.

34

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

recidivism outcomes. Third, we include studies regardless of whether or not they are published in
peer-reviewed journals. We do this in order to be broadly inclusive of reports, dissertations, and
other ways in which research findings are often reported, and also to minimize publication bias,
which may result if studies with positive findings are more likely to be published in journals than
those with negative or null findings (Borenstein et al., 2009). We nevertheless take care to rate
the rigor of studies using the Maryland Scale, which we operationalize using rules from the
WWC scale, as described above. This clarifies the extent to which results are internally valid and
free of likely selection bias. To address the issue of external validity—that is, generalizability—
we report on sample sizes, sample demographics (where given) and, where possible, on the
statistical significance of the effects.

Results
The 18 studies formally included in this systematic review, which are summarized in Table A.1,
address six different categories of interventions: Corrective Reading (again, a commercially
available curriculum), computer-assisted instruction, personalized academic instruction, remedial
academic instruction, vocational training, and passing the GED test. We limit the systematic
review to studies in which these interventions are administered within correctional facilities
serving juveniles. The studies examine four types of outcomes: reading skills, diploma
completion, postrelease employment, and recidivism, as measured by re-arrest or reincarceration. It is notable that passing the GED assessment (GED Testing Service, 2013) serves
as both an independent and dependent variable of interest in our review. This is because two of
the studies examine the relationship between intensive, personalized instruction and academic
attainment (including earning a GED), while two others examine the relationship between GED
completion in a juvenile facility and postrelease employment and recidivism. Given that many
correctional educators in workshops and individual discussions have pointed to implementation
of the new online GED assessments as a current and pressing challenge (see Chapter Four), the
evidence examining the effect of receipt of the GED was of particular interest to the research
team.
Table 3.2 summarizes results for the two intervention categories—Corrective Reading and
computer-assisted instruction—that use measures of reading skill as dependent variables. Table
3.3 summarizes results for the other three categories—personalized academic instruction,
vocational training, and earning a GED—that focus on diploma completion, employment, and/or
recidivism outcomes. Table A.1 presents additional, study-level details about each of the studies
summarized in Tables 3.2 and 3.3.

35

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Table 3.2. Summary of Findings for Reading Interventions
Intervention
Type
Corrective
Reading

Studies and Maryland
Scale Ratings
n Studies=4
n Participants=49
Level-5 studies: 3
(1 randomized controlled
trial, 2 single-case)
Level-3 studies: 1

Reading Effects

Preponderance of Evidence

Average gains of 9 to 36 words
read correctly per minute in 2 level
5 single-case design studies (no
hypothesis test);

The preponderance of evidence
about Corrective Reading in
juvenile correctional settings is
positive for reading skills based on
two level 5 single-case designs
and 1 level-3 study, but samples
are small in all cases. Level-5
studies should be largely free of
selection bias, but none of the
level-5 studies provides a
hypothesis test of the overall
effects of Corrective Reading.

Non-significant gain of 0.66 SD
across reading domains in level-3
study (p=0.36);
Non-significant gain of 0.21 SD
across reading domains for group
size of 4 versus 12 in level 5
randomized trial (p=0.65)

Computerassisted
instruction

n Studies=3
n Participants=1,399
Level-5 studies: 3

Gain of 0.21 SD in reading for
Read 180 relative to default
curriculum in level-5 study
(p<.001);

The preponderance of evidence
about Read 180 in juvenile
correctional settings is positive for
reading. The preponderance of
evidence about Fast ForWord is
slightly negative but statistically
Nonsignificant decrease of 0.17
SD across reading domains for
nonsignificant for reading. The
Fast ForWord relative to default
preponderance of evidence
curriculum in level-5 study (p>.05); onTune in to Reading is slightly
positive but statistically
Nonsignificant gain of 0.21 SD in nonsignificant for reading. All
reading for Tune in to Reading
estimates should be largely free of
relative to FCAT Explorer in level- selection bias.
5 study (p>0.05)

NOTE: We do not provide MD Scale or WWC effectiveness ratings because the number of studies is small and some
have very few students. Thus, we think it is premature to declare any of these interventions effective or ineffective
based on available evidence. Instead, we provide a column reporting on the preponderance of existing evidence,
which may be positive, mixed or negative for each outcome examined.

Corrective Reading
Research Context

Corrective Reading is a commercially available, intensive reading program designed for students
whose reading skills are below grade-level. The curriculum, which emphasizes direct instruction
over inductive or student-driven approaches, includes an instructional strand focused on
decoding skills (i.e., identifying unfamiliar words) and another strand focused on
comprehension, though the two strands can also be taught together (McGraw Hill Education,
2013).
In 2007, the WWC evaluated Corrective Reading and deemed it to have potentially positive
effects on alphabetics (e.g., phonics and decoding) and fluency (e.g., rate and accuracy), but no
discernible effects on comprehension (What Works Clearinghouse, 2007). This evaluation was
based on a randomized trial undertaken in third-grade classrooms at eight elementary schools
and thus did not focus on correctional education settings per se (Torgesen et al., 2006). The other

36

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

24 Corrective Reading studies screened in the WWC review did not meet WWC evidence
standards. However, the WWC review occurred before the WWC had established separate
standards for single-case design studies. This is important because two of the studies screened
out in that review—those by Drakeford (2002) and Allen-DeBoer and colleagues (2006)—were
single-case design studies that we include in our systematic review because they were
undertaken in juvenile correctional facilities and now comply with the set of WWC standards for
single-case designs (Kratochwill et al., 2010). In addition, the WWC review screened out a study
that we evaluate to be a level 2 on the Maryland Scale and thus include in our systematic review
(Scarlato and Asahara, 2004). Most of the Corrective Reading studies reviewed by the WWC did
not focus on incarcerated youth.
In our systematic review of Corrective Reading implemented in juvenile correctional
facilities, we consider four studies that meet an evidence level of 2 or higher on the Maryland
Scale. All of the studies focus on reading skills as the dependent variables of interest. The
studies’ findings are reported in Table 3.2, and additional details about each study appears in
Table A.1. Three of the studies compare Corrective Reading to the default reading intervention
in the juvenile facility. These include two well-implemented single-case design studies (AllenDeBoer et al., 2006; Drakeford, 2002) that each warrant a WWC “Meets Standards” rating, and
thus we give them a corresponding Maryland Scale rating of 5.19 The other, by Scarlato and
Asahara (2004), is a nonrandomized comparison group study that does adjust for baseline
performance, and so rates a level 3 on the Maryland Scale. In addition, we examine a
randomized trial with low attrition by Houchins and colleagues (2008), which rates a 5 on the
Maryland Scale but differs from the others in that it compares two approaches to Corrective
Reading implementation—one implemented with a small group of four students, and another
with a larger group of twelve students.
Studies with Ineligible (Level-1) Designs in Juvenile Correctional Settings

Before we turn to the eligible studies, it is worth noting that the research context includes two
studies conducted in juvenile correctional education settings that did not meet our systematic
review standards because they used pre/post measures without a comparison group, rendering
them a level 1 on the Maryland Scale. As noted, we reference these studies (and include them in
shaded rows of Table A.1) because they are often cited as part of the evidence base on Corrective
Reading. However, because they lack comparison conditions within the study, we do not
consider them to be part of our systematic review of the evidence base for the intervention.
A study by Coulter (2004) focused on 12 students, age 15 on average, with baseline reading
levels of grades one to six. The students were assigned to a nine-week program of one-to-one
tutoring using direct instruction (Carnine et al., 1997) and Corrective Reading (Engelmann et al.,
19

The original Maryland Scale did not consider single-case designs, but the inclusion of these designs is relevant for
the juvenile correctional education literature, so we continue our practice of operationalizing the WWC and
Maryland Scales in corresponding ways.

37

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

1999) strategies. The average instructional dosage received was 21 sessions, or about a month of
instruction, and the average gains during that time were nine months on the combined rate-andaccuracy subtest of the Gray Oral Reading Test, 3rd edition, and nine months on the
comprehension subtest. In addition, the gain in words read correctly per minute was 3.57 per
week, which the authors compared to an expected gain of 1 word per week for students on a
fourth-to-sixth grade level receiving intensive instruction. In addition, a study of Corrective
Reading implemented in a juvenile facility in combination with whole-language instruction
(Malmgren and Leone, 2000) found statistically significant gains in the Gray Oral Reading Test
rate-and-accuracy score (0.35 of a standard deviation) as well as a statistically nonsignificant
gain in comprehension (0.34 of a standard deviation), among 45 juveniles ages 13–18 exposed to
a six-week summer reading program for about three hours per day, five days per week. As with
Coulter (2004), this study is not our eligible for our systematic review because it did not include
a comparison group. Unlike Coulter’s study, Malmgren and Leone (2000) did not present their
findings in terms of national or other norms, so it is especially difficult to estimate whether
similar gains would have been made using a different instructional approach over the same
period of time. Still, the findings suggest that oral reading gains ranging from a third to a half of
a pretest standard deviation are feasible within six to nine weeks for incarcerated students
receiving Corrective Reading instruction in combination with whole-language teaching. We turn
now to evidence from studies that were formally eligible for the systematic review.
Eligible Studies in Juvenile Correctional Settings: Single-Case Designs

As noted above, we consider single-case designs to meet strong evidence standards (i.e., to merit
a 5 on the Maryland Scale in terms of the strength of their causal inferences) if they meet the
single-case design standards set forth by the WWC. The two single-case design studies that
examine Corrective Reading met that standard in terms of establishing clear pre- and postintervention performance trends in at least three cases (Kratochwill et al., 2010). Moreover, the
two single-case studies of Corrective Reading, both using multiple-baseline designs, show
positive effects of the Corrective Reading intervention on the number of words read correctly per
minute.20 The mean gain for the Allen-DeBoer et al. (2006) study, implemented for an average of
30 30-minute lessons with four students ages 16–18, was 35.8 words per minute. This was
relative to mean baseline of about 93 words per minute, suggesting a roughly 38 percent gain on
average. Single-case designs, though they allow for causal inference, do not permit hypothesis
testing or the calculation of traditional effect sizes (Kratochwill et al., 2010). The study,
however, showed no evidence of an effect on word errors per minute. The mean gain for the
Drakeford (2002) study, which was undertaken with six students ages 12–21 for an average of 20

20

In a multiple-baseline design, the intervention is introduced to multiple students but at staggered starting points,
to separate any secular time trend from the treatment effect.

38

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

one-hour lessons, was 9.2 words read correctly per minute, on a mean baseline of about 77 words
per minute. Relative to the baseline level, this represents about a 12 percent gain.
Pooling the two estimates, the mean gain associated with Corrective Reading in the two
studies appears to be 1.66 words read correctly per minute for every hour of Corrective Reading
instruction, though this value is a rough average at the study-level and is, of course, based on
very small samples. Though single-case designs are considered a rigorous method for
undertaking studies with small samples (such as small groups of learning-disabled students), they
do pose a challenge for generalizability, since they require extrapolation from very small groups
of students and do not permit traditional hypothesis testing.
Eligible Studies in Juvenile Correctional Settings: Comparison-Group Designs

Two studies of Corrective Reading used eligible comparison-group designs. As noted, one was
an observational study that adjusted for baseline differences in test scores and thus merited a 3 on
the Maryland Scale. This 19-week study by Scarlato and Asahara (2004) compared 180 minutes
per week of Corrective Reading to 345 minutes of weekly specialist-directed reading instruction
in the comparison group among nine students ages 16–17. The study found substantial positive
effects of Corrective Reading on subscales of the Woodcock Reading Mastery Test Revised,
ranging from about 0.3 of a standard deviation in word identification and word comprehension,
to about 0.9 in passage comprehension and total reading. However, the comparison group
showed lower reading skills at baseline than the treatment group, so it is possible that the groups
also differed in unobserved ways. Moreover, the treatment effects were due in part to the
substantial losses (as large as 0.5 of a standard deviation) made by the comparison group who
worked with a reading specialist. Whether this is due to the particular skills of that reading
specialist in the comparison condition is unclear. In other words, the design makes it difficult to
disentangle treatment effects from effects of the types of students who were selected to receive
the treatment and the particulars of the comparison condition. For this reason, the results should
be interpreted with caution. Moreover, none of the observed effects, despite their large
magnitude, was statistically significant, which is the result of the very small sample size of nine
students.
Houchins et al.’s (2008) randomized trial comparing Corrective Reading with smaller versus
larger instructional groups included 20 students ages 13–17 receiving instruction for about 21
one-hour sessions. The findings for use of Corrective Reading in smaller groups (1:4 instead of
1:12) were substantial and positive for several outcomes, including word identification, word
attack, and silent reading (0.6, 0.5, and 0.7 of a standard deviation, respectively), but were not
statistically significant. (The word identification coefficient was reported to be significant at the
1-percent level, but the p-value increased to 0.058 when we adjusted for multiple comparisons
using the formula 1–(1–p)k, where p is the p-value, and k is the number of comparisons.) The
findings were positive but very small and statistically nonsignificant for oral reading at the third
grade level, and negative but nonsignificant for oral reading at the fourth and fifth grade levels.
39

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

The size of the small Corrective Reading group in this particular evaluation (four per group)
appears similar to that in the aforementioned studies, based on the size of the treatment groups
examined, which ranged from 4 to 6 across the studies.
Because Houchins et al.’s (2008) study was a rigorously designed randomized trial with low
attrition, it merited a 5 on the Maryland Scale. Its results are therefore likely free from
unobserved differences between treatment and comparison groups. However, because the study
did not include a condition without Corrective Reading, it permits only an assessment of the
program with a smaller group relative to a larger group. It is possible that the positive effects of
the smaller group size are not specific to the Corrective Reading curriculum but would have been
observed with other curricula as well. Still, the findings suggest that for those seeking to use
Corrective Reading, smaller instructional groups may be warranted.
Summary

Taken together, the preponderance of evidence on Corrective Reading in correctional settings
suggests that the intervention’s effects are positive for reading skills. The evidence seems
strongest for basic skills like word identification and decoding, since positive evidence for
comprehension comes only from the Scarlato and Asahara (2004) study, which merits only a 3
on the Maryland Scale. In addition, none of the studies was large enough to yield statistically
significant effects. This makes it difficult to generalize broadly from these findings. Based on
these four eligible studies, it would be difficult to state definitively that Corrective Reading is an
effective reading program for incarcerated juveniles. However, taken in conjunction with the
literature in the research context section, and in the absence of better-supported remedial reading
alternatives for correctional settings, Corrective Reading shows some promise. This promise is
best substantiated with lower-level skills like word identification and decoding.
Computer-Assisted Instruction
Research Context

Meta-analyses of the effects of computer-assisted learning outside of correctional education
settings have produced mixed signals about the effectiveness of these programs for raising
student achievement. In a meta-analysis of 17 studies spanning kindergarten through 12th grade
and published in 1982 through 1999, Soe, Koki, and Chang (2000) found positive effects of
computer-assisted instruction on reading achievement. The average impact estimate was 0.132 of
a standard deviation, though the estimates ranged widely, from a low of 0.045 to a high of 0.762
of a standard deviation. In a more recent systematic review of computer-assisted instruction
effects on reading, Slavin et al. (2008) found a weighted mean effect of 0.1 of a standard
deviation across eight eligible studies. Importantly, all of these studies focused on computerassisted instructional packages as standalone, supplemental interventions rather than as

40

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

components of integrated, blended curricula.21 In our own analysis of computer-assisted
instruction in adult correctional education settings, we found no statistically significant
difference between computer-assisted and face-to-face instruction in math and reading, though
the estimates were based on only four relatively underpowered studies, and the effects were
generally positive in magnitude (Davis et al., 2013).
However, computer-based instructional interventions vary widely, which makes it difficult to
generalize about them as a class of interventions. We therefore focus our discussion in this
section on the three computer-assisted interventions that were eligible for inclusion in our
juvenile correctional education systematic review: Read 180, Fast ForWord®, and Tune in™ to
Reading (TIR).
Read 180, published by Scholastic, is a complete reading curriculum for upper elementary
through high school–aged students that includes an adaptive, computer-assisted component, but
also includes teacher-led direct instruction and independent and small-group reading
components. The prescribed dosage is typically 90 minutes per day, five days per week
(Loadman, 2011; Scholastic, n.d.). It is part of a class of interventions that Slavin et al. (2008)
refer to as “mixed-methods models” and what others call “blended learning” models (Horn and
Staker, 2011), but, given that it prominently features an adaptive, computer-based component,
we classify it as a computer-assisted approach in our analysis. Based on seven studies that meet
evidence standards with reservations (comparable to a level 4 on the Maryland Scale), the WWC
found potentially positive effects for the impact of Read 180 relative to comparison curricula on
comprehension, with an average increase of 0.11 of a standard deviation, or 4 percentile points,
and on general literacy achievement, with an average increase of 0.31 of a standard deviation, or
12 percentile points. Based on eight studies, four of which overlapped with the WWC analysis,
Slavin et al. (2008) provide a weighted mean effect of 0.24 of a standard deviation for Read 180
relative to comparison curricula. However, none of the studies included in Slavin et al.’s
systematic review of computer-assisted instruction, or in the WWC review of Read 180, were
conducted in correctional education settings.
The Fast ForWord software-based intervention is published by Scientific Learning
Corporation. Unlike Read 180, a middle-grade-oriented reading program in which computerassisted instruction is one component, Fast ForWord is a completely computer-based curriculum
and is designed for beginning readers (Scientific Learning Corporation, 2004). It is intended for
use five days per week, for 30–100 minutes per day over a 4–16-week time period (What Works
Clearinghouse, 2013). In a review of seven studies that met its evidence standards and two that
met standards with reservations (comparable to a level 4 on the Maryland Scale), the WWC
found positive effects of Fast ForWord on alphabetics (e.g., decoding), no effect on reading
fluency (e.g., rate and smoothness), and mixed (i.e., both positive and negative) effects on
21

We use the term “blended” to refer to curricula that incorporate both computer-based and face-to-face
instructional methods (see, for instance, Horn and Staker, 2011; also, Childress, 2013).

41

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

reading comprehension. However, none of these studies was conducted in correctional education
settings.
The final computer-assisted intervention eligible for consideration in our systematic review is
Tune in to Reading (TIR), which was developed in 2005 and is published by Electronic Learning
Products. The TIR software, which contains more than 600 songs analyzed for readability level,
promotes reading fluency by teaching students to sing written words with the correct pitch and
tone (Calderone et al., 2009). Prior to being studied in a juvenile correctional education setting,
the intervention, known at the time as Carry-a-Tune, was evaluated for 90 minutes a week over
nine weeks in a rural west Florida middle school. The study, which was undertaken by Biggs and
colleagues (Biggs et al., 2008), found that the 24 students in the treatment group improved by
0.98 of a pooled pretest standard deviation, as compared with no change in the performance of
the 22 matched-comparison group students who were assigned to required reading for the same
amount of time (p < .001). Though the study was small, its substantial effects and use of a
matched comparison group (warranting a level 4 on the Maryland Scale) suggested the promise
of this intervention with struggling adolescent readers.
Studies with Ineligible (Level 1) Designs in Correctional Settings

As part of our consideration of the research context, we did find one study of the aforementioned
interventions that was administered in a correctional setting for juveniles, but with a design that
made it ineligible for the systematic review. The developer of Fast ForWord, Scientific Learning
Corporation (2004), conducted a study of the intervention with 29 youth incarcerated by the
Virginia Department of Correctional Education. Focusing on 18 students at one of the two
schools in the study, evaluators found that students improved by 1.5 grade equivalents on
average over a ten-month time period on the Woodcock Johnson Tests of Achievement–Broad
Reading. In the other participating school, the 11 students in the study gained nearly 1.5 grade
equivalents in a four-month time period as measured by the STAR Reading assessment. Both
sets of pre-to-posttest gains were statistically significant at the 5 percent level. However, because
neither subgroup included a comparison group, this study is rated as a level 1 design on the
Maryland Scale, we do not officially include it in our systematic review, and it therefore appears
in a shaded row within Table A.1.
Eligible Studies in Juvenile Correctional Settings

The three studies of computer-assisted instruction that qualified for the systematic review
focused on the three aforementioned interventions: Read 180, Fast ForWord, and Tune in to
Reading. Results for the studies are synthesized in Table 3.2 and are summarized in greater detail
in Table A.1. In the largest of the three studies, Loadman et al. (2011) conducted a randomized,
controlled trial of Read 180 with students incarcerated in eight correctional facilities run by the

42

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Ohio Department of Youth Services facilities.22 The study randomized five cohorts of students,
for an initial sample of 1,982 students. Because the final rates of overall and differential attrition
(37.2 percent and 2.5 percentage points, respectively) fell beneath the liberal threshold of the
WWC, the study merited a 5 in our operationalization of the Maryland Scale. The final analytic
sample included 1,245 students, ranging in age from 14 to 22. The students randomly assigned to
receive 90 minutes of daily instruction with the Read 180 curriculum instead of the same amount
of instruction with the default language arts curriculum showed gains on the Scholastic Reading
Inventory that were 0.21 of a standard deviation higher than the control group at the end of the
20-week intervention, a statistically significant difference (p < .001). Moreover, among the 243
students still enrolled at the correctional institution a year after baseline testing, the gains at that
time were 0.26 of a standard deviation greater on the California Achievement Test than for the
control group. The latter difference was also statistically significant (p = 0.011), though the oneyear effect could be rated only a 4 on the Maryland Scale due to the small proportion of the
sample still enrolled a year after baseline testing. Still, the Read 180 study represents one of the
largest and most well-executed studies we uncovered within a juvenile correctional education
setting, and should be viewed as a model for future efforts.
Fast ForWord was also evaluated in a randomized, controlled trial in juvenile correctional
setting, in this case in a long-term, maximum-security juvenile facility in Alabama (Shippen et
al., 2012). The study, which we rated a 5 on the Maryland Scale due to its complete lack of
attrition, involved 51 students ages 11–20. The study estimated the impact of exposure to Fast
ForWord reading relative to a default individualized instructional program. The duration of the
program was 45 minutes per day, five days per week, for 11 weeks, though in practice the
average exposure duration was only 24 days. The study found a positive impact on spelling skills
of 0.23 of a standard deviation, but a negative effect on the Test of Word Reading Efficiency (–
0.14 of a standard deviation) and on the Woodcock Reading Mastery Test-Revised/Normative
Update (–0.21 of a standard deviation). Due to the study’s limited statistical power, none of the
estimates was statistically distinguishable from zero, but the magnitude and direction of the
reading effects suggest that Fast ForWord was not an effective tool for raising reading
achievement within the study population.
Finally, Calderone et al. (2009) undertook a randomized trial in which Tune in to Reading
was compared against the default language arts instruction program in six juvenile correctional
facilities in Florida. The default control-group curriculum was FCAT Explorer, which is another
computer-assisted program, but one focused largely on drill and practice for the Florida
Comprehensive Assessment Test. The study randomized 138 incarcerated males in grades 7
through 11 to TIR versus the control condition within each site. After attrition, which fell
22

The findings for the first two cohorts, which were similar, though slightly smaller in magnitude, were reported in
Zhu et al. (2010). The findings discussed above reflect the full five-year study documented in the project’s final
report to the Institute of Education Sciences.

43

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

beneath even the conservative boundary of the WWC (at 25.4 percent overall and a differential
rate of 0.6 percentage points), the analytic sample included 103 students, and we rated the study
a 5 on the Maryland Scale. Students received computer-assisted instruction for 45 minutes, two
times per week, for nine weeks and were assessed with a computer-adaptive cloze (i.e., fill-inthe-blank) reading assessment developed by TIR but validated against the Qualitative Reading
Inventory (Leslie and Caldwell, 2000) and the FCAT.23 Students in the treatment group showed
gains that were 0.21 of a pooled pretest standard deviation greater than those of the control
group. This is a positive effect of nontrivial magnitude, though, given the small sample size, the
effect estimate did not approach statistical significance (p = 0.3).24
Summary

The three eligible studies that focused on computer-assisted instruction in juvenile correctional
settings were all well-executed, low-attrition randomized trials. Taken together, they showed
mixed results with regard to students’ reading skills. The evidence from correctional settings
comports with the larger bodies of evidence for Read 180 and Fast ForWord, respectively, in that
the extant evidence for Read 180 suggests that it raises achievement more than comparison
interventions, and Loadman et al. (2011) found a clear positive effect of the curriculum.
Meanwhile, the larger body of evidence for Fast ForWord is indeterminate regarding its
effectiveness, and the Shippen et al. (2012) study conducted in a juvenile correctional setting
showed positive effects for spelling and negative effects for reading, none of which were
significant. The Tune in to Reading study in a correctional setting did show a positive effect, but
the effect was not statistically significant and was much smaller in magnitude than what had
been found in the pilot study conducted in a regular middle school setting (Biggs et al., 2008). In
light of these studies and the larger body of extant evidence, there is little evidence to currently
support the use of Fast ForWord in a correctional setting, and larger studies are needed to
strengthen the research base on Tune in to Reading. Considered in combination with the large
and favorable body of research outside of correctional settings, the evidence in favor of using
Read 180 is the strongest for any of the interventions considered in this review.

23

The correlations were reported to be 0.7 with QIR and 0.56 with the FCAT, and test/retest reliability was
estimated at 0.86.

24

Wexler et al.’s (2013) systematic review reports a negative effect for Calderone et al. (2009), possibly due to
transposing the estimates for the treatment and comparison groups.

44

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Table 3.3. Summary of Findings for Other Juvenile Correctional Interventions
Intervention
Type
Personalized
and intensive
instruction

Academic
Effects
n Studies=3
n Participants=895
Level-5 studies: 1
Level-3 studies: 2
Effects: 27.1 percentagepoint increase in diploma
completion in level-5 study
(p<.01); 2 months of total
achievement gain (math,
reading, language) in level-3
study (no hypothesis test);
nonsignificant effect of 0.045 SD for peer- versus
teacher-managed approach
in level-3 study (p>.05)

Employment
Effects
n Studies=1
n Participants=714
Level-5 studies: 1
Effect: 8.0
percentage-point
increase in
employment 1 year
postrelease (p<.02)

Other remedial n Studies=1
instruction
n Participants=63
Level-3 studies: 1
Effect: 0.38 SD reading gain
per 10 hours of instruction
with Orton/Gillingham
reading curriculum relative
to default
(no hypothesis test)

Vocational
n Studies=1
education/CTE n Participants=1,046
Level-2 studies: 1
Effect: 7.6 percentage-point
increase in GED pass rates
(p<.001).

n Studies=1
n Participants=1,502
Level-3 studies: 1
Effect: 39 percent
increase in the odds
of employment 1 year
postrelease (p<.01).

45

Recidivism
Effects

Preponderance of
Evidence

n Studies=2
n Participants=759
Level-5 studies: 1
Level-2 studies: 1
Effects: 1
percentage point
increase in rearrest
within a year in
level-5 study
(p>.05); 29
percentage-point
reduction in rearrest
within 6 months in
level-2 study (p<.05)

The preponderance of
evidence about
personalized
instructional
approaches in juvenile
correctional settings is
positive for degree
completion, based on
one level-5 study. It is
also positive for
employment based on
one level-5 study. It is
mixed for recidivism
avoidance, based on
one level 5 and one
level-2 study. The
level-5 study should be
largely free of selection
bias.

n Studies=2
n Participants=568
Level-3 studies: 1
Level-2 studies: 1
Effects: 22
percentage-point
reduction in level-3
study (p=.015); 9.4
percentage-point
increase in level-2
study (p<.05)

The preponderance of
evidence about other
remedial programs is
positive for reading,
based on one level-3
study. It is mixed for
recidivism based on
one level 3 and one
level-2 study. Selection
bias remains a likely
threat.

n Studies=1
n Participants=1,905
Level-3 studies: 1
Level-2 studies: 1
Effects:
nonsignificant 3
percent reduction in
odds of rearrest
within a year in
level-3 study
(p=0.8); 17.1
percentage-point
reduction in
reincarceration
within 5 years in
level-2 study (p<.05)

The preponderance of
evidence about
vocational
education/CTE
participation while in a
juvenile facility is
positive for diploma
completion based on
one level-2 study and
is positive for
employment, based on
one level-3 study. It is
mixed for recidivism
avoidance, based on
one level-3 and one
level-2 study. Selection
bias remains a likely
threat.

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

GED
completion

n Studies=2
n Participants=2,266
Level-3 studies: 1
Level-2 studies: 1
Effects: 5.8
percentage-point
reduction in
reincarceration
within a year in
level-3 study
(p<0.1); 12.5
percentage-point
reduction in rearrest
within 3 years in
level-2 study
(p<.001)

The preponderance of
evidence about GED
completion while in a
juvenile facility is
positive for recidivism
avoidance, based on
one level-3 study and
one leve- 2 study, but
selection bias remains
a likely threat.

NOTE: We do not provide Maryland Scale or WWC effectiveness ratings because the number of studies is small and
some have very few students. Thus, we think it is premature to declare any of these interventions effective or
ineffective based on available evidence. Instead, we provide a column reporting on the preponderance of existing
evidence, which may be positive, mixed or negative for each outcome examined.

Personalized and Intensive Instruction
Research Context

Personalized learning is a broad term in the education literature, indicating that instruction is
adjusted to fit the unique needs and developmental trajectories of each student. The Bill and
Melinda Gates Foundation’s director of Next Generation Learning recently defined it as meaning
that “students’ learning experiences are tailored to their individual needs, skill levels, and
interests” (Childress, 2013). The term can serve as a loose synonym for several related concepts,
including student-centered instruction and differentiated instruction, and it shares many features
with competency-based education (Jobs for the Future, 2013; U.S. Department of Education,
2011). This is especially true among the three studies in the juvenile correctional education
systematic review, each of which is described as including competency-based attributes.
Competency-based learning is an approach in which students progress at their own pace, earning
credit not for the time spent in a course but for mastering a specified sequence of content targets
(Sturgis and Patrick, 2010; Priest et al., 2012). Given the high mobility and highly variable skill
levels of students in juvenile correctional facilities (Leone et al., 2002), personalized and
competency-based approaches may be particularly well suited for correctional education settings.
Competency-based models are well established in settings that serve students with atypical
academic progression. They are mainstays of “credit recovery” programs that help students who
lack the credits to graduate to catch up with their peers on via an accelerated schedule (Sturgis et
al., 2010; U.S. Department of Education, 2011).
Nevertheless, systematic evidence about the effectiveness of personalized and competencybased instructional models is quite limited, with a literature focused largely on anecdotes (Priest
et al., 2012). A few high-performing charter schools, such as Carpe Diem in Arizona (NBC
News and the Hechinger Report, 2013) and Young Women’s Leadership Charter School in

46

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Chicago (U.S. Department of Education, 2010) have outperformed similar schools on aggregate
measures, and a few other schools of choice that use competency-based measures have also
shown stronger academic performance than demographically similar schools (Steele et al., 2013).
But it in all cases, it is difficult to attribute these differences to the competency-based models
themselves rather than to unmeasured characteristics of the schools themselves and the families
that choose them.
Studies with Ineligible (Level 1) Designs in Correctional Settings

Our examination of the research context includes two studies that were conducted in correctional
education settings that were not eligible for inclusion in the systematic review because they did
not include appropriate comparison groups. As with other relevant studies that we rated a 1 on
the Maryland Scale, these ineligible studies appear in shaded rows in Table A.1.
The first of these is a very small study of personalized instruction within a juvenile
correctional setting conducted by Muse (1998). The study was ineligible for the systematic
review because it was based on student-level data only for the treatment group, and its
comparison condition used aggregate, school-level data from four comparison schools. Given the
lack of a same-level comparison group, we assign it a rating of 1 on the Maryland Scale. The
study focused on the GED completion rates in academic classes taught by the author in one
North Carolina juvenile correctional facility (about 66 students in total), relative to the average
completion rate of students at 4 other North Carolina juvenile correctional facility schools in the
state in the same years. Students in the study ranged in age from 12 to 17 years old. Examining
GED completion rates over a three-year period, the author reported that 67.1 percent of his
students earned GEDs, as compared with 8.0 percent of students in the other schools, though it is
not clear that the groups were comparable in terms of risk levels, age, length of stay, and so
forth. Though the effect size is substantial, it is difficult to say how much of the 59.1 percentagepoint difference in completion rates was due to the intensive, personalized instructional method
as opposed to other factors. Results should therefore be viewed with great caution.
The other study of personalized instruction that was relevant to the research context but rated
a 1 on the Maryland Scale was by Hill, Minifie, and Minifie (1984). They undertook a program
of diagnostic evaluation and personal tutoring as an 18-hour (nine-week) supplement to
traditional classroom instruction for 31 incarcerated juveniles in South Carolina who were
defined as handicapped. They found that students improved by about three months in silent
reading, one month in oral reading, and an unspecified amount in mathematics, but in the
absence of a comparison condition, it is not possible to say how much improvement the students
would have made without treatment during the same time period. For this reason, the study does
not meet eligibility standards for the systematic review.

47

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Eligible Studies in Juvenile Correctional Settings

Among the 18 studies that were eligible for the systematic review, four focused on personalized
or individualized instructional models in which adult instructors worked intensively with
students, tailoring instruction to the individual needs of each student. In each of these studies, the
comparison groups received the default instructional programming for students in their
respective correctional systems. These studies are synthesized in Table 3.3, and additional details
about each study are provided in Table A.1.
The largest and most rigorous of the studies was a randomized trial by the National Council
on Crime and Delinquency (2009) evaluating the Avon Park Youth Academy operated by the
Florida Department of Juvenile Justice. The study randomized 714 youth ages 16–18 who were
incarcerated in Florida to either the Avon Park program or to a control condition was
incarceration in a default juvenile justice program in Florida. Unlike the default program, which
featured traditional academic instruction, the Avon Park Youth Academy used an intensive,
personalized, competency-based instructional model tailored to each student’s academic
development. It also included vocational programming through the Home Builder’s Institute, and
aftercare upon release. The study was not able to disentangle the vocational and aftercare effects
from the program’s personalized overall approach. However, because the randomized groups
were tracked with administrative data, there was no attrition from the randomization sample,
meaning the study earned a level 5 on the Maryland Scale and a WWC “Meets Standards” rating.
Despite randomization, however, the dosages of the treatment and control models were different;
students in Avon Park were incarcerated for an average of 14.2 months, versus an 11.2-month
average for the comparison group. The study examined three different outcomes: rates of
diploma completion at release, employment one year postrelease, and recidivism (defined here as
re-arrest) within a year after release. With regard to diploma completion, it found a 26.9
percentage point higher rate among the treatment group relative to the control group (44.1 versus
22.0 percent, p < .01), where diploma completion was defined as earning a high school diploma,
a GED, or a special diploma for students with special education needs. Differences were
pronounced and statistically significant in all three diploma categories, though they were greatest
in the GED and special diploma categories. The study also found an eight percentage-point
difference in employment one year postrelease, with an average employment rate of 72.4 percent
among the treatment group and 64.4 percent among the control group (p < .05). However, it
found no statistically significant difference in recidivism rates. Within a year after release, the
treatment and comparison groups were rearrested at comparable rates (57.2 percent, and 56.2
percent, respectively).
The second study we identified as focusing on personalized instruction in a juvenile
correctional facility was considerably smaller than the Avon Park study and did not use a
randomized design. Skonovd and colleagues (1991) examined recidivism rates six months after
release among 45 youth ages 16–17 who were incarcerated in San Bernardino County. Twentyfive youth were assigned to the Regional Youth Educational Facility (RYEF), which was an
48

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

intensive, competency-based education program that included six months of education within the
juvenile facility, followed by four to six months of aftercare supervision by a probation officer
familiar with the youth through RYEF. (The standard caseload for the probation officers was
reduced from 65 to 15 for the Avon Park program.) The comparison group of 20 students
received the default educational program in the county. Assignment to treatment or comparison
group was determined by the juvenile courts, and the comparison group was constructed so that
it met RYEF eligibility criteria, even though the students in the comparison group had not been
referred to RYEF by the courts. The treatment and comparison groups consisted only of students
who successfully completed either the treatment or default program. Students who changed
programs or failed complete were excluded from the analysis, and the average length of stay for
the control group was two months longer than for the treatment group. Though baseline risk
indices were similar for the two groups, no statistical adjustments were made for observed
demographic differences between treatment and comparison-group students. For these reasons,
the study rates a level 2 on the Maryland Scale. The study found that rates of re-arrest or
probation violation within six months after release were only 16 percent in the treatment group,
versus 45 percent in the comparison group. Despite the small sample size, this 29 percentagepoint difference was large enough to be statistically significant (p < .05).
We also identified two studies of personalized learning that focused on academic test scores
as the dependent variables of interest. Mayer and Hoffman (1982) compared pretest to posttest
gains on the California Achievement Test Total Battery for 68 incarcerated students assigned to
individualized instruction with the gains for 75 students assigned to group instruction. The total
battery score included mathematics, reading, and language skills. In the individualized scenario,
lessons and activities were tailored to students’ prior skill assessments; in the group approach,
classes of about 12 students received daily instruction appropriate to their assigned grade level.
During a ten-month period, students in the individualized program showed nine months of
academic growth on average, while those in the group program made seven months of growth,
for a relative treatment effect of two months. Because effects accounted for the baseline
performance of each student, the study merits a 3 on the Maryland Scale. However, the study did
not provide information for assessing the statistical significance of this difference, nor did it
provide extensive information about other institutional and sample differences between groups.
For these reasons, it is difficult to attribute the difference in gains to the individualized program
itself, and the results must be interpreted with caution.
The final eligible study we identified that focused on personalized instruction did not actually
assess the personalization itself, but rather, compared a personalized instructional model
managed by a peer tutor to a similar model managed by a classroom teacher. This study,
conducted by Kane and Alley (1980), is relevant insofar as a peer-tutoring model may offer a
cost-effective alternative to a teacher-directed approach. Twenty-one students in the study were
assigned to a peer-managed classroom, in which the ratio of peer tutors to students was 1:1 or
1:2, and 17 were assigned to a teacher-managed classroom, in which the teacher-student ratio
49

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

ranged from 1:3 to 1:7. Importantly, the mean pretest math grade level of the tutors was 11.7, as
compared with a mean of 6.0 for the students, and the curriculum was the SRA Computation
Skills Development Kit published by Science Research Associates. The SRA kits provided selfpaced, competency-based instruction that allowed each student to demonstrate mastery of
content at his or her own pace (Proctor and Johnson, 1965). Therefore, in both the peer-managed
and teacher-managed conditions, student progress was largely self-paced, and the peer tutor or
teacher served to help keep students on track and answer their questions rather than to lead
whole-class instruction. After 38 lessons of 45 minutes each, researchers found that students in
the peer-managed classrooms made less mathematics progress than their peers, as measured by
the SRA assessment survey, with a relative change of –0.045 of a standard deviation. Because
this effect was not statistically significant at the 5 percent level, the researchers suggested that
there was no difference in performance between the peer-managed and teacher-managed
classrooms. However, given that the study was underpowered with only 38 students, only a large
effect would be sufficient to reject the null hypothesis. Thus, the fact that the effect estimate was
negative in magnitude provides a tentative cautionary note about the use of peer tutors as
opposed to classroom teachers. Also, because the study did not adjust for between-group
differences other than baseline performance, it merits a 3 on the Maryland Scale and suggests
mainly that more evidence is needed on the merits of peer- versus teacher-managed classrooms
in correctional settings.25
Summary

The eligible literature on personalized instruction for juveniles in correctional facilities includes
one large, no-attrition randomized trial: the Avon Park study conducted by the National Council
on Crime and Delinquency. Though it is just one study, it provides a convincing endorsement for
the effectiveness of the kind of intensive and personalized approach that the Avon Park program
employed in improving diploma completion and employment rates. However, evidence for the
effect of personalized and intensive learning on recidivism is mixed: The Avon Park study found
no effect on rates of recidivism within a year, while the less rigorous RYEF study found a large
and statistically significant reduction in recidivism. Given that the Avon Park study was 15 times
as large as the RYEF study and that the latter used a less-rigorous, level-2 design, evidence of a
positive relationship between intensive instruction and recidivism reduction remains thin at best.
One possibility is that some as-yet-undetermined threshold of impact on skills or employment
must be reached before such programs yield a reduction in recidivism.26
Regarding the effects of personalized learning on academic achievement, the Mayer and
Hoffman (1982) study showed a positive effect with unclear statistical significance, and the Kane
25

Wexler et al.’s (2013) review described the effect as positive rather than negative but did not adjust for the
baseline between-group differences reported in the article.
26

We thank a peer reviewer for raising this point.

50

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

and Alley (1980) study showed a slight negative—but not statistically significant—effect for
peer-managed personalized learning relative to teacher-managed personalized learning.
It is important to note that the definition of personalized learning varies not only in the
current education literature but also among these four studies. In the Avon Park and RYEF, the
intervention itself was multifaceted, encompassing competency-based academic instruction,
targeted services for special needs students, vocational education, and aftercare supervision by a
probation officer familiar with the students’ progress while incarcerated. However, in the studies
by Mayer and Hoffman (1982) and Kane and Alley (1980), the intervention was much narrower,
focusing mainly on self-paced classroom instruction. The field would benefit from studies that
compared simpler and more-complex personalized models using common outcome metrics, and
also that considered the relative costs of each model.
Other Remedial Instruction Programs
Research Context

In recent years, a growing body of literature has questioned the effectiveness of remedial
education for improving student outcomes. However, this literature has largely focused on
postsecondary education, where remedial education can slow a student’s progress and increase
the cost of earning a degree, thereby acting as a potential deterrent to degree completion
(Martorell and McFarlin, 2008; Caldagno and Long, 2008; Scott-Clayton and Rodriguez, 2012).
In secondary education, the need to remediate the learning gaps of students who fall behind
seems less controversial, the question being how best to do so. In many ways, the
aforementioned categories of Corrective Reading, computer-assisted instruction, and
personalized learning each offer answers to that question. All of the interventions in these
categories aimed, at least in part, to remediate the low academic achievement of students in
juvenile correctional facilities. However, our systematic review also uncovered two studies of
remedial programs that did not fit directly into one of the aforementioned categories, and we
consider evidence for those studies in this section.
Studies with Ineligible (Level 1) Designs in Correctional Settings

Interestingly, the number of studies of remedial education models that we deemed ineligible for
the systematic review but relevant to the research context—four—was twice the number of
studies in this category that were eligible for the systematic review. This is perhaps because our
“other remedial intervention” category acts as a catch-all for small studies that have examined
various remediation efforts on a small scale, without the benefit of robust research designs.
These four studies, which appear in shaded rows within Table A.1 due to their ineligibility,
collectively examine the effects of teaching efforts for reading road signs (Murph and
McCormick, 1985), completing a job application (Heward et al., 1980), using metacognitive
learning strategies (Platt and Beech, 1994), and using three distinctive writing prompts (Sinatra,
1984). All are small studies, and all show gains in the target behaviors, but none uses a
51

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

comparison group. Because all four ineligible studies are included in the recent systematic
review by Wexler (2013), we consider them here in some detail.
Two of the studies focus on students with very low reading levels who have been classified,
based on IQ testing, as having mild mental retardation. These two studies, by Murph and
McCormick (1985) and Heward and colleagues (1980), use single-case designs and focus on
improvement in fundamental, reading-related life skills—namely, reading nine road signs in the
former study and filling out an entry-level job application in the latter. The study by Heward
does not meet WWC single-case design evidence standards because it does not establish long
enough baseline trends (i.e., with at least four or five observations per phase) before introducing
the intervention, but it does show that students taught to fill out the applications raised their
average accuracy rate by 18 fill-in-the-blank items (out of 35) during 11 45-minute instructional
sessions. The study by Murph and McCormick does meet WWC standards in terms of the
number and length of pre- and post-intervention phases, but it suffers from what the WWC terms
an over-alignment of instruction and assessment (What Works Clearinghouse, 2008): After 9–24
instructional sessions of 15 minutes each, students who are repeatedly drilled in reading nine
road signs are able to read all of them. Though an effect of instruction is clearly established, the
measure of reading is simply too narrow to be construed as a measure of reading skill, and the
intervention itself is also too narrow to generalize to other contexts in which reading must be
taught. Though the authors justify the reading of road signs as a necessary life skill for driving, it
seems unlikely that 16-to-18-year-olds who struggle to identify common road signs will be able
to read and pass a written driving test—a fact that undermines the authors’ assertion about the
relevance of the task to their lives.
The two other studies deemed ineligible in this category also have serious design flaws. A
study by Platt and Beech (1994) used a single-case design approach, but with only one pretest
observation in each case, meaning that it does not meet WWC standards for single-case designs.
The authors show mean gains of 12 percentage points in words read correctly and 19 percentage
points in passage comprehension after an unspecified amount of instruction in metacognitive
learning strategies. However, they selectively present data for only five of the students taught by
the 27 educators trained to use the method, noting that only the “most complete” reports were
presented. Overall data for all students exposed to the method are not presented, nor are any data
presented about the number and demographics of the student sample. In other words, the data
should be construed as anecdotal at best.
Finally, Sinatra (1984) considers the effects of a writing intervention implemented in his own
classroom, in which struggling readers were given structured writing prompts that focused on
visual, imagery, and report-writing tasks. He finds a noteworthy gain of 16 percentage points on
students’ writing assignments after several months of instruction, but it is unclear how well the
students would have fared with writing instruction that did not include those prompts.

52

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

In short, the research context on remedial instruction for struggling incarcerated learners is
replete with research designs that, despite the benevolent intentions of authors and teachers, do
little to advance the field’s understanding of what works in remediation.
Eligible Studies in Juvenile Correctional Settings

We turn now to two studies of remedial interventions that had comparison groups and were
deemed eligible for the systematic review. Results from these studies are synthesized in Table
3.3 and are shown in greater detail in Table A.1. A study by Simpson, Swanson, and Kunkel
(1992) examined the effects of a structured remedial reading program, the Orton/Gillingham
reading curriculum, as compared with the default language arts program in two juvenile youth
detention facilities. The Orton/Gillingham program was described in the article as phonics-based,
with a focus on reading, writing, and spelling, and incorporating auditory, visual, and kinesthetic
learning modes. Researchers identified 55 learning disabled students for treatment, 32 of whom
persisted to analysis, and asked teachers to recommend a similar set of 61 students for the
comparison group, 31 of whom persisted to analysis. The treatment group received the
Orton/Gillingham curriculum for 90 minutes a day in groups of 1–6; the comparison group
students received default language arts instruction for only 45 minutes a day in classes of about
12. Thus, the treatment group received a larger instructional dosage per week and in smaller
classes. Treatment students may also have been released sooner, given that their mean
instructional hours were reported as 51.9 (about 35 days) versus 46.0 hours (about 61 days) for
the control group. Students in the analysis were tested at pretest and again before release using
the Woodcock Johnson Test of Reading Mastery, and their re-arrest rates were tracked within the
first year after release. Those in the treatment group gained 0.86 years of growth in reading more
than the control group, and the difference was statistically significant (p = .007), though the
treatment group also received a greater instructional dosage and in smaller classes. To eliminate
the instructional dosage confound, the authors estimated that the treatment group gained 0.38
years of growth more than the treatment group for every ten instructional hours, though they did
not provide a hypothesis test for this estimate. Finally, they found a substantial relationship
between the treatment condition and recidivism avoidance; those in the treatment group had a
one-year re-arrest rate of 41 percent, versus 63 percent in the comparison group, and this
difference was statistically significant (p = 0.015). Because the treatment and comparison groups
were both identified (at least by teachers) as learning disabled and shared similar baseline
reading scores (4.3 vs. 4.6) and ages of first arrest (13.7 and 13.8), and because the reading
analysis adjusts for baseline scores, we give the study a rating of 3 on the Maryland Scale.
However, it does remain vulnerable to unobserved differences between groups. For instance,
given that the treatment students appear to have been released sooner, on average, it is possible
that they also had a lower baseline risk of re-arrest that was not accounted for in the study.
The other eligible study of remedial education, by Archwamety and Katsiyannis, focused
only recidivism outcomes, measured between one and seven years after release. The study
53

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

compared 339 youth ages 12–18 enrolled in a mathematics or reading remedial program (not
both) while incarcerated to 166 students who were not enrolled in a remedial program. Students
who were assigned to remediation were at least one grade level behind at baseline in the remedial
subject, and had a lower mean baseline IQ than their nonremediated peers (91.8 versus 99.3).
The authors found that students assigned to remediation were actually 9.4 percentage points
more likely to recidivate (definition unspecified) within 1–7 years after release. Their recidivism
rate was 23.3 percent in the treatment group versus 13.9 percent in the comparison group, and
the difference was statistically significant (p < .05). However, the study did not adjust for
baseline differences, resulting in a rating of 2 on the Maryland Scale. In other words, given that
the treatment group clearly demonstrated higher baseline risk than the comparison group in terms
of weaker academic and cognitive skills, the study does not provide convincing evidence that
remediation caused higher recidivism among the treatment group, and it would be inappropriate
to draw conclusions about the impact of remediation based on this study.
Summary

The two studies in the systematic review yield different conclusions about the relationship
between remedial education and recidivism. Though the Simpson et al. (1992) study of the
Orton/Gillingham remedial curriculum was only about one-eighth the size of the Archwamety
and Katsiyannis (2000) study, it showed stronger equivalence at baseline and also provided
clearer details about differences between the treatment and control conditions. As such, it makes
a more convincing case that students remediated with the Orton/Gillingham program rather than
receiving standard language arts instruction improved faster in their reading skills and were less
likely to be rearrested. Nevertheless, it leaves open the possibility that treated students may have
had shorter sentences or other unobserved differences that may at least partially explain the
treatment effects, and thus the evidence supporting the Orton/Gillingham program remains
underdeveloped. Archwamety and Katsiyannis’s study is highly vulnerable to selection bias and
thus says little about the relationship between remediation and recidivism. Beyond the evidence
presented in previous sections about Corrective Reading, computer-assisted instruction, and
personalized instruction, it is difficult to draw broad conclusions about the effectiveness of other
remedial programs in juvenile correctional settings.
Vocational /Career Technical Education
Research Context

Among the population of U.S. secondary school students at large, the prevalence of vocational
training—now commonly termed career technical education (CTE)—declined between 1982 and
2004, with CTE credits accounting for 21 percent of the credits earned by high school graduates

54

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

in 1982, versus only 14 percent in 2004 (U.S. Department of Education, 2013b).27 In part, this
was a response to concerns that lower-achieving students were being tracked into vocational
pathways that did not prepare them to succeed in an increasingly competitive and dynamic labor
market (U.S. Department of Education, 2013b). Evidence on the effectiveness of vocational
education/CTE in raising academic outcomes is somewhat mixed. Using a nationally
representative sample of high school students tested in 10th and 12th grade, Bozick and Dalton
(2013) found no evidence that CTE course-taking improved or hurt students’ mathematics
performance. Studies that have used lottery-based random assignment have also found little to no
impact on test scores, though they have found benefits in terms of other outcomes. For example,
Kemple and Willner (2008) randomized New York high school students to career academies that
provided both vocational and academic training in combination with internships; they found
positive effects on high school persistence, as well as subsequent earnings benefits for males.
Neild, Boccanfuso, and Byrnes (2013) capitalized on students’ lottery-based random assignment
to five CTE high schools in Philadelphia, finding that students assigned to CTE had higher rates
of college preparatory mathematics coursework and higher graduation rates, though they
performed no better than their peers on mathematics and reading assessments. Insofar as a lack
of marketable skills increases the appeal of criminal behavior (Becker, 1968), it is possible that
juveniles involved in the criminal justice system may be especially likely to benefit from
programs that emphasize vocational skills.
In a meta-analysis of 548 effect estimates from 361 studies focused on juvenile offenders,
Lipsey (2009) considered evidence about a variety of programs designed to reduce recidivism.
His study, which included studies published between 1958 and 2002, differed from ours in that it
included both education and non-education programs, it focused on recidivism as the sole
outcome of interest, and only 22 percent of the estimates in his analysis were based on programs
administered within correctional facilities. Lipsey found that skill-building interventions—
defined to include behavior management, cognitive-behavioral therapy, social skills training,
challenge programs, academic training, and job-related interventions including vocational
training—reduced subsequent recidivism by about 6 percentage points, though the effect was not
statistically significant at the 5 percent level. He also found that the effects of skill-building
interventions did not depend on the implementation context—whether in juvenile correctional
facilities or with non-incarcerated offenders. Disaggregating skill-building effects by program
subtype, he estimated that job-related training programs reduced recidivism by about 2.8
percentage points; however, this estimate was not statistically significant and was smaller in
magnitude than the 6 percentage-point estimate for the skill-building category as a whole.
Our own meta-analysis of adult correctional education also looked in particular at the effects
of correctional vocational education/CTE for adults, finding that participation in vocational
27

We use the term “vocational education/CTE” for consistency with our adult meta-analysis, and to reflect the
language in the studies we are able to synthesize in this section, though CTE is the more-contemporary term.

55

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

education/CTE programs while incarcerated reduced adults’ odds of recidivism by 36 percent
relative to no participation in correctional education (Davis et al., 2013).
Eligible Studies in Juvenile Correctional Settings

Our systematic review identified three eligible studies that examined the effects of participation
in a standalone vocational education/CTE program in a juvenile correctional facility. These are
summarized in Table 3.3, and additional details about each are provided in Table A.1. Roos
(2006) examined the employment and recidivism rates for participants of the Re-Integration of
Offenders–Youth (RIO-Y) career development course operated by the Texas Youth
Commission. The sample included 1,502 incarcerated individuals ages 18–21, an age group
classified as juveniles within the program. Five hundred eight-two of these juveniles participated
in the 30-day RIO-Y program; another 920 received no career development course or other
instruction of note during that time period. (The RIO-Y study is unique in our systematic review
in that the comparison group did not receive an alternative instructional program during the
intervention period. This is possible because all were 18 years of age or older.) Students were not
randomly assigned to the program, but the analysis adjusted for 17 baseline demographic and
risk-related covariates, so it warrants a level 3 rating on the Maryland Scale. The study reported
that the odds of employment one year after release were 39 percent higher among the treatment
than the comparison group, and the difference was statistically significant (p < .01). This would
correspond to about a 7.1 percentage point increase in the probability of employment, using the
64.4 percent employment base rate one year postrelease reported in the aforementioned National
Council on Crime and Delinquency (2009) Avon Park study. (Roos’s study does not report a
base rate.) The study also found that the odds of re-arrest within a year after release were 3
percent lower in the treatment group, but this very small effect did not approach statistical
significance (p = 0.8), so the author concluded that the program had affected employment but not
recidivism.
Wilson (1994) also investigated the effects of vocational education/CTE in a juvenile
correctional facility. This study examined the re-arrest rates within five years after release among
403 juveniles, ages 11–18, incarcerated by the Colorado Division of Youth Services. Two
hundred sixty of the students participated in a vocational education/CTE training program while
incarcerated, and 143 did not. The study did not provide details about program attributes or
dosage levels. Assignment to treatment status was not random, and the analysis did not adjust for
baseline differences, so the study merits a level 2 rating on the Maryland Scale. Bearing in mind
that there was no adjustment for selection, the study found a statistically significant 17.2
percentage point reduction in re-arrest rates, from 78.3 percent in the comparison group to 61.2
percent in the treatment group (p<.05).
Finally, DelliCarpini (2010) examined the effect of offering a new vocational education/CTE
program to youth ages 16–21 within a county jail in New York State. The new program
augmented the existing academic instructional program with classes in business, drafting, and
56

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

carpentry. All youth incarcerated in the facility were enrolled in the program (though dosage
information was not reported), and outcomes for the first-year cohort (2008–2009) of 465
students were compared with outcomes for the prior-year cohort of 581 students. The rate at
which students earned a GED increased by 7.6 percentage points from 5.5 to 13.1 percent—a
statistically significant gain (p < .001). The gain was driven in part by an 8.2 percentage point
increase in eligibility to test (from 7.1 to 15.3 percent), as determined by scores on a practice
test, and also by an 8 percentage point increase in pass rates (from 78 to 86 percent) among those
who took the GED. The study rates a level 2 on the Maryland Scale because it did not adjust for
(or report on) baseline differences between the two cohorts, and because it did not adjust for
secular time trends that may have accounted for differences in outcomes between two sequential
cohorts.
Summary

Based on these three studies, the preponderance of evidence for vocational education/CTE in a
juvenile correctional facility is positive for GED completion and postrelease employment and
mixed for avoidance of recidivism. None of the studies are impervious to selection bias, but the
Roos (2006) study does attempt to mitigate such bias through statistical controls, and it is also
the largest study. Bearing that in mind, the positive evidence for employment effects seems
firmer than for GED completion or recidivism effects. Still, the results from these studies are not
sufficiently rigorous to be considered definitive. Taken as preliminary evidence, the
preponderance of extant research on vocational education/CTE in correctional settings does
appear to support further study of such programs, preferably with rigorous methods that allow
for causal inferences about their effects.
GED Completion
Research Context

Two of the studies in our systematic review examine the relationship between earning a GED
while incarcerated and subsequent recidivism. Among adult incarcerated populations, our metaanalysis found that completion of education while incarcerated reduced inmates’ odds of
recidivating by about 43 percent. Among the 22 studies (and 28 effect size estimates) that
focused in particular on high school credential or GED programs in correctional facilities, the
corresponding rate reduction in odds was about 30 percent (p < .05). However, these studies
pertain to the approximately 37 percent of adults in prisons who lack high school diplomas
(Crayton and Neusteter, 2008).28 In contrast, nearly all juveniles in correctional facilities lack
diplomas because most are still of high school age. The question, then, among the studies in
juvenile facilities that have focused on GED completion, is whether juveniles who earn a GED
28

This figure applies to state prisons in 2004. The comparable figure for the general U.S. population ages 16 and
older is 19 percent, according to the same report.

57

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

while incarcerated are subsequently less likely to recidivate. The question is very difficult to
answer in a causal sense—that is, does earning a GED reduce recidivism?—because juveniles
remain in the facilities for very different lengths of time, and their ability to complete a GED
during that time may depend on many factors, including their length of stay, their overall
academic preparedness, and their opportunities to pursue a traditional high school diploma
instead of a GED. Unfortunately, these and other potential confounds are not fully captured in
the studies we consider, and thus it is not appropriate to view the two studies, which rate at levels
2 and 3, respectively, on the Maryland Scale, as estimating the causal impact of the GED on
recidivism. Still, the studies help to illuminate the association between GED completion and
postrelease recidivism.
The broader evidence on GED attainment is mixed with regard to educational and economic
outcomes. In the general population of high school dropouts, Heckman and Rubinstein (2001)
found that earning a GED is negatively related to subsequent earnings, hourly wages, and levels
of additional schooling, after controlling for cognitive ability as measured by the Armed Forces
Qualifying Test. They attributed this negative effect to lower levels of noncognitive skills, such
as persistence, planning, and adaptability, among those who earned GEDs relative to other
individuals who did not complete high school. This still raises the question of what the value of
the GED credential is, holding all else constant, including motivation to pursue a GED. Tyler,
Murnane, and Willett (2000) exploited between-state variation in GED passing scores to separate
the effects of the GED credential from the underlying ability and motivation levels among high
school dropouts who take the GED. They found that the GED credential itself improved white
recipients’ earnings five years later by 10–19 percent, though it appeared to have no effect on
recipients from other racial/ethnic groups.
However, the aforementioned studies did not estimate the effects of GEDs earned in prison
settings. In a rigorously executed follow-up study by Tyler and Kling (2007), the authors found
that, among incarcerated adults, earning a GED in prison yielded earning gains of 15 percent in
the first two years after release, though this benefit was concentrated among nonwhite GED
earners and dissipated after the second year. In addition, they found that most of the benefit came
from participating in a GED education program rather than from actually earning the GED itself.
The implication is that incarcerated individuals may benefit more from the human capital they
acquire by participating in a GED program than from the credential itself. In other words, the
acquisition of skills is what matters.29 This finding is consistent with our own meta-analysis of
adult correctional education, in which we found that participation in an academic educational
program while incarcerated increased the odds of postrelease employment by 8 percent, and that

29

As one reviewer noted, it is possible that the availability of the credential is a factor that motivates inmates to
acquire the skills, and that they would learn less if the credential were not available to them. These studies do not
address the role of the credential in motivating learning.

58

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

participating in a high school diploma or GED program, in particular, reduced the odds of
recidivism by 30 percent (Davis et al., 2013).
Eligible Studies in Juvenile Correctional Settings

Given that all juveniles are expected to pursue a high school diploma or GED while incarcerated,
we did not identify studies that specifically examined the effects of GED program participation
(relative to no participation) for juvenile offenders. However, we did identify two comparisongroup studies that examined the effects of earning a GED while incarcerated in a juvenile
correctional facility. We summarize our findings about these studies in Table 3.3, and we
provide additional details about each study in Table A.1. Unlike the aforementioned GED
literature, which focused mainly on postrelease earnings, the two GED studies in our systematic
review examined the relationship between earning a GED in a juvenile facility and the
probability of recidivating after release. The larger study, by Jeffords and McNitt (1993),
examined re-incarceration rates within a year after release among 1,717 juveniles ages 16–21
held in the Texas Youth Commission or Gulf Coast Trades Center correctional facilities in
Texas. Among the 475 youth who earned GEDs while incarcerated, the estimated rate of reincarceration was 5.8 percentage points lower than among the 1,242 youth who did not, though
the difference was significant at only the liberal 10-percent level. Though students were not
randomized to their GED completion status, this estimate is based on a regression model that
controls for gender, ethnicity, age at release, risk classification, previous felony referrals,
previous adjudications and incarcerations, and severity of most recent offense. For this reason, it
merits a level 3 on the Maryland Scale. The unadjusted difference without controls was 9
percentage points (p < .1), which suggests that including the controls mitigated some selection
bias that exaggerated the GED effect.
The second study, by Katsiyannis and Murray (1999), examined re-incarceration rates within
three years after release among 549 youth who had been incarcerated for at least four months in a
Nebraska rehabilitation and treatment facility. Two-hundred eighty-four of the students
completed GEDs while incarcerated, and 265 did not. The re-incarceration rate among those who
earned GEDs was 47.5 percent, versus 60.0 percent among those who did not. The 12.5
percentage point difference was statistically significant (p < .01). The study also fits a logistic
regression model with a subset of 260 students that controls for age at first commitment and
improvements in test scores. Though that analysis shows a positive and significant effect, its
sample restriction and inclusion of post-baseline controls lead us to use the unadjusted rates
instead. The comparison is rated a level 2 on the Maryland Scale.
Summary

The preponderance of evidence for earning a GED while incarcerated suggests that juveniles
who do so experience lower rates of re-incarceration within one to three years after they are
released. This effect was estimated at 12.5 percentage points in the level-2 study by Katsiyannis
59

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

and Archwamety (1999, but it was only 5.8 percentage points in the level-3 study by Jeffords and
McNitt (1993). Given that the Jeffords and McNitt study found larger effect in the unadjusted
model, their more-rigorous estimate of 5.8 percentage points is likely a better true estimate of the
GED benefit, and even that may be positively biased by selection on unobservable
characteristics. Because GED completion is at least partly a function of student motivation and
aptitude rather than of differences in institutional programming, estimates of the impact of GED
completion are especially vulnerable to selection bias. A more convincing approach would
capitalize on external forces affecting incarcerated students’ access to GED testing, while
holding academic skills and other observable attributes constant. In the absence of more-rigorous
GED studies in juvenile correctional settings, we can conclude only that those who succeed in
earning a GED while incarcerated appear less likely to recidivate after release. The extent to
which obtaining the GED causes this difference remains an open question.

Discussion
Limitations of Our Approach
Our systematic review of correctional education interventions for incarcerated juveniles reveals
great heterogeneity in terms of interventions, methods, and outcomes of interest. Among the 18
eligible studies we identified, we classified the interventions into six categories: Corrective
Reading, computer-assisted instruction, personalized instruction, other remedial education,
vocational education, and GED completion. Studies in the first two categories focused on
packaged and branded reading interventions (Corrective Reading, Read 180, Fast ForWord, and
Tune in to Reading) and focused on reading performance as the dependent variables of interest.
Studies in the latter three categories focused on a broader set of outcomes, including not only
reading and mathematics performance but also measures such as diploma completion,
postrelease employment, and postrelease recidivism. The wide variety of hypotheses tested in the
studies, the broad array of outcomes examined, and the small number of studies in each
hypothesis-by-outcome category make it difficult to synthesize the findings into statements about
the relative effectiveness of each approach. In fact, because the number of studies eligible for
review is quite small, and because many of the studies are themselves very small in terms of
sample sizes, we mostly refrain from giving definitive pronouncements or ratings about the
effectiveness of any particular type of intervention. Instead, we comment on the preponderance
of evidence for each intervention category, based on the 16 comparison-group studies and the
two well-executed single-case design studies featured in the evidence summary tables, Tables 3.2
and 3.3. To contextualize those studies, we have also provided short descriptions of the broader
research context for each intervention type.
The effectiveness of any one of the intervention types in the study is likely to depend on how
well the program is implemented and for what period of time. Given the small number of studies

60

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

in each category, we cannot easily extrapolate the effects of differential dosages or
implementation approaches. However, the review does include a couple of exceptions. One of
the Corrective Reading studies was specifically designed to measure the effects of instructional
group size; it found a positive but statistically nonsignificant impact of smaller groups relative to
larger groups that were using the Corrective Reading curriculum (Houchins et al., 2008). And
one of the studies of personalized instruction found a small but statistically nonsignificant
negative effect of a peer-managed versus teacher-managed approach (Kane and Alley, 1980).
We are also unable to extrapolate differential effects by participant characteristics. The
studies vary in the level of detail they provide about participant characteristics, and we lack
enough common studies testing the same hypothesis to examine differential effects by subgroup.
Still, to facilitate policy decisionmaking among those using this review, we do report in Table
A.1 on the demographic and baseline achievement features of the samples in each study, insofar
as that information was present in the source material.
Key Insights for the Research Community
We found that the methods employed in the studies varied markedly by intervention type.
Studies of the packaged reading interventions were generally fairly small, because these studies
involve administering particular curricula at the classroom or student level, as well as
administering pre- and post-tests to individual students. Two of the Corrective Reading studies
utilized single-case designs that involved just four and six students each, but even the two
comparison-group studies included only nine and 20 students, respectively. The designs of these
studies were fairly robust, with one level-5 randomized trial and two level-5 single-case designs,
but the small size of the studies and limited power for hypothesis testing still makes it difficult to
generalize broadly from their findings. The studies of computer-assisted instruction also
employed well-executed randomized trials, though the two studies varied in size. The Fast
ForWord randomized trial included only 51 students, again providing limited statistical power,
but the Read 180 randomized trial was well powered, with 1,245 students; it was able to detect a
positive and statistically significant effect using a rigorous design that was fairly impervious to
selection bias. Additional studies of this type within juvenile correctional facilities should be
encouraged where possible.
The two studies of personalized and intensive instructional approaches were heterogeneous
in size and design; the National Council on Crime and Delinquency’s Avon Park study was a
well-powered and well-executed randomized trial that included 714 youth and was able to
examine diploma completion, employment, and recidivism-related outcomes, finding positive
and statistically significant effects on the first two variables. The San Bernardino County study
was much smaller, with only 45 students, and was more vulnerable to selection bias, which may
have in part accounted for the substantial recidivism reduction effect it uncovered.
The vocational education/CTE and GED studies were similar in that they took advantage of
large, administrative datasets to compare students exposed to particular treatments while
61

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

incarcerated (vocational programs, or GED credential completion) to those who were not
similarly exposed. In both categories, one study employed an array of statistical controls to
adjust for observed differences between treatment and comparison groups, earning a level 3
rating on the Maryland Scale. The other studies in each category compared those exposed to
those not exposed without adjusting for selection, and thus warranted a rating of 2. Comparing
the estimates of the level-3 and level-2 studies in the same intervention category and on a
common outcome (namely, recidivism) suggests that selection bias may, indeed, have inflated
the estimates in the level-2 studies, and even level-3 studies cannot adjust for selection on
unobserved characteristics.
It is also notable that none of the studies in the systematic review earned a level-4 rating,
which requires that the treatment and comparison groups be nearly identical on relevant baseline
characteristics, as is sometime achieved by propensity score matching or other matching methods
(Rosenbaum and Rubin, 1983; Shadish et al., 2008). Instead, the studies in the systematic review
can be generally characterized as small-to-mid-scale randomized trials or as large observational
studies with minimum-to-moderate use of statistical methods to adjust for unobserved
differences.
This suggests that the field is ripe for larger-scale randomized trials. The Loadman et al.
(2011) Read 180 study and the National Council on Crime and Delinquency’s (2009) Avon Park
study suggest that such studies, though challenging to undertake, are feasible. The literature is
also ripe for rigorous evaluations of natural experiments such as Aizer and Doyle’s (2013) study
of the effects of juvenile incarceration using naturally occurring random assignment to harsh
judges. Studies that take advantage of rigorous causal methods in juvenile settings can shed
much-needed light on what works in these settings. Several of the smaller randomized trials we
include here have noted the difficulties of high student turnover in correctional facilities, and of
simply gaining permission to undertake research in these facilities (Shippen et al., 2012,
Calderone et al., 2009). Such research efforts will clearly take time to develop and execute. They
will ideally be realized through long-term partnerships between researchers and correctional
facilities. Because such partnerships take time to establish, there may also be a federal role in
galvanizing them. The U.S. Department of Education Institute of Education Science’s recent
grant program for supporting research partnerships between school systems and researchers
offers one potential model. Guided by such partnerships, facilities can make increasingly
evidence-based decisions that not only improve their students’ prospects but also reduce the
social incidence of crime and delinquency.
Key Insights for Policymakers and Practitioners
Taken in conjunction with the broader research literature on each of the interventions examined,
our systematic review does identify two interventions for which the evidence base is strongest:
Read 180 (for reading improvement) and the kind of personalized and intensive intervention
administered at the Avon Park Academy (for diploma completion and postrelease employment).
62

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Both of these interventions are supported by a large and rigorous study within juvenile
correctional settings, and the effectiveness of Read 180 is further substantiated by several large
and well-executed studies outside of correctional facilities. Beyond these strong bodies of
research, we find that evidence for Corrective Reading and Tune in to Reading is positive but
based on very small studies from which it is difficult to generalize. Our review also highlights a
few juvenile correctional education interventions, such as the Fast ForWord software program
and peer-managed instruction, for which the current (though limited) body of knowledge offers
little support at the present time.
We are more reluctant to offer even cautious endorsement for interventions in which the
strongest studies are rated 3 or below on the Maryland Scale, even if they show positive effects,
as is the case with the Orton/Gillingham remedial education program. This limitation in research
quality also applies to the personalized interventions we reviewed other than Avon Park, the
vocational education/CTE interventions, and GED completion as an intervention.
In fact, the benefits of earning a GED while incarcerated, though estimated as positive in the
systematic review, remain especially unclear, since these studies’ comparisons of students who
earn a GED with those who do not are especially vulnerable to selection bias at the student level.
The most rigorous research from the literature on incarcerated adults suggest that it is the
education acquired in GED programs rather than the GED credential itself that confers the
greatest postrelease benefits (Tyler and Kling, 2007; Davis et al., 2013). This finding is largely
consistent with the GED research outside of correctional settings as well (Heckman and
Rubinstein, 2001).
Though the evidence base about what works in juvenile correctional education remains
incomplete, the existing research does offer guidance about promising directions for future
programmatic investments. In the interim, program directors who make decisions based on
extant evidence can play a critical role in documenting their interventions and reporting on the
outcomes by using the most rigorous methods at their disposal.

63

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

4. RAND Correctional Education Survey

Introduction
When we began our correctional education study, we recognized early on that the 2008 recession
had a substantial effect on the field of correctional education, with many states reporting cuts in
funding for programs and changes to their delivery models for educating incarcerated adults,
including such changes as shortening the length of time individuals spent in programs, reducing
the number of teachers, reducing the number of program slots, and cutting some programs
altogether. Such changes mean that today correctional education in the U.S. likely looks very
different from correctional education during the time that many of the studies in our metaanalysis were undertaken. Understanding these differences helps us to put in context the metaanalytic results described in Chapter Two and to provide the basis for forward-looking policy
recommendations presented in Chapter Five.
In July 2013, we fielded the RAND Correctional Education Survey to better understand the
key issues facing correctional education today. State correctional education directors’ responses
to this survey provide us with insights into how states dealt with the recession of 2008, how
correctional education is currently provided to incarcerated adults in the United States, what
information technology is being used, and how states fund correctional education. We also
gathered information on preparations for the new 2014 GED exam.
In this chapter, we first summarize our approach for the survey and then present the results of
the survey analyses, concluding with a discussion of what the survey results inform us about the
field of correctional education for incarcerated adults.
Approach
Survey Design

The purpose of the RAND Correctional Education Survey was to gather information about the
organization and delivery of correctional education for incarcerated adults in U.S. state prisons,
about the use of computer technology and preparations for the 2014 GED exam, and about the
impact of the 2008 recession on the field. The intent of this national survey was to help fill a
critical void in our understanding of the organization and delivery of academic and vocational
education/CTE to incarcerated adults and of how the landscape of correctional education is
changing.
To inform the development of the survey, we held discussions with experts in the field and
conducted four 90-minute focus groups with state correctional education directors responsible
for adult education, adult vocational education/CTE, and/or juvenile correctional education.
Specifically, we conducted two focus groups on adult academic programs, one focus group on
64

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

vocational education/CTE programs for adults, and one focus group on juvenile correctional
education. The participants were recruited in advance using a purposive sampling design to attain
geographic representation and to include representatives from states considered to be leaders in
field. A total of 30 individuals participated in these focus groups, which were conducted at the
2011 Correctional Education Association Leadership Forum. The focus group discussions
identified several key themes facing the field. The purpose of these group discussions was to
help us identify some of the key trends in this field. During the focus group discussions, we
learned about a number of issues that were facing the field of correctional education, including
the effects of budget cuts as a result of the 2008 recession, the need to gain efficiencies in
providing education to incarcerated adults and juveniles, the increasing role of computer
technology in academic education and vocational education/CTE, and the challenges the new
2014 GED exam and computer-based testing. We leveraged the insights and information from
these discussions to inform the development of the specific survey items, which were designed
by the project team and then underwent several reviews by our research partners at the
Correctional Education Association and the U.S. Department of Education. The final
questionnaire was loaded into a web survey and fielded using RAND’s Multimode Interviewing
Capability (MMIC) system, whose staff programmed and fielded the web survey.
Sample

The survey was distributed to the state correctional education directors in all 50 states. The
contact list was generated by searching public documents and verified by the Correctional
Education Association.
Fielding the Survey

An advance letter from RAND accompanied by a letter by the BJA and the U.S. Department of
Education explaining the importance of the survey was sent to each director approximately two
weeks before the survey was fielded in July 2013. This was followed by an email invite to the
directors to participate in the web survey, which provided them with the web link and their
unique login name and password. In addition, we made available a PDF version of the web
survey for those directors that preferred to fill out a paper version of the survey or who asked for
a copy so they could see what information they would need to collect to complete the web
survey.
We sent several follow-up email reminders to those directors who had not yet completed or
begun the web survey. In addition, the Director of the State Council of Directors, Correctional
Education Association, assisted us in sending out several reminders to their membership
encouraging participation. Two team members also called individual directors whose states that
had not yet started or completed the web survey to encourage participation. Survey responses
were accepted through October 2013.

65

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Measures

Shaped by our discussions with correctional education leaders across the country, the survey (see
Appendix B for the questionnaire) included questions about the following topics:
•
•
•
•
•
•
•

Key components of correctional education programs within each state
Capacity of correctional education programs and how it changed between 2009 and 2012
Impact of budget cuts or other fiscal pressures
Use of technology
Preparations for the 2014 GED exam and computer-based testing
Outcome and performance indicators tracked by states’ correctional education programs
Budget and financing.

Data Cleaning and Analysis

As a result of the various forms of outreach, data came to RAND in several different forms. Most
data were received through the MMIC system, but respondents also sent emails or added
information in an open-ended comment field at the end of the survey to provide data on specific
items or to otherwise clarify responses. Once the survey was closed, data cleaning involved
identifying the data provided through these methods and merging these with the final dataset.
Because skip patterns were built directly into MMIC, the logic of responses that depend on
earlier responses was maintained. However, respondents sometimes added information to the
“other” category that could be coded as a previously listed response. In these cases, we recoded
responses for consistency.
We appended to the dataset information on the size of each state’s adult prison population in
2009 and 2012 using data from the Bureau of Justice Statistics (Carson and Sabol, 2012; Carson
and Golinelli, 2013). We undertook a descriptive analysis and present the results overall, by size
of state and by type of lead agency responsible for administering adult correctional education
within states. Because this is a census of all state correctional education programs, we do not
calculate inferential statistics. Using data from the Bureau of Justice Statistics (Maruschak 2012),
we classified states by the size of their adult prison populations in 2012. We consider small states
to have had an adult prison population in the range of 1–24,999, medium states to have had an
adult prison population in the range of 25,000–49,999, and large states to have had 50,000 or
more adult prisoners in 2012.

Results
The overall response rate for the survey was 46 out of 50 states, or 92 percent, and importantly
included at least partial participation from all of the states with large prison populations (n =
50,000 adult inmates or greater). In addition, 40 of the respondents (87 percent) had
responsibility for both academic education and vocational training for incarcerated adults within
their state. Only five respondents were responsible for academic education only, and one
respondent was responsible for vocational training only. Of the state directors who responded,
66

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

only four stopped before completing the entire questionnaire. The number of states on which the
findings are based is noted for each table and figure presented below.
Overview of Correctional Education Programs Today
For most responding states (36), the majority of correctional education program authority is
vested within one central state agency; it is shared among several state agencies in three states. In
30 of the states, the authority for correctional education programs resides primarily within their
states’ departments of corrections or public safety; only four states indicated that the primary
authority was a state department of education or department of adult education, and one
respondent indicated that it resided within their state department of labor (data not shown).
Table 4.1 summarizes the types of educational programs available to adult state prisoners.
Most states (44) reported offering adult basic education, general education development (GED)
courses, and vocational skills training/CTE. Forty states also reported offering special education.
Thirty-two states also offered adult secondary education and/or adult postsecondary education,
and 33 states offered ESL courses. Smaller states were somewhat less likely to offer adult
secondary education and postsecondary education courses. In addition, 30 out of 46 states
indicated that within their state’s prison system correctional education is considered to be an
offender work assignment (data not shown), meaning that work assignments are also considered
part of correctional education.
Table 4.1. Number of States Offering Educational Programs to Adult State Prisoners, by Type of
Program
Size of State
Overall

Small

Medium

Large

Type of Program

N (%)

N (%)

N (%)

N (%)

Adult basic education

44
(96%)

27
(93%)

10
(100%)

7
(100%)

Adult secondary education

32
(70%)

19
(66%)

7
(70%)

6
(86%)

GED test preparation

44
(96%)

27
(93%)

10
(100%)

7
(100%)

Adult postsecondary
education/college courses

32
(70%)

18
(62%)

9
(90%)

5
(71%)

Vocational skills training/CTE

44
(96%)

27
(93%)

10
(100%)

7
(100%)

English as a second language
(ESL) courses

33
(72%)

22
(76%)

7
(70%)

4
(57%)

Special education

40
(87%)

25
(86%)

9
(90%)

6
(86%)

Other

17
(37%)

12
(41%)

3
(30%)

2
(29%)

67

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Total number of states responding

46

29

10

7

NOTE: Percent represents the percentage for each column of the number of states who responded.

In recent years, there has been an increasing emphasis on offering vocational education/CTE
programs that lead to a nationally or industry-recognized certificate. The data in Table 4.2
provide insights as to what types of vocational training programs are currently being provided to
adult inmates. Construction and automotive training were two important trades in which
instruction was provided. Overall, the trade certifications most commonly reported were in
construction (28 states), occupational safety (20 states), plumbing or electrical apprenticeships
(20 states), automotive service (19 states), and welding certification (14 states). The perceived
importance of general computing skills is underscored by the reports that 24 states out of the 42
states that responded to this question also offered Microsoft Office certification. Very few states
indicated that no nationally or industry-recognized certifications were offered to adult inmates.
Table 4.2. Number of States Offering Nationally or Industry-Recognized Certifications
Size of State
Overall

Small

Medium

Large

Certifications Offered

N (%)

N (%)

N (%)

N (%)

National Center for Construction
Education and Research

28
(67%)

16
(57%)

7
(88%)

5
(83%)

Microsoft Office certification

24
(57%)

15
(54%)

5
(63%)

4
(67%)

Occupational Safety and Health
Administration training programs

20
(48%)

13
(46%)

4
(50%)

3
(50%)

Apprenticeship cards (e.g.,
plumbing, electrical)

20
(48%)

11
(39%)

5
(63%)

4
(67%)

National Institute for Automotive
Service Excellence

19
(45%)

12
(43%)

4
(50%)

3
(50%)

American Welding Society

14
(33%)

11
(39%)

1
(13%)

2
(33%)

Our state does not offer nationally
or industry-recognized certificates

2
(5%)

2
(7%)

0
(0%)

0
(0%)

Total number of states responding

42

28

8

6

NOTE: Percent represents the percentage for each column of the number of states who responded.

Overall, participation in correctional education programs is mandatory in 24 states for adult
inmates without a high school diploma or GED and mandatory in 15 states for adult inmates
below a certain grade level (Table 4.3). We hypothesized that as a result of the 2008 recession
and budget cuts, some states might change their requirements to make correctional education

68

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

voluntary as one way of trimming costs. In fact, 32 out of 35 states indicated that they had not
changed their state’s policy from mandatory to voluntary participation (data not shown).
Table 4.3. Degree to Which Participation in Correctional Education Programs Is Mandatory
Size of State
Overall

Small

N (%)

N (%)

N (%)

N (%)

Participation is mandatory for adult
inmates without a high school
diploma or GED

24
(52%)

14
(48%)

5
(50%)

5
(71%)

Participation is mandatory for adult
inmates below a certain grade
education level

15
(33%)

6
(21%)

7
(70%)

2
(49%)

Participation in correctional
education programs is voluntary for
all inmates (i.e., not mandated by
state policy or by legislation)

21
(46%)

16
(55%)

3
(30%)

2
(29%)

Other

2
(4%)

1
(3%)

0
(0%)

0
(0%)

46

29

10

7

Total number of states responding

Medium

Large

NOTES: Respondents were asked to check all that apply. Therefore, there is overlap in the responses between the
first two categories (mandatory for adults without a high school diploma or GED and mandatory for adult inmates
below a certain grade level). For the “Overall” column, the “other” category included two responses that education
was mandatory based on age.

Funding of Correctional Education and Impact of the 2008 Recession
Overall, the recession resulted in a decrease on average in the size of states’ correctional
education budgets. In 2009, the mean correctional education budget reported by survey
respondents was $114,546,927, and this dropped to $100,760,235 in 2012 (Table 4.4). Overall,
36 states reported that between FYs 2009 and 2012 their state’s correctional education programs
(academic and/or vocational education/CTE) experienced a decrease in funding (Table 4.6). The
average change in the total correctional education budget in each state between 2009 and 2012
was a decrease of 6 percent, but there were differences by size: Small states experienced a 2
percent increase on average, whereas medium states experienced an average decrease of 20
percent and large states an average decrease of 10 percent (Table 4.4).
Table 4.4. Mean Change in State Correctional Education Budgets Between FY2009 and FY2012
Size of State

69

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Total Correctional Education
Budget
Mean budget in FY2009
(number of states)

Overall

Small

Medium

Large

$114,546,927
(n = 30)

$7,281,225
(n = 17)

$18,444,125
(n = 8)

$633,014,800
(n = 5)

Mean budget in FY2012
$100,760,235
$6,567,571
$15,550,286
$529,846,167
(Number of states)
(n = 34)
(n = 21)
(n = 7)
(n = 6)
1
Mean change in budget
–6%
2%
–20%
–10%
(number of states the calculation
(n = 29)
(n = 17)
(n = 7)
(n = 5)
is based on)
1
Mean changes in budget are calculated as the mean of each state’s change in budget from 2009 to 2012. They are
calculated on the sample of states that provided valid data in 2009 and 2012 and thus, may not represent the
average change for all states.

Another way to examine the decrease in states’ correctional education budgets is to calculate
the mean change in dollars spent per student. In FY 2009, the mean dollars spent per student in
correctional education programs was $3,479, and this decreased to $3,370 in FY 2012 (Table
4.5). Overall, the average change in the mean correctional education dollars spent per student
was a decrease of 5 percent between 2009 and 2012. Similar to the results shown in Table 4.4,
medium-sized states experienced the largest average decrease, 16 percent in the mean dollars
spent per student.
Table 4.5. Mean Change in States’ Correctional Education Budgets Per Student Between FY2009
and FY2012
Size of State
Overall

Small

Medium

Large

Mean dollars per student in FY2009
(number of states)

$3,479
(n = 24)

$1,710
(n = 13)

$1,213
(n = 6)

$10,794
(n = 5)

Mean dollars per student in FY2012
(number of states)

$3,370
(n = 21)

$1,590
(n = 19)

$1,666
(n = 6)

$10,711
(n = 6)

Mean change in dollars per student

–5%

4%

–16%

–1%

23

13

5

5

Number of states the change
calculation is based on

1Mean changes in dollars per student are calculated as the mean of each state’s change in mean dollars per
student from 2009 to 2012. They are calculated based on the sample of states that provided valid data in 2009 and
2012 and thus, may not represent the average change for all states.

We asked respondents specifically about the impact of budget cuts or other fiscal pressures
on different aspects of their states’ correctional education systems. Of the 36 states that reported
a decrease in funding for their correctional education systems, 27 states reported a reduction in
the number of teachers for academic programs, and 25 states reported a reduction in the number
of instructors for vocational educational/CTE programs (Table 4.6).

70

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Table 4.6. Number of States Reporting Decrease in Funding Between FY2009 and FY2012
Size of State
Overall

Small

Medium

Large

Change Implemented in
Response to Budget Cuts or
Other Fiscal Pressures

N (%)

N (%)

N (%)

N (%)

Number of states reporting a
decrease in funding for their state’s
correctional education programs
between FYs 2009 and 2012

36
(84%)

21
(75%)

9
(100%)

6
(100%)

Total number of states responding

43

28

9

6

Number of states that indicated
changes were made to the number
of teachers or instructors

31
(86%)

16
(76%)

9
(100%)

6
(100%)

Total number of states responding

36

21

9

6

Number of states that reported
reductions in the:
•

Number of teachers for
academic programs

27
(87%)

14
(88%)

8
(89%)

5
(83%)

•

Number of instructors for
vocational education/ CTE
programs

25
(81%)

11
(69%)

8
(89%)

6
(100%)

31

16

9

6

Total number of states responding

In response to reduced budgets, states in general reduced their staffing levels and the capacity
of their correctional education programs. Of the 36 states that reported a decrease in funding for
their state’s correctional education programs (Table 4.6), a major cost-cutting measure for 31
states was to not fill vacant teaching or instructor positions (Table 4.7). Other cost-cutting
measures included the implementation of hiring freezes (21 states), delayed or canceled pay
increases for staff (20 states), staff furloughs (11 states), reductions in the number of course
offerings for academic programs (20 states) and for vocational education/CTE programs (16
states), and reductions in or elimination of contracts with community colleges or technical
schools (17 states). Further, ten states reported that they anticipated additional budget cuts to
correctional education programs in FY2013 (data not shown).
Table 4.7. Of Those States That Reported A Decrease in Funding Between FY2009 and FY2012,
Changes Made in Staffing Levels and Capacity in Response to Budget Cuts
Size of State
Overall

Small

Medium

Large

Changes Made to Staffing Levels
and Capacity

N (%)

N (%)

N (%)

N (%)

Did not fill vacant teaching/instructor

31

17

8

6

71

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

positions

(89%)

(81%)

(100%)

(100%)

Hiring freeze of teachers/instructors
was implemented

21
(60%)

9
(43%)

7
(88%)

5
(83%)

Delayed and/or cancelled pay
increases for teachers/instructors

20
(57%)

10
(48%)

6
(75%)

4
(67%)

Staff furloughs of
teachers/instructors were made

11
(31%)

6
(29%)

3
(38%)

2
(33%)

Reduced salaries and/or benefits for
teachers/instructors

6
(17%)

2
(10%)

2
(25%)

2
(33%)

Reduced or eliminated contracts
with community or technical
colleges

17
(49%)

8
(38%)

5
(63%)

4
(67%)

Reduced the number of course
offerings for academic programs

20
(57%)

10
(48%)

5
(63%)

5
(83%)

Reduced the number of course
offerings for vocational
education/CTE programs

16
(46%)

8
(38%)

4
(50%)

4
(67%)

Other

8
(23%)

5
(24%)

2
(25%)

1
(17%)

None

1 (3%)

0 (0%)

1 (13%)

0 (0%)

35

21

8

6

Total number of states responding

Smaller states appear to have been less likely than medium-sized and larger states to have
reduced the number of course offerings for academic and vocational education/CTE programs
(Table 4.7). In general, smaller states appeared to be less likely to use the range of cost-cutting
measures listed in Table 4.7 than medium-sized or large states. It may be that smaller states had
less leeway to reduce staff or course offerings than other states. We did not ask whether teachers
or instructors were given additional duties instead, though this may help explain why fewer
smaller states did not employ the full range of cross-cutting measures listed.
The impact of these budget cuts was a reduction in the mean number of students participating
in academic education and vocational education/CTE programs, particularly within mediumsized and large states (Table 4.8). In 2009, the mean number students enrolled in academic
programs was about 8,300. By 2012, the mean dropped to 6,918. Between FYs 2009 and 2012,
the average change reported by states was an overall decrease of 4 percent. Medium-sized and
large states in particular saw reductions in the number of students in academic programs. Among
small states, the average change in the number of students enrolled in academic programs was a
decrease of 1 percent, but medium and large states reported larger average decreases (10 percent
and 8 percent, respectively)
The average number of vocational education/CTE students in each state was smaller (3,935
in 2009; 3402 in 2012) than that for academic education programs (Table 4.8). On average, there
was a 1 percent increase in the number of students enrolled in vocational education/CTE

72

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

programs (Table 4.8).30 It appears that this was largely due to an increase of 7 percent in the
mean number of students enrolled in vocational education/CTE programs within small states;
whereas the medium-sized and large states reported on average a decrease of 4 percent and 11
percent, respectively, in the number of students enrolled.
Table 4.8. Impact on Mean Number of Students Enrolled in Academic Programs and Vocational
Education/CTE Programs, FYs 2009 and 2012
Size of State
Impact on Students

Overall
(Mean)

Small
(Mean)

Medium
(Mean)

Large
(Mean)

Number of students enrolled in
academic programs
•

FY2009

8,321 (n=31)

3,524 (n=19)

10,180 (n=6)

21,656 (n=6)

•

FY2012

6,918 (n=39)

3,356 (n=25)

8,985 (n=8)

19,002 (n=6)

–4%

–1%

–10%

–8%

Mean change in academic program
1
enrollment
Number of students enrolled in
vocational education/CTE programs
•

FY2009

3,935 (n=30)

1,683(n=17)

5,352 (n=7)

8,663 (n=6)

•

FY2012

3,402 (n=37)

1,777 (n=23)

4,807 (n=8)

7,758 (n=6)

Mean change in vocational
1
education/CTE students

1%

7%

–4%

–11%

1

Calculated as the mean of each state’s change in the number of students from 2009 to 2012. It is calculated on the
sample of states that provided valid data in 2009 and 2012 and may not represent the average change for all
states.
NOT: For one small state, we set the value for number of students enrolled in academic programs in 2009 to
missing because of uncertainty in the data the state reported. As a result, this state does is not included in the
calculation of the mean change in academic program enrollment.

Based on our discussions with state correctional education directors, we hypothesized that
the recession would result in a decrease in the number of teachers and instructors who were
employees and a possible increased reliance on contract personnel. The effect of the staffing and
capacity changes summarized in Table 4.7 was an overall decrease in the mean number of
academic teachers who were employees from an average of 110 in 2009 to an average of 85 in
2012—representing on average a 24 percent decrease (Table 4.9). This decrease was largely
driven by what was occurring in the medium-sized and large states. Small states experienced a
modest average decrease of 5 percent; however, medium-sized and large states on average
experienced a 44 percent and 20 percent decrease, respectively. At the same time, we expected to
see an increased reliance on contract academic teachers. On average, states experienced a
30

The overall mean shows a decrease (3,935 to 3,402), likely because the average decrease in the larger states
offsets the small increase in the smaller states.

73

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

decrease of 1 percent in the number of academic teachers who were contract personnel. When we
look at size of state, both the small and large states experienced a decrease on average of 20
percent and 40 percent, respectively. The large increase of 136 percent reported by the mediumsized states in the mean number of academic teachers who were contract personnel was largely
driven by what was occurring within a few states in this size category.
In terms of instructors for vocational education/CTE, on average, states reported an increase
of 8 percent in the mean number of instructors who were employees between 2009 and 2012 and
reported (Table 4.9). In the small and medium-sized states, we see on average an increase of 8
percent and 24 percent, respectively in the number of instructors who were employees, which
suggests a modest expansion of vocational education/CTE programs in these states. At the same
time, the large states reported a 7 percent decrease, on average, in the number of vocational
instructors that were employees. In general, the large percentage changes among academic
contract personnel and vocational contract personnel among medium-sized states reflect very
large changes by a couple of states, but relatively minor changes in the others. For example, one
state increased its academic contract personnel from 0 to more than 70 teachers; another state
increased its vocational contract personnel from 0 to more than 30 during this time period.
Table 4.9. Mean Change in the Number of Employee or Contract Teachers by Type of Program,
FYs 2009 and 2012
Size of State
Number of Teachers or
Instructors

Overall
(Mean)

Small
(Mean)

Medium
(Mean)

Large
(Mean)

Number of academic teachers who
are employees
•

FY2009

110 (n=38)

34 (n=24)

165 (n=8)

342 (n=6)

•

FY2012

85 (n=41)

39 (n=27)

96 (n=8)

275 (n=6)

–24%

–5%

–44%

–20%

Mean change in number of
academic teachers who are
1
employees
Number of academic teachers who
are contract personnel
•

FY2009

11 (n=35)

15 (n=22)

6 (n=8)

5 (n=5)

•

FY2012

12 (n=41)

12 (n=27)

13 (n=9)

5 (n=5)

–1%

–20%

136%

–40%

Mean change in number of
academic teachers who are contract
1
personnel
Number of vocational instructors
who are employees
•

FY2009

56 (n=39)

18 (n=25)

105 (n=8)

152 (n=6)

•

FY2012

56 (n=43)

19 (n=28)

116 (n=8)

141 (n=6)

74

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Mean change in number of
vocational instructors who are
1
employees

8%

8%

24%

–7%

Number of vocational instructors
who are contract personnel
•

FY2009

2 (n=37)

3 (n=24)

2 (n=9)

0 (n=5)

•

FY2012

4 (n=41)

3 (n=28)

9 (n=8)

0 (n=5)

27%

–10%

250%

N/A

Mean change in number of
vocational instructors who are
1
contract personnel
1

Mean change calculated as the mean of each state’s change in the number of teachers (by type) from FYs 2009 to
2012. It is calculated on the sample of states that provided valid data in both FYs 2009 and 2012 and may not
represent the average change for all states.
NOTE: N/A indicates that there were no such instructors in either year.

Other changes to instructional support included an increased reliance on inmates as peer
tutors in the classroom. Between FYs 2009 and 2012, 14 out of 36 states reported increasing
their use of inmate/peer tutors in the classroom. Of those that did, 14 states used inmates as peer
tutors to assist students with coursework, nine states had inmates assist with vocational
education/CTE programs, eight states had inmates assist with administrative tasks, and three
states had inmates help oversee computer labs (data not shown).
Postsecondary Education
Historically, a key piece of legislation that helped to make postsecondary education more
available to incarcerated adults is the Higher Education Act of 1965 (Pub. L. 89-329), which
provided student loans to any qualified student, including prisoners, for postsecondary education.
In part due to this federal funding, in the 1970s through the 1980s there was a growth in the
number of state prison systems offering postsecondary education courses. However, in the early
1990s, Congress excluded prisoners from the Pell Basic Education Opportunity Grant with
passage of the Violent Crime Control and Law Enforcement Act of 1994 (Pub. L. 103-322). The
elimination of Pell Grant funding for prisoners led to the closure of approximately half of the
existing postsecondary correctional education programs within correctional facilities (Taylor,
2005).
The result was a decrease in the percent of state prison facilities that offered college courses
from 30 percent in 1995 to 26 percent in 2000 (Crayton and Neusteter, 2008). Participation by
state prison inmates in college courses also declined over time from 14 percent in 1991 to 7
percent in 2004 (Crayton and Neusteter, 2008).

75

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Our survey results provided an update on these historical trends. In 2013, as noted in Table
4.1, 32 states reported offering postsecondary education or college courses31 to adult inmates,
with the medium and larger states more likely to offer such courses than the smaller states. With
the Pell Grant exclusion, our survey results indicate that today postsecondary education or
college courses in many states (28) are paid for primarily by the individual inmate or through the
use of family finances, or by private funding such as foundations or individual donations (20
states) (Table 4.10). State funding is used by 16 states. Only 12 states use college or university
funding to cover the costs of postsecondary education, and very few states use inmate benefits or
welfare funds. Larger states are less likely to rely on inmates’ personal or family finances and
more likely to use state funding or college or university funding to cover postsecondary
education costs.
Table 4.10. Funding Sources Used to Pay for Adult Inmates’ Postsecondary Education or College
Courses
Size of State
Overall

Small

Medium

Large

Funding Sources

N (%)

N (%)

N (%)

N (%)

Personal or family finances

28
(62%)

17
(61%)

8
(80%)

3
(43%)

Private funding (e.g., foundations,
religious/community group,
individual donation)

20
(44%)

12
(43%)

6
(60%)

2
(29%)

State funding (e.g., department of
corrections’ budget allocation)

16
(36%)

7
(25%)

5
(50%)

4
(57%)

College or university funding

12
(27%)

5
(18%)

4
(40%)

3
(43%)

Inmate benefits or welfare funds

7
(16%)

4
(14%)

2
(20%)

1
(14%)

Not Applicable, our state does not
offer postsecondary/college courses
to adult inmates

7
(16%)

5
(18%)

1
(10%)

1
(14%)

45

28

10

7

Total number of states responding

NOTE: Respondents were asked to mark all that apply and so the column totals exceed 100 percent.

Use of Technology and Preparedness for Implementation of the 2014 GED Exam
A growing trend in the field of education and in correctional education is the use of computer
technology. Gorgol and Sponsler (2011) surveyed 43 states on their postsecondary correctional

31

One should keep in mind that postsecondary education can include courses that lead to a vocational certificate as
well as college coursework in general.

76

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

education programs and concluded that correctional educators looked to technology as an
innovative way to improve the delivery of postsecondary education and to increase access.
Based on the insights from the focus groups we conducted and discussions with key experts, we
included specific questions in our survey of state directors to gather data on what type of
computer technology is currently being used for educating adult inmates, on instruction methods
that leverage technology, and the degree of access to the Internet by teachers and inmate
students. Forty-one states reported that at least one of their state’s prison facilities had a
computer lab, with the median number of facilities with a computer lab being 11 (data not
shown).
The use of computers in correctional education programs is common. Thirty-nine states
reported use of desktop computers (either standalone or networked), and 17 states reported use of
laptops for their correctional education programs (Table 4.11). The use of tablets, such as
Kindles or iPads, was reported by only two of the small states, and 13 states reported the use of
other technology (specifically, eight states reported the use of Smartboards). In terms of
networks, 26 states (62 percent) reported their correctional education program utilized a local
area network (LAN), and 11 states (26 percent) reported using a statewide or wide area network
(WAN). Only 10 states reported using closed-circuit television for correctional education
instruction.
Table 4.11. Types of Technology Hardware and Networks Used in States’ Correctional Education
Systems
Size of State
Overall
(n=42)

Small
(n=28)

Medium
(n=8)

Large
(n=6)

Type of Hardware or Networks
Used

N (%)

N (%)

N (%)

N (%)

Desktop computers (standalone or
networked)

39
(93%)

27
(96%)

7
88%

5
(83%)

Local area network (LAN)

26
(62%)

15
(54%)

5
(63%)

6
(100%)

Mobile laptops

17
(40%)

9
(32%)

5
(63%)

3
(50%)

Statewide or wide area network
(WAN)

11
(26%)

7
(25%)

2
(25%)

2
(33%)

Closed-circuit TV

10
(24%)

5
(18%)

4
(50%)

1
(17%)

Tablets (e.g., Kindles, iPads)

2
(5%)

2
(7%)

0

0

Other technology

13
(31%)

6
(21%)

5
(63%)

2
(33%)

42

28

8

6

Total number of states responding

77

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Small states were more likely to rely on desktop computers (standalone or networked),
whereas medium-sized and large states were more likely to utilize laptops (Table 4.11). Mediumsized states were more likely to report use of smart boards and closed-circuit TV for instruction.
Table 4.12. Instruction Methods for Academic Programs or Vocational Education/CTE Courses
Size of States
Overall

Small

Medium

Large

N (%)

N (%)

N (%)

N (%)

On-site instruction

42
(100%)

28
(100%)

8
(100%)

6
(100%)

Correspondence courses

15
(36%)

9
(32%)

4
(50%)

2
(33%)

Interactive Video/satellite instruction

3
(7%)

1
(4%)

2
(25%)

0
(0%)

One-way Video/satellite instruction

2
(5%)

1
(4%)

1
(13%)

0
(0%)

One-way Internet-based instruction

1
(2%)

1
(4%)

0
(0%)

0
(0%)

Interactive Internet-based
instruction

1
(2%)

1
(4%)

0
(0%)

0
(0%)

Other technology

4
(10%)

2
(7%)

1
(13%)

1
(17%)

42

28

8

6

Instruction Methods Used

Total number of states responding

In terms of methods of instruction, 42 states reported use of on-site instruction to provide
academic or vocational/CTE courses, and 15 states reported the use of correspondence courses
(Table 4.12). Although ten states reported that they had closed circuit television (Table 4.11),
few states actually used it to provide one-way or interactive video/satellite instruction (Table
4.12). The use of the Internet-based instruction (one-way or interactive) was only reported by 1
state.
In general, student access to the Internet is very limited in most states. Thirty states (73
percent) of states reported that only teachers and instructors have access to live Internet
technology in the classroom (Table 4.13). In 26 states, students do not have access to any
Internet technology, and in 16 states students have access to only simulated Internet programs.
Use of simulated Internet programs appeared to be more prevalent in medium-sized states.
Our survey results are further supported by the findings from a recent survey of state
correctional executives. The Association of State Correctional Administrators (ASCA) asked
their membership about whether their agency planned to provide their state prison inmate
population access to online education courses to obtain a GED diploma or advance degree. Only
four states indicated their agency planned to provide inmates with access to online education
78

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

courses for the GED, two states indicated so for inmates to earn advanced degrees, and two states
for inmates to earn professional or vocational certification (ASCA, 2013).
Table 4.13. Number of States Offering Access to the Internet for Correctional Education Programs
Size of State
Overall

Small

Medium

Large

Type of Access to the Internet

N (%)

N (%)

N (%)

N (%)

Only teachers/instructors have
access to live internet technology in
correctional education classrooms

30
(73%)

21
(75%)

7
(88%)

2
(40%)

41

28

8

5

Total number of states responding
Student access to the internet
•

Students do not have
access to any Internet
technology

26
(62%)

18
(64%)

5
(63%)

3
(50%)

•

Students may only use
simulated Internet
programs

16
(38%)

10
(36%)

4
(50%)

2
(33%)

•

Students have restricted
access to live Internet

6
(14%)

4
(14%)

1
(13%)

1
(17%)

42

28

8

6

Total number of states responding

2014 GED Preparedness

In 2014, a new GED assessment will be implemented. The new, more rigorous test will be
aligned with the Common Core State Standards (CCSS) and will use a new test delivery
model—computer-based testing to replace the paper-and-pencil examination. These two changes
have important implications for correctional administrators and educators in terms of preparing
for and implementing the new test. Educators will need to be prepared to teach the CCSS and
prepare students for a more rigorous GED test that will require students to demonstrate highlevel thinking skills and exhibit deeper levels of knowledge in four subject areas. In addition, the
new test delivery model will require educators to prepare students to have a level of computer
literacy and skills necessary to successfully navigate the test using a computer. These changes, in
turn, have implications when it comes to agency budgets and professional development needs of
educators and present a number of logistical concerns when it comes to preparing to implement
computer-based testing. We asked correctional education directors about their preparations for
the new GED exam and for their views regarding early concerns about what these changes might
mean for their correctional education systems.
Thirty-one states reported that their state planned to implement the 2014 GED exam, with
nine states indicating that their state was exploring other high school equivalency exams and two
states indicating their state did not plan to implement the new GED exam (data not shown).

79

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Of those that planned to implement the 2014 GED exam, on average 14.5 of their state’s
correctional facilities (median was 11 facilities) were currently set up or were expected to be set
up by January 1, 2014, to implement the exam (data not shown).
Of those planning to implement the 2014 GED exam, 17 states planned to use a combination
of computer workstations and laptops for inmates to take the exam. Eight states indicated that
they planned to use only computer workstations, and two states planned to use laptops only (data
not shown).
Of the 31 states who planned to implement the 2014 GED exam, all but one indicated that
they would provide professional development training for their teachers and instructors to
prepare them to teach the new GED exam; the one exception indicated no such training would be
provided (data not shown). Table 4.14 summarizes the type of professional development training
states are providing or plan to provide to assist teachers and instructors in preparing to teach and
implement the 2014 GED exam and computer-based testing. Most states planned to conduct
training on the testing process, test protocols, and test security requirements. Most states also
planned to train on instruction aligned with the common core standards and on computer literacy.
Table 4.14. Professional Development Training for Teachers/Instructors to Prepare Them to Teach
the New 2014 GED Exam
Size of State
Overall

Small

Medium

Large

Topics Training Will Address

N (%)

N (%)

N (%)

N (%)

Training on instruction aligned with
the common core standards

29
(97%)

21
(100%)

3
(75%)

5
(100%)

Training on the testing process

28
(93%)

20
(95%)

3
(75%)

5
(100%)

Training on the test protocols

26
(87%)

18
(86%)

3
(75%)

5
(100%)

Training on test security
requirements

26
(87%)

18
(86%)

3
(75%)

5
(100%)

Training on computer literacy

25
(83%)

17
(81%)

3
(75%)

5
(100%)

Other

2
(7%)

1
(5%)

1
(25%)

0
(0%)

30

21

4

5

Total number of states responding

State correctional education directors are concerned about the more rigorous 2014 GED
exam and the implementation of computer-based testing. In our survey, 14 states expected that
these changes may have a negative effect on the number of adult inmates who will be prepared to
take the new exam, 13 states expected a negative effect on the amount of time needed to prepare
for the exam, and 16 states expected a negative effect on GED completion rates (data not
shown). Our findings are similar to that of a recent survey by the Association of State
80

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Correctional Administrators in which 21 states reported that they anticipated a sizable drop in
their pass rate for inmate students as a result to the switch to computer-based testing for the GED
(Association of State Correctional Administrators, 2013).
Table 4.15 summarizes what concerns, if any, state correctional education directors have
with respect to the 2014 GED exam and the move to computer-based testing. Of the 31 states
planning to implement the 2014 GED exam, 24 of the states expressed concerns about the length
of time it may take to prepare students to take the new GED exam, and 22 of the states were
concerned about the cost to their institution or correctional education program to prepare for
computer-based testing. Teachers being adequately prepared to teach the new exam (19 of the
states) and to implement computer-based testing (14 of the states) were concerns for a number of
states. Twelve of the states reported concerns that limited access to computers may possibly
preclude some students from taking the new GED exam, and ten of the states were concerned
about the cost of the 2014 GED exam to the individual student. Only two of the states indicated
they had no concerns about the new exam or computer-based testing.
Table 4.15. Concerns About Forthcoming Changes to the 2014 GED Exam and the Move to
Computer-Based Testing
Size of State
Overall

Small

Medium

Large

Areas of Concern

N (%)

N (%)

N (%)

N (%)

Length of time it will take to prepare
students to take the GED exam

24
(83%)

16
(76%)

3
(100%)

5
(100%)

Cost to the institution or program of
preparing for computer-based
testing

22
(76%)

14
(67%)

3
(100%)

5
(100%)

Teachers may not be prepared to
teach the new GED exam’s
components

19
(66%)

11
(52%)

3
(100%)

5
(100%)

Teachers may not be prepared to
implement computer-based testing

14
(48%)

8
(38%)

2
(67%)

4
(80%)

Limited access to computers may
preclude some students from taking
the GED exam

12
(41%)

7
(33%)

1
(33%)

4
(80%)

Cost of the 2014 GED exam to the
student

10
(34%)

6
(29%)

1
(33%)

3
(60%)

Security concerns about access to
the Internet for the GED exam

7
(24%)

4
(19%)

1
(33%)

2
(40%)

Other

6
(21%)

3
(14%)

1
(33%)

2
(40%)

No concerns

2
(7%)

2
(10%)

0
(0%)

0
(0%)

29

21

3

5

Total number of states responding

81

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Smaller states tended to express fewer concerns about the 2014 GED exam and computerbased testing than medium-sized or larger states. All of the medium-sized and large states that
answered this question were concerned about the cost of the new GED exam, about whether their
teachers would be prepared to teach the new exam, and the length of time it would take to
prepare students. In addition, most of the large states were concerned that teachers may not be
prepared to implement computer-based testing and that limited access may preclude some
students from taking the exam. These results suggest that states with larger prison populations
may encounter more challenges in terms of implementing the new GED exam and that smaller
states may fare better.
Outcome Indicators and Postrelease Measures of Success
Another area of interested that we asked state correctional education directors about is what
outcome indicators and measures of postrelease success for correctional education programs are
of value to both (1) assess student progress and attainment and (2) meet correctional goals of
increased safety within the institution and reductions in recidivism.
Table 4.16 summarizes which outcome indicators states’ correctional education systems
track. A majority of states (40) track GED certificates and nationally or industry-recognized
certificates earned (36 states). Thirty-two states also tracked gains in reading or math skills, and
about half of states tracked academic program completions. College credits earned and degrees
were tracked by 17 and 18 states, respectively. Other outcome indicators tracked by states’
correctional education systems included reading level performance, the Wide Range
Achievement Test (WRAT) scores, the Tests of Adult Basic Education (TABE), WorkKeys
certificates awarded, vocational training program completions, reductions in discipline, and state
and local vocational certificates.
Table 4.16. Outcome Indicators Tracked by States’ Correctional Education Systems
Indicator

Size of State
Overall

Small

Medium

Large

Indicators

N (%)

N (%)

N (%)

N (%)

GED certificates earned

40
(95%)

27
(96%)

8
(100%)

5
(83%)

National or industry-recognized
certificates awarded

36
(86%)

22
(79%)

8
(100%)

6
(100%)

Gains in reading or math skills

32
(76%)

20
(71%)

7
(88%)

5
(83%)

Academic program completions
(e.g., adult basic education, adult
secondary education, ESL)

23
(55%)

15
(54%)

5
(63%)

3
(50%)

High school degrees awarded

21
(50%)

15
(54%)

3
(38%)

3
(50%)

82

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

College credits earned

17
(40%)

10
(36%)

5
(63%)

2
(33%)

College degrees earned (e.g.,
Associate degrees)

18
(43%)

11
(39%)

5
(63%)

2
(33%)

Other

7
(17%)

4
(14%)

2
(25%)

1
(17%)

42

28

8

6

Total number of states responding

By size of state, tracking of GED certificates earned, nationally or industry-recognized
certificates, and gains in reading or math skills were reported by a number of the states in each
size category (Table 4.16). Medium-sized states were nearly twice as likely to report also
tracking college credits and college degrees earned, suggesting that perhaps college coursework
and contracting with community colleges to provide courses might be more prevalent in these
states.
We also asked state correctional education directors what postrelease indicators they
considered to be important outcome measures of academic or vocational education/CTE program
success. The majority indicated reductions in recidivism and postrelease employment as being
two important measures to track. Many states also cited enrollment in vocational training
programs and in postsecondary education/college courses. Less cited were postrelease indicators
of college attainment or degrees awarded. Other postrelease indicators mentioned included
Department of Labor statistics for their population and continued skill training, and one
respondent noted that all of the indicators listed in Table 4.17 are important but cannot be
tracked at this time.
Table 4.17. Postrelease Indicators States Consider to Be Important Outcome Measures for
Correctional Education
Size of State
Overall

Small

Medium

Large

Indicators

N (%)

N (%)

N (%)

N (%)

Recidivism

41
(98%)

27
(96%)

8
(100%)

6
(100%)

Post-release employment

38
(90%)

24
(86%)

8
(100%)

6
(100%)

Job retention

29
(69%)

18
(64%)

5
(63%)

6
(100%)

Enrollment in vocational training
programs

24
(57%)

15
(54%)

6
(75%)

3
(50%)

Enrollment in postsecondary
education/ college courses

22
(52%)

14
(50%)

6
(75%)

2
(33%)

College attainment

16
(38%)

10
(36%)

4
(50%)

2
(33%)

Degrees awarded

12

7

4

1

83

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Other
Total number of states responding

(29%)

(25%)

(50%)

(17%)

3
(7%)

3
(11%)

0
(0%)

0
(0%)

42

28

8

6

Medium-sized states were more likely to consider enrollment in vocational training programs
and in postsecondary education/college courses as well as college attainment and degree awarded
as being important outcome measures (Table 4.17). These results are consistent with the findings
in Table 4.16 that medium-sized states also were more likely than the small or large states to
track college credits and college degrees earned.
Participation in Federal, State, Local, and Private Grant Programs

In addition to funding from states, correctional education programs also can benefit from federal
funding such as from Title I, Part D, of the Elementary and Secondary Education Act of 1965
(ESEA) (Pub. L. 89-10) to be used to improve educational services for children and youth in
local and state institutions for neglected or delinquent children and youth.32 Further, federal grant
programs and foundation funding can be used to support specific programs or research efforts.
Twenty-seven state correctional education programs reported participation in ESEA, Title I,
Part D and the Workforce Investment Act (Pub. L. 105-220), Title II programs (Table 4.18).
Regardless of size, approximately two-thirds of states participated in the ESEA Title I, Part D
and Workforce Investment Act, Title II programs. Only nine states indicated that their
correctional education programs had received funding under the SCA and six states from
foundations such as the Sunshine Lady Foundation.33
Table 4.18. Federal, State, or Private Grant Programs States’ Correctional Education Systems
Participate in
Size of State

Type of Program

Overall
(n=42)

Small

Medium

Large

N (%)

N (%)

N (%)

N (%)

32

The purposes of Title I, Part D are to (1) improve educational services for children and youth in local and state
institutions for neglected or delinquent children and youth so that they have the opportunity to meet the same
challenging state academic content and State student achievement standards that all children in the State are
expected to meet; (2) provide these children with services to enable them to transition successfully from
institutionalization to further schooling or employment; and (3) prevent at-risk youth from dropping out of school as
well as to provide dropouts and children and youth returning from correctional facilities or institutions for neglected
or delinquent children and youth, with a support system to ensure their continued education (U.S. Department of
Education, 2006).

33

Other foundations and specific grant programs mentioned included the Bill and Melinda Gates Foundation, the
Open Society Foundations, the Perkins Leadership grant program, and one state’s Department of Labor career
technical grants.

84

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

ESEA, Title I, Part D

27
(64%)

17
(61%)

5
(63%)

5
(83%)

ESEA, Title II, Part A

4
(10%)

1
(4%)

1
(13%)

2
(33%)

Workforce Investment Act, Title II
(also known as the Adult Education
Family Literacy Act)

24
(57%)

16
(57%)

5
(63%)

3
(50%)

Federal Second Chance Act (SCA)
grants

9
(21%)

4
(14%)

2
(25%)

3
(50%)

Office of Juvenile Justice and
Delinquency Prevention (OJJDP)
grants

1
(2%)

1
(4%)

0
(0%)

0
(0%)

Bureau of Justice Assistance (BJA)
grant funding (other than Second
Chance Act)

3
(7%)

0
(0%)

1
(13%)

2
(33%)

National Institute of Justice (NIJ)
grants

1
(2%)

0
(0%)

0
(0%)

1
(17%)

Foundations (e.g. Sunshine Lady)
(please specify):

6
(14%)

4
(14%)

2
(25%)

0
(0%)

None/Don’t Know

4
(10%)

3
(11%)

1
(13%)

0
(0%)

42

28

8

6

Number of states responding

Of those states (n = 18) that received Workforce Investment Act, Title II dollars, the mean
amount received was $26,014,500 in FY2012 (Table 4.19). In terms of the Perkins Act funding,
states (n = 30) that received funding under this grant program received a mean amount of
$4,114,150. Eight states also reported receiving the states’ higher education/aid resources in
FY2012, with the mean amount being $1,306,031.
Table 4.20. Amount of Funding States’ Correctional Education Programs Received in 2012 from
the Workforce Investment Act, Perkins Act, and States’ Higher Educational/Aid Resources
Type of Program

Mean

Median

Workforce Investment Act, Title II
(also known as the Adult Education
Family Literacy Act) (n=18)

$26,014,500

$284,000

Carl D. Perkins Career and
Technical Education Act (Perkins
Act) (n=28)

$4,114,150

$69,000

States’ higher education/aid
resources (n=6)

$1,306,031

$596,125

85

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Discussion
Variation in Correctional Education Programming Across the States
In 2013, most states offered adult basic education, GED courses, and vocational skills training.
In addition, most states also reported having special education courses available. Higher-level
educational programming, including adult secondary education and postsecondary education, is
offered in about 70 percent of states (32 states and 33 states, respectively). However, we found
that smaller states were less likely to offer such courses, suggesting that inmates in smaller states
have fewer opportunities for adult secondary and postsecondary education. In 24 states,
participation in correctional education programs is mandatory for adult inmates without a high
school diploma or GED, and in 15 states mandatory for adults below a certain grade level.
However, smaller states were less likely than medium-size and large states to require mandatory
participation in correctional education programs. Smaller states though were more likely to
emphasize vocational education/CTE training for state prisoners than medium-sized or large
states.
An emerging trend is a growing emphasis on providing vocational education/CTE
programming that will lead to industry or nationally recognized certifications. For example, 28
states reported offering the National Center for Construction Education and Research
certification. Our survey suggests that more than half of reporting states offer certification
training in construction and in Microsoft Office skills. Occupational safety and plumbing and
electrical apprenticeships are offered in nearly half of reporting states, and welding is offered in
about a third of them.
Impact of the 2008 Recession
The effect of the 2008 recession was a 6 percent decrease on average in states’ correctional
education budgets between FYs 2009 and 2012. However, the effect of the recession differed by
size of state. The largest decrease in budgets was felt by medium-sized and large states. On
average, small states experienced a 2 percent increase in their state’s correctional education
budget, compared with a 20 percent and 10 percent decrease in medium and large states. Another
way to look at this is to calculate the dollars spent per student during this time period. Overall,
the mean dollars spent per student for correctional education was $3,479 in FY2009, compared
with $3,370 in FY2012—a 5 percent decrease on average in the dollars spent per student.
The reductions in states’ correctional education budgets reportedly led to a dramatic
contraction in the capacity of academic education programs, and to a reduction in the number of
students on average who participated in these programs. For academic programs, these budget
cuts and resulting cost-cutting measures yielded, on average, a 4 percent decrease in the mean
number of adult students enrolled in academic programs between Fiscal Years 2009 and 2012.
Medium-sized and large states on average experienced a larger decrease in the number of adult

86

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

students enrolled in academic programs (10 percent and 8 percent decrease, respectively), than
did small states who reported an average decrease of 1 percent in the number of students in these
programs.
The effect of the staffing and capacity cost-cutting measures on teachers was particularly felt
in medium-sized and large states. Overall, there was on average a 24 percent decrease in the
number of academic teachers who were employees, from an average of 110 in 2009 to an
average of 85 in 2012. All size states experienced a decrease in the number of teachers who were
employees, but the largest decrease occurred in medium-sized (44 percent) and large states (20
percent).
In addition, 20 states also reduced the number of course offerings for academic programs
during this time period; this was especially true in the larger states. In the short run, these cuts
saved states money by reducing the direct costs of correctional education programming,
However, in the long run they may have added to the future costs of reincarceration, given that
inmates are now returning to local communities having had fewer educational opportunities
while incarcerated. Long-term costs are important to bear in mind. Our meta-analysis results in
Chapter Two suggest that participation in correctional education programs is associated with a
13-percentage point reduction in recidivism, and that for every dollar spent on correctional
education programs, five dollars are saved in three-year reincarceration costs.
Vocational education/CTE programs seem to have fared somewhat better during the
recession than academic programs in terms of reductions in the number of students enrolled in
vocational training programs, and in the number of instructors. On average, there was a 1 percent
increase in the number of students enrolled in vocational/CTE programs between 2009 and 2012.
However, this appears to be largely driven by an increase on average of 7 percent in smaller
states. In comparison, the medium-sized and large states experienced a reduction on average of 4
percent and 11 percent, respectively, in the number of students enrolled in these programs. Small
and medium-sized states in fact saw a modest increase between FYs 2009 and 2012 in the mean
number of vocational education/CTE instructors who were employees (8 percent and 24 percent,
respectively). Combined, this suggests a modest expansion of vocational education/CTE
programs in small and medium-sized states during this time period. Still, 38 percent of small
states and 44 percent of medium-sized states reported that in response to budget cuts they had
reduced the number of course offerings for vocational education/CTE programs.
Use of Information Technology
One of the major trends that will shape the future of work in the 21st century is the growing role
of information technology in our society, with technological change resulting in an increased
demand for a skilled workforce (Karoly, 2013). In today’s job market, basic computer skills are
virtually a necessity in searching for job opportunities, applying online for jobs or benefits, and
undertaking simple clerical tasks in the workplace. The importance of computing skills for
today’s job market is recognized by state correctional education directors and reflected by the
87

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

fact that 24 states reported offering a Microsoft Office certification as part of their vocational
education/CTE programs.
Further, distance learning and online instruction are growing trends in the United States, with
increasingly more educational courses being offered online either by colleges or virtual high
schools. These online courses are appealing in that they offer an opportunity to address key
barriers that correctional educators face in terms of limited classroom space and the need to scale
back on instructional staff in recent years. In addition, the frequent movement of inmates from
facility to facility makes it difficult to ensure continuity of coursework and learning
opportunities, while distances between facilities (especially in rural states) make it difficult to
provide instruction in all facilities. Computer-assisted instruction is also appealing in offering the
opportunity to tailor instruction and coursework to the needs of the individual student.
Yet, our survey results indicate that the role of computer technology in correctional education
is complicated. We found that the use of computers for instructional purposes is common, with
39 states reporting the use of desktop computers (either standalone or networked) and 17 states
reporting the use of laptops. However, access to the Internet, and the use of Internet-based
instruction (one-way or interactive), is reported to be limited in most states’ correctional
facilities. Thirty states reported that only teachers and instructors have access to live Internet
technology. In 26 states, inmate students lack access to any Internet technology, and in only 16
states do inmate students have access to simulated Internet programs. In focus group discussions,
state correctional education directors cited corrections’ opposition to access to computer or to the
Internet as a key barrier to using technology in the classroom. In terms of instructional methods
that use some type of technology, only ten states reported that they had closed-circuit television,
and only a few states reported using it to provide one-way or interactive video/satellite
instruction.
Readiness for the 2014 GED Exam and Computer-Based Testing
The GED is the predominant way that inmates earn their high school equivalency diplomas, and
GED completion is often a prerequisite for many vocational training programs (Harlow, 2003;
Lockwood et al., 2013).
The new 2014 GED exam and the move to computer-based testing will further push
correctional education systems to use information technology in the classroom and to find
solutions to some of these barriers. Of the 31 states planning to implement the 2014 GED exam,
17 plan to use a combination of computer workstations and laptops for inmates to take the exam.
The 2014 GED exam not only represents a more rigorous test, being aligned with the
Common Core State Standards (CSS), but also will rely on a new test delivery model—namely,
computer-based testing to replace the old paper-and-pencil exam (Lockwood et al., 2013). This
represents a profound change to states and at the same time presents some key challenges. GED
completion rates are seen as important outcome indicator to track by 95 percent of states that
took part in our survey. Of the 31 states planning to implement the 2014 GED exam, 14 states
88

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

expected that the more rigorous GED exam and the use of computer-based testing may have a
negative effect on the number of adult inmates who will be prepared to take the new exam, and
16 states expected a negative effect on GED completion rates. This was particularly true for the
medium-sized and large states. Nineteen states were concerned about their teachers being
adequately prepared to teach the new exam, and 24 states were concerned about the length of
time it may take to prepare students for the more rigorous exam.
In recent discussions with state correctional education directors at a 2013 Correctional
Education Association conference and workshop we facilitated a discussion on preparations for
the 2014 GED exam. One of the issues the state directors debated was how to assess whether an
inmate student had sufficient computer skills to take the timed exam. Anecdotal reports from
state correctional education directors with early experience with computer-based testing were
that some inmate students did not have adequate computer skills to finish the test within the
allocated amount of time. In addition to keyboarding tasks, the new GED exam and computerbased testing require a range of computing skills, such as knowledge of how to access tool bars,
navigate “HOT SPOTS,” use “drag and drop” and “point and click” skills, and use a drop-down
online calculator (Lockwood et al., 2013). The directors discussed possible workarounds to help
students, including the use of standalone calculators and having students practice writing in longhand their essays before typing their answers on the computer. The directors also mentioned
including as part of the GED preparation time in the computer lab for students. In our survey, 12
states reported concerns that limited access to computers may preclude some students from
taking the new GED exam. Also, responding directors in 14 states reported concerns that their
teachers may not be adequately prepared to implement computer-based testing.
Only two states reported no concerns about the new exam or computer-based testing. In
general, smaller states expressed fewer concerns; however, our survey results suggest that states
with the majority of the prison population (i.e., medium-sized and large states) expect to
encounter a number of challenges in implementing the new exam and test delivery system.
Given these concerns, the survey results suggest that the United States may experience a
dramatic drop in the number of GED completion rates for incarcerated adults, which will merit
close monitoring and an assessment of the long-term implications for this population in terms of
effects on their opportunities to participate in vocational training programs and postsecondary
education, as well as effect on employment opportunities. These results also suggest that states
may need technical assistance in preparing teachers and students for the new GED exam. The
fact that not all states will be using the GED exam as a high school equivalency test raises
questions about whether the use of alternative exams will be accepted by vocational training
programs and college programs.

89

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Postsecondary Education
As noted earlier, the history of postsecondary education for incarcerated adults is one of an initial
growth in the number of programs and then a significant reduction in response to the elimination
of Pell grants in the 1990s for this population.
Our survey results provide updated information about these trends. Our survey did not ask
about the number of inmates in postsecondary courses but does provide information on the
degree to which states offer them and how inmate students are paying for these courses. We
found that in 2013, 32 states reported offering postsecondary education or college courses to
adult inmates (especially true of medium-sized and larger states). However, these courses today
are primarily paid for by the individual inmate or by family finances. In 16 states, state funding
from the department of corrections, for example, is used to cover the costs of postsecondary
education. Only 12 states reported using college or university funds to pay for these courses. Our
survey results suggest that reinstatement of the Pell grants for this population may have a
substantial effect in expanding postsecondary opportunities for state prisoners.
In recent years, there has been a growing interest in providing postsecondary education to
inmates in state prison. Such programs as the Bard College Initiative and the Prison University
projects are two examples. Importantly, a group of foundations recently joined together to fund a
demonstration project in three states called Pathways from Prison to Postsecondary Education
led by the Vera Institute of Justice to support postsecondary education and degree attainment for
individuals who are within two years of release. Of particular note is that these various initiatives
are focused on degree attainment, whereas traditionally courses offered within prisons often were
not aimed at credential attainment or building a core of courses that would allow individuals to
continue and ultimately, to obtain a postsecondary education degree either while incarcerated or
upon release from prison.

90

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

5. Conclusions and Recommendations

Introduction
The key finding from this comprehensive study of correctional education in the United States is
that correctional education is effective in reducing recidivism for incarcerated adults and that
there is reasonable evidence that it is also effective, especially vocational training, in improving
individuals’ likelihood of postrelease employment. Our cost analysis further showed that
correctional education is highly cost-effective for incarcerated adults—for every dollar spent on
correctional education, five dollars are saved on three-year reincarceration costs. Our report also
provides the most comprehensive systematic review we are aware of on what works in
correctional education for incarcerated juveniles. For example, we found compelling—if still
preliminary—evidence for Scholastic’s computer-enhanced reading intervention, Read 180, and
for the highly intensive and personalized education model exemplified by Florida’s Avon Park
Youth Academy.
Thus, the debate should no longer be about whether correctional education is effective or
cost-effective; rather, the debate should focus on where the gaps in our knowledge are and
opportunities to move the field forward.
In this chapter, we offer some recommendations and next steps, drawn from our evaluation
results; while this report is to the U.S. Attorney General, these recommendations will also be of
interest to other federal departments and agencies focused on reentry and are intended to provide
a roadmap for building on the gains made to-date in educating incarcerated individuals to
improve their chances of success upon release and reentry into local communities.

Correctional Education for Adults
Our survey results provide solid evidence about the dramatic impact the 2008 recession had on
correctional education in the United States. Specifically, the results show that as budgets were
reduced, the reported capacity for academic programs contracted, which led to a corresponding
drop in the number of incarcerated adults participating in these programs and in the number of
teachers who were employees. In the long run, such a lack of educational opportunities may
contribute to future reincarceration trends and future incarceration costs. This raises the question
of whether the trade-offs we are making in terms of cost savings today with reductions in
educational programming are worthwhile considering the future costs of reincarceration as well
as the effect that such lost opportunities have on individuals’ chances of finding employment and
being successful in reintegrating back into society.

91

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Throughout this project, at various conferences, workshops, and as part of individual
discussions with state correctional education directors, these directors have repeatedly said that
their legislature or department of corrections is asking them for evidence about how effective
their programs are to inform budget decisions and that they now are providing correctional
education programming with fewer dollars. The directors strongly desired information on how
they might modify their models of education to trim their budgets while still maintaining the
effectiveness of their programs. The results of our meta-analysis answers the first question about
effectiveness—correctional education programs are dramatically effective in reducing
recidivism, and there is modest evidence of improvements postrelease employment outcomes.
Our findings also clearly indicate that correctional education programs are highly cost-effective
for incarcerated adults.
However, because of limitations in quality of the evidence base (as discussed further below),
we cannot answer the other critical questions needed to inform discussions about modifications
to educational programming in a resource-constrained environment. We note, as did MacKenzie
(2008), that we are unable to get at what is inside the “black box” of what works in correctional
education, to answer such questions as:
•
•
•
•

What dosage is associated with effective programs, and how does it vary for different
types of academic programs and students?
What models of instruction and curriculum delivery (e.g. one-on-one, traditional
classroom lectures, computer-based learning) are most effective in a correctional
environment?
Who benefits most from different types of correctional education programs?
What principles from adult education and learning may be applicable to correctional
education?

Thus, we recommend focusing research and evaluation efforts at the federal and state levels
to address these questions so that policymakers and state correctional education directors
can make informed trade-offs in budget discussions. Where feasible, researchers should be
encouraged to make as much use of administrative data as possible to help reduce evaluation
costs.
Apart from this limitation, our survey results underscore that how correctional education is
being provided today is very different from how it was provided when many of the studies in the
meta-analysis were undertaken. This includes different models of instruction and delivery,
reductions in the number of teachers who are employees, the increased use of peer tutors, and the
growing role of computer technology in the classroom and in instruction. Thus, a program
provided ten years ago may be operating today in a different context altogether and under a
different set of budget constraints. Thus, moving forward, we recommend that federal and
state governments and philanthropy fund (1) evaluations of programs that illustrate
different educational instructional models with the goal of getting inside the black box, (2)
evaluations of programs that are trying innovative strategies to implement technology and
92

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

leverage distance learning in the classroom, and (3) an analysis of what lessons from the
larger literature on adult education may be applied to correctional education.
The new 2014 GED exam, which requires implementing computer-based testing, represent a
profound change for the field of correctional education. The GED certificate continues to be an
important mechanism by which many inmates earn their high school equivalency and is a key
outcome indicator tracked by departments of corrections. Yet, because the updated exam is more
rigorous than its predecessor and because of the new requirement of computer-based testing, the
majority of state correctional education directors expect to see a negative impact of the new GED
exam on completion rates and on the number of inmates prepared to take the new exam.
These directors also have expressed concern that lower GED completion rates will hurt
educational and recidivism outcomes more broadly. However, existing research suggests that this
concern may be overstated because it appears that it is the skills inmates acquire while preparing
for the GED, more than the credential itself, that reduces their postrelease recidivism (Tyler and
Kling, 2007), and this finding is corroborated by broader evidence that the GED’s effect as a
signal of worker quality is quite limited (Heckman and Rubenstein, 2001; Tyler, Murnane, and
Willett, 2000). Consistent with that conclusion, our own meta-analysis also found a positive
impact of GED preparation, though it was not possible to disentangle preparation from
completion in some of the less-rigorous studies (Davis et al., 2013). As such, it is possible that a
more-rigorous GED will actually improve the long-term outcomes of inmates who pursue it.
What is clear is that well-designed research is needed to estimate and document the impact of the
new GED on inmates’ educational skills, attainment, employment, and recidivism, as well as the
implementation challenges it imposes on the correctional facilities themselves. We recommend
that the federal government monitor and evaluate the impact of the new GED and
computer-based testing on the field and consider opportunities to provide technical
assistance to states and training to help prepare educators to teach the more rigorous GED
exam and to implement computer-based testing.
The role of computer technology in correctional education is a growing trend, and the new
computer-based testing requirement for GED exam administration is likely to accelerate the
adoption of computer technology in correctional settings. Given these changes, it will be
important to document how correctional settings overcome security and resource challenges to
computer-based testing and how they maintain their technology infrastructure in resourceconstrained environments. These lessons are important as computer-enhanced instruction
becomes increasingly commonplace in the broader secondary and postsecondary educational
landscape nationally. With the rise of blended learning technologies and massively open online
courses (MOOCs), the question is not whether computers should play a substantial role in
educating incarcerated adults, but how best to facilitate their adoption and use. Further, educators
need assistance in measuring readiness for the GED exam including computer literacy, as well as
assistance in adopting computer-aided instruction and incorporating online courses into the
correctional education curriculum. In addition, there is a need for in-depth case studies and
93

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

evaluation of innovative examples of the use of computer technology in the classroom to aid in
identifying exemplary practices. An analysis of the larger literature on the use of computer
technology in adult education may be informative here as well. Thus, we recommend further
evaluation and research on the use of computer technology in the correctional education
setting to help answer such questions.
States are increasingly offering nationally and industry-recognized certificates, which is a
positive trend as corrections focuses increasingly on training programs that will lead to
meaningful credentials and enable individuals to earn a living wage. However, it is not yet clear
to the degree to which these certificates will enhance the post-employment prospects of those
leaving prison, given the historically difficult time former inmates have getting hired in jobs that
provide a living wage—particularly in the sub-baccalaureate labor market. We need to assess the
effectiveness of these programs, and the credentials they provide, in helping returning
individuals find and sustain employment and to assess the degree to which existing barriers to
employment persist that may dampen the effects of having these changes in vocational training
programs. Given the changes in the U.S. economy and the 21st century workforce needs, we
recommend an assessment at the federal and state levels about what such changes mean for
the criminal justice–involved population and that a summit at the state and federal levels
with private industry be supported to explore what opportunities are available to formerly
incarcerated individuals and what skills will be needed in the future.
Finally, when we began this study, we conducted a wide search to identify what other
surveys had been conducted on this topic. We found very little information available, and what
was available tended to be out-of-date and limited in scope. The nationwide survey we
conducted of state correctional education directors can serve as a baseline moving forward.
Repeating a nationwide survey of correctional education annually or biennially would
enable the field and policymakers to assess progress in specific areas and the impact of
different policies.

Correctional Education for Juveniles
For juveniles, a key question is how best to provide services that will lower young offenders’
risk of future crime and increase their chance of success in the legitimate economy. In educating
juveniles, correctional facilities must serve a highly transient population of students who bring a
widely varied set of educational and emotional needs (Sedlak and McPherson, 2010; Meisel et
al., 1998; Leone, Meisel, and Drakeford, 2002). Further, youth with learning disabilities tend to
be overrepresented in juvenile correctional facilities (Meisel, Henderson, Cohen and Leone,
1998). The literature in this area reflects the reality of what correctional education looks like for
juveniles in the United States.
We focused our systematic review on education provided to juveniles in institutional settings.
Overall, the 18 studies in our systematic review can be generally characterized as small-to-mid94

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

scale randomized trials or as large observational studies with minimum-to-moderate use of
statistical methods to adjust for unobserved differences. We found that the methods employed in
the studies on juvenile correctional education varied markedly by intervention type. For example,
studies of the packaged reading interventions were generally fairly small, because these studies
involve administering particular curricula at the classroom or student level, as well as
administering pre- and post-tests to individual students. The designs of these studies were fairly
robust, but the small sample sizes of the studies and limited power for hypothesis testing makes
it difficult to generalize broadly from their findings. This suggests that the field is ripe for largerscale randomized trials.
The field is also ripe for rigorous evaluations of natural experiments such as Aizer and
Doyle’s (2013) study of the effects of juvenile incarceration using naturally occurring random
assignment to harsh judges. Studies that take advantage of rigorous causal methods in juvenile
settings can shed much-needed light on what works in these settings. Several of the smaller
randomized trials we include here have noted the difficulties of high student turnover in
correctional facilities, and of simply gaining permission to undertake research in these facilities
(Shippen et al., 2012; Calderone et al., 2009). As such, we recommend that the focus be on
implementing larger-scale randomized trials and rigorous evaluations of natural
experiments. Such research efforts will clearly take time to develop and execute. They will
ideally be realized through long-term partnerships between researchers and correctional
facilities. Informed by such partnerships, facilities can make increasingly evidence-based
decisions that not only improve their students’ prospects but also reduce the social incidence of
crime and delinquency.
Taken in conjunction with the broader research literature on each of the interventions
examined, we did identify two interventions that show particular promise: a blended learning
reading curriculum by Scholastic called Read 180, which combines teacher-directed instruction
with computer-enhanced, self-paced instruction, and the Avon Park Youth Academy in Florida,
which is a highly intensive program that includes personalized academic instruction and
consistent mentoring during and after incarceration by the same parole officer (who is given a
markedly reduced caseload). Beyond these stronger studies, we found positive effects from very
small studies of Corrective Reading and Tune in to Reading, but we think it is premature to
generalize from such small samples.
Finally, the benefits of earning a GED while incarcerated, though estimated as positive in the
systematic review, remain especially unclear, since these studies’ comparisons of students who
earned a GED with those who did not are particularly vulnerable to selection bias at the student
level. Further, as noted above, the most rigorous research from the literature on incarcerated
adults suggests that it is the education acquired in GED programs rather than the GED credential
itself that confers the greatest postrelease benefits (Tyler and Kling, 2007; Davis et al., 2013).

95

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Improving the Evidence Base for Adult and Juvenile Correctional Education
In our meta-analytic report (Davis et al., 2013), we laid out a number of recommendations to
improve the evidence base and they merit summarizing here. The questions we would like to
have answered were not feasible because of limitations in the quality of the evidence base and
the unevenness of the research designs used to assess the evidence and identify promising
practices. There are four things that we recommend that the federal and state governments
and philanthropy invest in to help further develop the evidence base for correctional
education.
Apply Stronger Research Designs
Establishing a causal relationship between correctional education participation and successful
outcomes for inmates requires ruling out the possibility of selection bias. This form of bias
occurs when inmates who elect to participate in educational programs differ in unmeasured ways
from inmates who elect not to participate in educational programs. Isolating the effects that can
be directly attributable to a program is crucial in supporting the design of effective policies—an
objective hampered by studies with research designs that are highly susceptible to selection bias.
In our meta-analysis, only seven of the 50 studies used to assess recidivism and one of the 18
studies used to assess employment were based on studies with high-quality research designs.
Further, many studies did not report sufficient information about the socio-demographic
characteristics and other characteristics of the treatment and comparison groups; reporting on
such information would allow researchers to assess meaningful differences between the two
groups to be evaluated and to quantify the potential threat of selection bias. To minimize this
potential for bias, future studies should ideally use such research designs as randomized
controlled trials and well-executed quasi-experimental designs.
In addition, identifying the appropriate comparison groups is important. Many of the studies
reviewed in our meta-analyses used comparison groups of nonprogram participants but did not
consider differences in terms of levels of education, certification, or training. Thus, the
comparison group might be a mixture of inmates with varying levels of academic achievement.
Gaes (2008) recommended that a study registry be established to help sort out the different
effect sizes found across studies. The vast array of programs currently administered and the
dearth of basic information on their design and their effectiveness in a centralized system
precludes the effective utilization of resources, particularly for states making strategic decisions
on whether and how to recalibrate their programs to adjust to changes in funding and changes in
the prisoner population. Funding of such a registry by the federal government to be operated by a
university or research organization would help advance the evidence base by including details
about each study, including information about the program and intervention, about the evaluation
design, about characteristics of the treatment and comparison groups, and about the outcomes
measures used. Such a registry could also provide technical assistance and evaluation guidance
96

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

for those working in the field. Throughout the course of the project, we have received repeated
requests from correctional educators and researchers for this type of information—clearly the
field sees a real need for such a registry.
Measure Program Dosage
Many practitioners have posed the question: What dosage level is associated with effective
correctional education programs? For instance, does it matter that an individual participates in 20
hours of academic instruction, or is 30 hours of academic instruction required for a given course?
Such questions about dosage levels are especially salient now, when many correctional education
programs have experienced significant budget cuts.
On average, the studies we reviewed lacked specific information about the dosage of the
program, such as the overall program duration, the number and grade level of the courses in
which inmates were enrolled, how many hours per day or week inmates were exposed to formal
class instruction, and how many hours per day or week inmates worked on assignments outside
the classroom. In many of the studies, particularly those that were secondary analyses of
administrative data sets, respondents were categorized simply as correctional education
participants and nonparticipants. This crude categorization undoubtedly masked variation in
exposure to the program among participants. For example, some inmates may have been enrolled
for a year, while others inmates may have been enrolled for a week and withdrawn.
Without being able to discern such differences, it is difficult to put the findings from
individual studies in their proper contexts. The lack of dosage information means that there is
little to no empirical evidence that can help inform policymakers on “how much” correctional
education is necessary to produce a change in the desired outcomes. In future studies, the proper
recording of program dosage when collecting data and monitoring the progress of inmates
through correctional programs will be critical to enable researchers to examine these questions.
Identify Program Characteristics
When we undertook our review of the literature on academic and vocational training programs
for incarcerated adults, our charge from BJA was to identify promising or evidence-based
programs that could be potentially replicated in other settings. We were unable to identify
specific exemplary programs—not because such programs do not exist, but because the evidence
base does not provide sufficient detailed information about such programs to allow us to do so.
Many of the studies in the literature review did not provide sufficient detail on the characteristics
of the program, such as the structure of the curriculum, the training and certifications of the
teachers, the instructional methods used by the teachers, the student-teacher ratio in classrooms,
and supplemental access to textbooks and technology. To the extent possible, we culled this
information from the studies that provided it and used it in an exploratory fashion in our metaanalyses. However, few studies consistently listed these details in their program descriptions;
consequently, our findings from these few studies are suggestive at best. Thus, from a meta97

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

analytic approach, we are unable to offer evidence-based prescriptions about what aspects of
correctional education are most or least effective. The field would be well served if future
research carefully documented the characteristics of the programs so that different models of
program organization and instruction could be empirically validated.
Examine More-Proximal Indicators of Program Efficacy
More research is needed on more-proximal measures that would better indicate how programs
actually affect thinking and behavior, such as changes in motivation, literacy gains, development
of concrete skills, or academic progress versus academic achievement. Overwhelmingly, the
research conducted to date has looked at recidivism as the major outcome indicator, which is
understandable given its importance as a marker of successful prisoner rehabilitation. However,
despite its salience in criminological research, the emphasis on recidivism has meant that we
know much less about the process through which correctional education helps shape how former
inmates re-integrate into the community. Correctional education is believed to improve the skills
and abilities of inmates (i.e., “human capital” in economics parlance), which, in turn, improves
their chances of continuing education/training upon release and then finding gainful
employment. Only four studies in our review looked at skills and abilities (as measured by
achievement test scores), and only 18 looked at employment. There were too few studies of
additional education/training to include in a meta-analysis. Applying these more-proximal
indicators of program efficacy will help to better elucidate the mechanisms that undergird the
role of education in the rehabilitation process.
In summary, to improve the evidence base, state and federal policymakers and
foundations should invest in well-designed evaluations of correctional education programs.
Also, researchers and program evaluators need to strive to implement rigorous research designs
to examine questions related to potential bias and program dosage. Funding grants and
guidelines can help further the field by requiring the use of more-rigorous research
designs. Funding mechanisms should also support partnerships between correctional educators
and researchers and evaluators to undertake rigorous and comprehensive evaluations of their
programs. A study registry of correctional education evaluations would further aid in
developing the evidence base in this field to help inform policy and programmatic
decisionmaking. Given that we know that these programs are cost-effective, if these programs
were refined based on this important missing information, correctional education could yield
even greater returns on investment.

Implications of Broader Trends in Corrections for Correctional Education
Several trends occurring in the field of adult and juvenile corrections have important
implications that merit further consideration. First, a key trend in corrections is efforts by states
to reduce the size of their state prison population, through a variety of means. This includes such
98

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

“front-end” strategies as reducing prison admissions, diverting offenders to county- rather than
state-level institutions, or changing felonies to misdemeanors. This approach is being tried in
many states, particularly with respect to drug offenders. Delaware, for example, repealed
mandatory minimums for certain drug offenses in 2007. Colorado modified penalties for certain
drug possession offenses in 2010. New York’s Rockefeller Drug Laws were changed to
eliminate mandatory minimums for certain first- and second-offense offenders (Division of
Criminal Justice Services, 2010). And Indiana proposed a sentencing reform plan to give judges
more leeway to sentence lesser felon to community corrections or treatment programs
(Associated Press, December 15, 2010).
In addition, states have implemented strategies focused on the “back end” of the system, such
as reducing sentence lengths though earned credits or good time and revocations for probationers
and parolees. For example, in April 2011, California Assembly Bill 109 shifted prisoner and
parolee responsibility to the counties to close the revolving door for low-level offenders because
of high parolee revocation rates. California’s Public Safety Realignment Plan, which went into
effect October 1, 2011, fundamentally changed the state’s criminal justice system. Under
Realignment, nonserious, nonviolent, and nonsex offenders no longer serve time in state prison,
nor are they supervised by state parole when released (CDCR, 2011). Instead, local counties are
now responsible for managing, housing, supervising, and rehabilitating these low-level
offenders. Many states also are reducing prison populations though accelerated release
mechanisms. Media reports contained in Crime and Justice News reports compiled by Ted Gest
revealed that more than 30 states are either in the planning stages or have implemented policies
for early release, some targeting large segments of the prison population, and others more narrow
segments, such as the terminally ill. All these changes in the correctional landscape have
implications for how we think about how to provide academic and vocational education/CTE to
incarcerated adults. For example, the movement in some states to have low-level offenders serve
their time in county jails versus state prisons has implications for how we think about providing
academic and vocational training to incarcerated adults at the local level. It raises policy
questions: Are there differences in access to academic and vocational education/CTE programs
depending on the setting where one services one’s sentence? Are there differences in education
and employment outcomes as a result?
Second, in the area of juvenile corrections, a related long-term trend has been to keep youth
in the community if at all possible instead of placing them in correctional institutions, and, when
they are incarcerated, to house them in local versus state facilities. A sharp decline in the juvenile
incarceration levels in the United States may partially reflect this trend. For example, the number
of juveniles detained, diverted, or committed on any given day in the United States declined
from 105,000 to 61,000 between 1997 and 2011. This suggests that the current emphasis is on
community-based educational services for juveniles who become involved in the criminal justice
system, such placement in nonresidential alternative schools. Given evidence that incarceration

99

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

itself reduces juvenile offenders’ educational attainment and increases their recidivism relative to
less-restrictive sentences (Aizer and Doyle, 2013), this is a promising development.
In addition, an important federal initiative to address the school-to-prison pipeline and reduce
at the front-end the chances of youth becoming involved with the justice system was launched in
2011. The U.S. Departments of Justice and Education announced the joint Supportive School
Discipline Initiative (SSDI) aimed at addressing the disciplinary policies and practices that can
push students out of school and into the justice system. As part of the SSDI, the U.S.
Departments of Justice and Education recently released a school discipline guidance package to
assist states, districts, and schools in developing practices and strategies to enhance school
climate, and ensure that those policies and practices comply with federal law (U.S. Department
of Education, 2014).
Our systematic review focused on what works with incarcerated youth in part because the
broader literature on educational interventions for juvenile offenders outside of correctional
facilities is even more nebulous. An important direction for future research is to identify
interventions that improve juveniles’ educational, employment, and recidivism outcomes in lessrestrictive settings, such as alternative schools or traditional schools. To facilitate such studies on
a large scale, it would of course be useful for longitudinal educational data systems to include
indicators of students’ involvement in the criminal justice systems. However, we recognize that
the inclusion of such indicators may raise both logistical and privacy concerns. Therefore, any
such indicators would likely need to be accompanied by rules governing their use (e.g., only for
program evaluation in de-identified datasets). Without such indicators, it is difficult to identify
juvenile offenders in larger educational data systems and thus, to conduct large-scale analyses of
what works for those populations outside of correctional facilities.
The growing policy emphasis on community-based schooling for juvenile offenders also has
implications for students’ transitions between correctional and noncorrectional settings. In our
discussions with juvenile correctional education directors, they identified these transitions as
important challenges in terms of transferring academic records and maintaining curricular
consistency. The extent to which these challenges are mitigated by placing offenders in
nonresidential alternative schools instead of correctional facilities is unclear, as are other best
practices for facilitating smooth transitions.
To guide policy improvements, stronger federal reporting requirements about local
correctional education practices could help facilitate improved state and local comparisons of
program effects. We currently know less at the federal level about education programs for
juvenile offenders than about education for the larger K–12 population. Although some
correctional education programs are included in the U.S. Department of Education’s Common
Core of Data, inclusion is variable, and these programs are often difficult to isolate in federal
data. Moreover, such data provide little information about local policies on incarceration versus
alternative placements and on standard sentence lengths, staffing policies, technology
infrastructure, and instructional programs offered. A central repository of such information,
100

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

whether collected federally or privately, would provide a valuable tool to policymakers and
researchers alike.
We recommend that policymakers seek to assess and understand the implications of
these trends in the field of corrections with respect to their impact on correctional
education.

Concluding Thoughts
There are more than 2 million incarcerated adults in the United States—more than any
industrialized nation. This study has demonstrated that education programs can help adults get
back on their feet upon release from prison and may help youth involved with the juvenile justice
system to improve their education and employment prospects. Moreover, our meta-analysis of
the literature on incarcerated adults suggests that correctional education programs are highly
cost-effective in helping to reduce recidivism and improve postrelease employment outcomes.
States will continue to operate in a reduced funding environment for at least the near future. The
findings and recommendations we have laid out here are intended to ensure that, moving
forward, we understand how to best to deliver education and vocational training to assist in
achieving positive reentry outcomes.

101

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Appendix A. Summary of Studies for the Juvenile Correctional
Education Review

In this appendix, we provide a detailed summary of the studies, sample sizes, and effect sizes
reported in the systematic review in Chapter Three. In addition, each study was rated for rigor on
the Maryland Scientific Methods Scale. Eighteen of the 27 studies that underwent scientific
review were deemed eligible for formal inclusion in the analysis. Shaded rows indicate studies
that were ineligible for systematic review (n = 9) due to design but that inform the research
context.

102

Treatment
Condition

Drakeford,
2002

Corrective
Reading

Corrective Reading
AllenCorrective
DeBoer et
Reading
al., 2006

Citation

Traditional
language
arts
instruction

Traditional
language
arts
instruction

Comparis
on
Condition

Oak Hill
Academ
y in
Marylan
d

Mental
health
treatme
nt unit
within a
juvenile
correcti
onal
facility

Setting

103

Age: 16–18;
100% Male;
75% African
American;
25% White;
100% with
learning
disabilities;
Baseline
grade
equivalent:
4th-5th grade
Age: 12–21
(mean: 17);
100% Male;
100% African
American;
100% with
history of
educational
disabilities

Demographi
cs

6

4

n
Tre
at

0

0

n
Co
mp
are

One hour,
3 times a
week, for
8 weeks
(20
lessons on
average)

30
minutes a
day, 5
days a
week, for
9 weeks
(30
lessons on
average)

Duration
and
Frequenc
y

Words Read
Correctly per
Minute
(WPM)

Words Read
Correctly per
Minute
(WPM) and
Word Errors
per minute
(WE)

Outcome 1
Metric

Mean gain:
9.2 WPM

Mean gain:
35.8 WPM;
No evidence
of WE effect

Outcome 1
Effect Size
Estimate

Outcome 2
(and 3)
Metrics

Outcome
2 (and 3)
Effect Size
Estimate

Table A.1. Summary of Studies, Samples, and Effects in the Systematic Review
(Shaded rows indicate studies that were ineligible for systematic review due to design but that inform the research context.)

5*

5*

MD
Scal
e

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Treatment
Condition

Corrective
Reading:
4:1
student:
teacher
ratio

Citation

Houchins
et al, 2008

Corrective
Reading:
12: 1
student:
teacher
ratio

Comparis
on
Condition
Longterm
juvenile
correcti
on
facility
in a
MidAtlantic
State

Setting

104

Age: 13–17
(mean: 16.5);
100% Male;
64% African
American;
18%
Hispanic;
18% White;
21% with
learning
disabilities or
mental
retardation;
58% with
emotional or
behavioral
disabilities

Demographi
cs
10

n
Tre
at

n
Co
mp
are
10

Duration
and
Frequenc
y
1 hour, 3
times a
week, for
7 weeks
(21
sessions)
Woodcock
Reading
Mastery Test
, Revised
(WRMT-R):
Word
Identification
(WI) and
Word Attack
(WA); Gray
Silent
Reading
Test
(GRST);
Dynamic
Indicators of
Basic Early
Literacy
Skills Oral
Reading
Fluency
(DORF)

Outcome 1
Metric
Relative
gains:
WI: 0.60 SD
(reported
p<.01, but
p=.058 if
adjusted for
multiple
comparisons
)
WA: 0.50 SD
GSRT: 0.72
SD
DORF3:
0.07 SD
DORF4:
-0.21 SD
DORF5:
-0.46 SD
(none with a
p<.05;
mean: 0.205
SD, p=0.65)

Outcome 1
Effect Size
Estimate

Outcome 2
(and 3)
Metrics

Outcome
2 (and 3)
Effect Size
Estimate
5

MD
Scal
e

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Treatment
Condition

Corrective
Reading
for (180
minutes
per week)

One-toone
tutoring
using
direct
Instruction
and
Corrective
Reading

Citation

Scarlato
and
Asahara,
2004

Coulter,
2004†

None

Reading
Specialist
for 60
minutes
twice a
week, plus
225
minutes of
additional
reading
instruction
(345
minutes
weekly)

Comparis
on
Condition

State
juvenile
detentio
n facility
in
Souther
n
Colorad
o

Residen
tial
juvenile
treatme
nt
facility

Setting

105

Mean=15.5;
83% Male;
33% African
American;
33%
Hispanic;
33% White;
83% with
disabilities;
42% with
emotional
disturbance;
8% with
mental
retardation;
IQ range: 55–
89

Age: 16–17;
100% Male;
100% with
learning
disabilities or
emotional
disturbance;
100% read
significantly
below grade
level

Demographi
cs

12

5

n
Tre
at

0

n
Co
mp
are
4

5 days a
week for 9
weeks
(mean=21
sessions,
range=5
to 48
sessions;
session
length not
given)

Duration
and
Frequenc
y
45
minutes, 4
times a
week for
19 weeks
Woodcock
Reading
Mastery
TestRevised
(WRMT-R):
Word
Identification
(WI), Word
Attack (WA),
Word
Comprehend (WC)
and
Passage
Comprehend (PC);
Total
Reading
(TR)
Gray Oral
Reading
rd
Test, 3
Edition
(Passage
[combines
Rate and
Accuracy]
and
Comprehens
ion)
Also, words
read
correctly per
minute
(WPM)

Outcome 1
Metric

Passage: 9
months of
gain for 1
month of
teaching;
Comprehens
ion: 9 month
gain for 1
month of
teaching.
3.57 correct
WPM per
week gain,
versus a 1
WPM
expected
gain

Relative
gains:
WI: 0.84 SD
WA: 0.30 SD
WC: 0.32
SD PC: 0.89
SD TR: 0.95
SD (none
significant)
(mean: 0.66
SD, p=0.36)

Outcome 1
Effect Size
Estimate

Outcome 2
(and 3)
Metrics

Outcome
2 (and 3)
Effect Size
Estimate

1

3

MD
Scal
e

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Corrective
Reading
Plus
Whole
Language
Instruction

Malmgren
and
Leone,
2000†

None

Comparis
on
Condition

Computer Assisted Instruction
Loadman
Read 180
Default
et al.,
(Scholastic English
2011
)
language
arts
instruction

Treatment
Condition

Citation

Age: 14–22,
most in
grades 9–10;
96% Male;
69% African
American;
24% White;
2% Hispanic;
5% Other;
48% with
disabilities;
100% below
proficient but
at least basic
readers at
baseline

Eight
Ohio
Depart
ment of
Youth
Service
s
facilities

106

Age: 13.8–
18.8
(mean=17.1);
100% Male;
100% African
American;
44% in
special
education;
22% with
emotional
disturbance;
7% with
mental
retardation

Demographi
cs

Urban
juvenile
detentio
n facility
on the
East
Coast

Setting

677

45

n
Tre
at

568

n
Co
mp
are
0

90 min., 5
days a
week, for
20 weeks

Duration
and
Frequenc
y
2 hrs 50
min per
day, 5
days a
week, for
6 weeks

California
Achievement
Test (CAT)
in reading
one year
after
baseline
testing

Scholastic
Reading
Inventory
(SRI) score
at end of
intervention

Gray Oral
Reading
rd
Test, 3
Edition
(Passage
[combines
Rate and
Accuracy]
and
Comprehens
ion)

Outcome 1
Metric

Relative
gain:
0.26 SD
(p=.011)
[CAT
analysis is
based on
only 133
treatment
and 110
comparison
students]

Relative
gain: 0.21
SD (p<.001)

Passage:
0.35 SD
(p=.02)
Comprehens
ion: .34 SD
(p=.13)

Outcome 1
Effect Size
Estimate

Outcome 2
(and 3)
Metrics

Outcome
2 (and 3)
Effect Size
Estimate

[4 for
CAT
analy
sis]

5

1

MD
Scal
e

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Treatment
Condition

Fast
ForWord
softwarebased
beginning
reading
program
(Scientific
Learning)

Tune in to
Reading
(TIR), a
program to
teach
reading
through
singing
(Electronic
Learning
Products)

Citation

Shippen et
al, 2012

Calderone
et al, 2009

Default
instructiona
l program
(namely,
FCAT
Explorer,
an online,
standardsbased
program)

Default,
individualiz
ed
academic
and
vocational
training

Comparis
on
Condition

Six
residenti
al sites
for
juvenile
s in the
Florida
correcti
onal
system

Longterm
maximu
m
security
juvenile
facility
in
Alabam
a

Setting

107

Ages not
given; grades
7–11;
100% Male
52% African
American
13% Hispanic
31% White
44% with
disabilities

Age: 11–20
(mean=16.3);
100% Male
53% African
American;
45% White;
2% Other;
Mean IQ: 78;
18% with mild
intellectual or
learning
disabilities

Demographi
cs

64

27

n
Tre
at

39

n
Co
mp
are
24

45
minutes,
twice a
week, for
9 weeks

Duration
and
Frequenc
y
45 min., 5
days a
week, for
11 weeks
(average=
24 days)
Test of
Written
Spelling-4
(TWS-4);
Test of Word
Reading
Efficiency
(TOWRE);
Woodcock
Reading
Mastery
TestRevised/Nor
mative
Update
(WRMTR/NU)
TIR
computeradaptive
cloze
reading
assessment

Outcome 1
Metric

(p>.05 in all
cases)
Relative
gain: 0.21
SD
(p>.05)

(Reading
domain
mean:
-0.172 SD)

Relative
gains: TWS4:
0.226 SD
TOWRE:
-0.142 SD
WRMTR/NU:
-0.201 SD

Outcome 1
Effect Size
Estimate

Outcome 2
(and 3)
Metrics

Outcome
2 (and 3)
Effect Size
Estimate

5

5

MD
Scal
e

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Fast
ForWord
softwarebased
beginning
reading
program
(Scientific
Learning
Corporatio
n)

Scientific
Learning
Corporatio
n, 2004

NA

Comparis
on
Condition

Personalized and Intensive
Instruction
National
Avon Park Default
Council on Youth
juvenile
Crime and
Academy:
correctional
Delinquen
Intensive,
programs
cy, 2009
personaliz
within the
ed,
state
vocational
and
academic
training
with
aftercare

Treatment
Condition

Citation

Florida
Depart
ment of
Juvenile
Justice
facilities

Two
facilities
in the
Virginia
Depart
ment of
Juvenile
Justice

Setting

108

Age: 16–18;
41% African
American;
14%
Hispanic;
44% White;
38% with
special
needs;
65% with
below 6th
grade reading
level and
100% with
below 6th
grade math
level at
baseline

Mean
baseline
reading grade
level: 6.6

Mean grade
level: 8.9

Ages not
given

Demographi
cs

369

29

n
Tre
at

345

n
Co
mp
are
NA

14.2
month
average
stay in
facility
(versus
11.2
months for
compariso
n group)

Duration
and
Frequenc
y
48
minutes or
more, 5
days a
week, for
10 months
(WJ test
group) or
4 months
(STAR
test group)

High school,
GED, or
special
diploma
completion
at time of
release

STAR
Reading
Assessment

Woodcock
Johnson
Tests of
Achievement
rd
, 3 Edition

Outcome 1
Metric

27.1
percentage
points
(p<.01)
T: 49.1%
C: 22.0%

Mean gain:
1.3 grade
levels in
STAR test
group (n=11,
p<.05)

Mean gain:
1.6 grade
levels in WJ
test group
(n=18,
p<.05)

Outcome 1
Effect Size
Estimate

Re-arrest
within 1
year after
release

Employment
1 year
postrelease

Outcome 2
(and 3)
Metrics

Re-arrest:
1.0
percentage
points
(p>.2)
T: 57.2%
C: 56.2%

Employme
nt: 8
percentage
points
(p=.02)
T: 72.4%
C: 64.4%,

Outcome
2 (and 3)
Effect Size
Estimate

5

1

MD
Scal
e

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Intensive,
competenc
y-based
education
with
vocational
training
and
aftercare

Individuali
zed
academic
instruction

Peermanaged
instruction
using an
individualiz
ed
curriculum;
tutorstudent
ratio of 1:1
to 1:2
Individuali
zed,
competenc
y-based
academic
instruction

Skonovd
et al.,
1991

Mayer and
Hoffman,
1982

Kane and
Alley,
1980

Muse,
1998†

Treatment
Condition

Citation

Teachermanaged
instruction
with
individualiz
ed
curriculum;
teacher
student
ratio of 1:3
to 1:7
Default
instruction
in other
schools in
the same
system

Group
(classroom
-level)
instruction

Default
programs
for
juveniles in
the same
county

Comparis
on
Condition

North
Carolina
juvenile
correcti
onal
facilities

Minimu
msecurity
juvenile
correcti
onal
institutio
n in
Minneso
ta

San
Bernardi
no
County
Probatio
n
Depart
ment
Juvenile
Hall
Four
youth
offender
facilities
in
Florida

Setting

109

Age: 12–17;
No additional
demographic
information
provided

Age: 12–17;
100%
identified as
learning
disabled;
mean pretest
math grade
level: 6.0

Age: 16–17;
Wards from
which sample
was drawn
were 21%
African
American;
29%
Hispanic;
50% White
Ages not
given; 100%
Male
52% African
American
48% White

Demographi
cs

66
(stu
den
ts
in 1
sch
ool)

21

68

25

n
Tre
at

4**
(sc
hoo
llev
el
ave
rag
es)

17

75

n
Co
mp
are
20

9-month
school
year
implied

8 weeks
(38 45minute
class
periods)

10 months
(frequency
not given)

Duration
and
Frequenc
y
6 months
in juvenile
facility and
4–6
months in
after care

GED
completion
rate over 3
years

California
Achievement
Test, version
3, Total
Battery
(math,
reading,
language)
Science
Research
Associates
(SRA)
Assessment
Survey
Multilevel
Edition in
mathematics

Re-arrest or
probation
violation
within 6
months after
release

Outcome 1
Metric

59.1
percentage
points (no
hypothesis
test)
T: 67.1%
C: 8.0%

Relative
gain: -0.045
SD (p>.05)

Relative
gain: 2
months of
learning (no
hypothesis
test
available)

-29
percentage
points
(p<.05)
T: 16%;
C: 45%

Outcome 1
Effect Size
Estimate

Outcome 2
(and 3)
Metrics

Outcome
2 (and 3)
Effect Size
Estimate

1

3

3

2

MD
Scal
e

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Diagnostic
evaluation
and
tutoring in
reading
and math

Hill,
Minifie,
and
Minifie,
1984

NA

Comparis
on
Condition
South
Carolina
Depart
ment of
Youth
Service
s
correcti
onal
facilities

Setting

Archwame
ty and
Katsiyanni
s, 2000

Remedial
education
in math or
reading

Nonremedial
education

Nebrask
a Youth
Rehabili
tation
and
Treatme
nt
Center

Other Remedial Reading and Writing Instruction
Simpson,
Orton/Gilli
Default
Two
Swanson,
ngham
language
juvenile
and
structured
arts
youth
Kunkel,
remedial
instruction
detentio
1992
reading
for 45
n
instruction
minutes a
facilities
for 90
day in
(location
minutes a
classes of
not
day in
about 12
given)
groups of
1–6

Treatment
Condition

Citation

110

Age: 13–18;
100% Male;
Baseline
reading grade
level: 4.4;
Treatment
students were
test-verified
as learning
disabled;
comparison
group
students were
teacheridentified as
similarly
disabled
Age: 12–18;
Mean IQ:
94.3;
Treatment
students were
at least one
grade level
behind in
remedial
subject

Ages not
given;
100%
identified as
handicapped;
all were 5–8
years below
grade level in
reading and
mathematics

Demographi
cs

339

32

31

n
Tre
at

166

31

n
Co
mp
are
NA

Not
specified

Actual
mean
dosage:
51.9
instruction
al hours in
treatment
group
versus
46.0 in
control
group

Duration
and
Frequenc
y
One hour
twice a
week for 9
weeks (18
sessions),
in addition
to regular
classroom
instruction

Recidivism
(definition
unspecified)
within 1–7
years after
release

Also, years
of growth for
every 10
hours of
instruction

Years of
growth on
the
Woodcock
Test of
Reading
Mastery

ARI oral
reading
accuracy

Analytical
Reading
Inventory
(ARI) silent
reading
comprehensi
on

Outcome 1
Metric

+9.4
percentage
points
(p<.05)
T: 23.3%
C: 13.9%

0.38 years of
growth per
10 hours of
instruction
(no
hypothesis
test
available)

Relative
gain: 0.86
years
(p=.007)

1 month
improvement
(p>.05)

3 months
improvement
(p<.05)

Outcome 1
Effect Size
Estimate

Re-arrest
within a
year
following
release

KeyMath
Diagnostic
Assessment

Outcome 2
(and 3)
Metrics

-22
percentage
points
(p=.015)
T: 41%
C: 63%

Outcome
2 (and 3)
Effect Size
Estimate
Statistically
significant
gain
(p>.05);
magnitude
unspecified

2

3

1

MD
Scal
e

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Treatment
Condition

Instruction
in reading
road signs

Visual
Response
System
training in
completing
a job
application

Learning
strategies
instruction
in
decoding,
paraphrasi
ng, taking
tests, and
setting
goals

Citation

Murph and
McCormic
k, 1985

Heward,
McCormic
k, and
Joynes,
1980

Platt and
Beech,
1994

Students’
previous
instructiona
l
experience
s in reading

Students’
previous
instructiona
l
experience
s

Students’
previous
instructiona
l
experience
s

Comparis
on
Condition

Adult
and
juvenile
detentio
n
centers
in
Florida

Training
Institute
of
Central
Ohio, a
juvenile
correcti
onal
facility
Correcti
onal
facility
for
juvenile
offender
s
(location
not
given)

Setting

111

Age: 15 to 18;
100% Male;
100%
classified as
“educable
mentally
retarded”;
mean
baseline
reading grade
level: 4.6
100% under
age 21;
No additional
demographic
information
provided

Age: 16 to 18;
100% Male;
IQs: 70–79;
mean
baseline
reading grade
level: 2.5

Demographi
cs

5
sel
ect
ed
stu
den
ts
tau
ght
by
27
tea
che
rs
trai
ned
in
the
met
hod

7

5

n
Tre
at

NA

NA

n
Co
mp
are
NA

Not
reported

11 45minute
sessions

Duration
and
Frequenc
y
9 to 24
fifteenminute
instruction
al
sessions
per
student

Passage
comprehensi
on (pre vs.
during-andpost)

Items
answered
correctly on
a 35-item
Master
Employment
Application
(pretest
versus
probes and
follow-up)
Words read
correctly
(pre vs.
during-andpost)

Road signs
recognized
out of 9 (pre
vs. duringand- post)

Outcome 1
Metric

[Note: the 5
students for
whom data
are given
are only a
small subset
of students
exposed to
treatment.]

Mean gain:
19.3
percentage
points

Mean gain:
11.9
percentage
points

Mean gain:
17.8 items

Mean gain:
8.1 signs

Outcome 1
Effect Size
Estimate

Outcome 2
(and 3)
Metrics

Outcome
2 (and 3)
Effect Size
Estimate

1*

1*

1*

MD
Scal
e

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Assignme
nt of
visual,
imagery,
and report
writing
tasks

Sinatra,
1984

Vocational Education
Roos,
Re2006
Integration
of
Offenders
–Youth
(RIO-Y)
career
developme
nt course
Wilson,
Vocational
1994
education
elective
participatio
n in facility
(auto,
business,
constructio
n, food,
special
cooperativ
e services)

Treatment
Condition

Citation

Texas
Youth
Commis
sion
facilities

Colorad
o
Division
of Youth
Service
s
facilities

Participatio
n in
nonvocatio
nal
education

Shortterm
adolesc
ent
treatme
nt
center
for
incarcer
ated
youth

Setting

No
participatio
n in a
career
developme
nt course

Students’
previous
instructiona
l
experience
s in writing

Comparis
on
Condition

112

Age: 11–18;
100% Male;
16% Black;
34%
Hispanic;
48% White;
2% Other

Age: 18–21;
34% African
American;
38%
Hispanic;
28% White

Mean age:
15.3;
20% Male;
Baseline
reading grade
level: 5–6

Demographi
cs

260

582

20

n
Tre
at

143

920

n
Co
mp
are
NA

30 days of
instruction
(versus no
comparabl
e
instruction
al hours in
compariso
n group)
Not
reported

Duration
and
Frequenc
y
Weekly
practice
over
several
months

Reincarceration
within 5
years after
treatment

Employment
1 year after
release

Writing
proficiency
score
(assessed
by three
raters)
averaged
across three
tasks, as
compared
with a
pretest

Outcome 1
Metric

-17.1
percentage
points
(p<.05)
T: 61.2%
C: 78.3%

Odds ratio:
1.39 (p<.01)

Mean gain:
16.3
percentage
points

Outcome 1
Effect Size
Estimate

Re-arrest
within 1
year after
release

Outcome 2
(and 3)
Metrics

Odds ratio:
0.97 (p=.8)

Outcome
2 (and 3)
Effect Size
Estimate

2

3

1

MD
Scal
e

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Vocational
education
program
availability
(business,
drafting,
and
carpentry)

DelliCarpin
i, 2010

Katsiyanni
s and
Murray,
2000

GED
completion
in facility

GED Completion
Jeffords
GED
and
completion
McNitt,
in facility
1993

Treatment
Condition

Citation

Texas
Youth
Commis
sion or
Gulf
Coast
Trades
Center
correcti
onal
program
s
A youth
rehabilit
ation
and
treatme
nt
facility
in
Nebrask
a

No GED
completion
in facility

No GED
completion
in facility

Eastern
Suffolk
BOCES
Progra
m for
Incarcer
ated
Youth in
NY
State

Setting

Participatio
n in default
educational
program

Comparis
on
Condition

113

Age: 12–18;
100% Male

Age: 16–21;
No additional
demographic
information
provided

Age: 16–21;
No additional
demographic
information
provided

Demographi
cs

284

475

465

n
Tre
at

265

124
2

n
Co
mp
are
581

At least 4
months
spent in
facility

Not
reported

Duration
and
Frequenc
y
8 week
module
(daily
instruction
implied)

Reincarceration
within 3
years after
release

Reincarceration
within 1 year
after release

GED pass
rate

Outcome 1
Metric

-12.5
percentage
points
(p<.01) T:
47.5%
C: 60.0%

-5.8
percentage
points
(p<0.1)

7.6
percentage
points
(p<.001)
T: 13.1%
C: 5.5%

Outcome 1
Effect Size
Estimate

Outcome 2
(and 3)
Metrics

Outcome
2 (and 3)
Effect Size
Estimate

2

3

2

MD
Scal
e

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

114

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Appendix B. RAND Correctional Education Survey Questionnaire

STATE CORRECTIONAL EDUCATION DIRECTOR
DATA COLLECTION FORM
Thank you for participating in this State Correctional Education Director data
collection form being conducted by the RAND Corporation. The questions in this form
focus on academic education and vocational or career/technical education (CTE)
provided in state prison or correctional facilities for incarcerated adults.
Our goal is to understand how correctional education is currently provided and to
whom, the effects of recent fiscal cuts on correctional education, the use of technology,
preparations for the 2014 GED Exam, and how correctional education is organized and
funded within your state. Your organization’s responses and your identity will be kept
confidential. This study is funded by the Bureau of Justice Assistance (BJA), Office of
Justice Programs, U.S. Department of Justice. We kindly request that you complete this
survey by August 30, 2013.
If you have any questions about the survey, please contact Lois Davis, RAND Project
at email: Lmdavis@rand.org, tel. 310.393.0411, ext. 7330. If you have any questions
about the project in general, please contact Dr. Gary Dennis, BJA Project Officer and
Senior Policy Advisor for Corrections, email: Gary.Dennis@usdoj.gov.
Please mail the completed survey to Lois Davis, Ph.D., RAND, 1776 Main Street,
Santa Monica, CA 90407-2138. Or email the survey to her at Lmdavis@rand.org.
CONTACT INFORMATION
Name of Person Completing this Form: ________________________________________
Title: _____________________________________________________________________
Department/Organization: __________________________________________________
Telephone: ( )__________________________________________________________
Email address: ____________________________________________________________
State: ____________________________________________________________________
A. In your current position, which of the following activities do you oversee or have responsibility
for:
___Academic education programs only
___Vocational education or career/technical education only
___Both academic programs and vocational education/CTE programs

115

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

I. OVERVIEW OF YOUR STATE'S CORRECTIONAL EDUCATION PROGRAMS
1. What types of educational programs are currently available to adult prisoners

(Please make sure to also include those programs that ore available in
your state, but not under your supervision.) (Mark all that apply)

within your state?

_ _Literacy training/Adult basic education (ABE) (i.e., basic skills instruction in
arithmetic, reading, and writing)
_ _Adult secondary education (ASE) (i.e., preparation to complete a high school
diploma program)
_ _ General Education Development (GED) test preparation
_ _Adult post-secondary education (PSE)/college courses
_ _Vocational skills training/Career Technical Education (CTE)
_ _English as a Second Language (ESL) courses
_ _Special education (e.g., for offenders with learning disabilities)
_ _Other, please specify: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

2. Is participation in correctional education programs mandated by your state (either
by legislative statute or policy)? (Mark all that apply)
__Yes, participation in correctional education programs is mandated by the state
for:

0 All adult inmates
D Adult inmates without a high school diploma or GED
D Adult inmates below a 6'h or Sth grade education level
D Other (please specify): _ _ _ _ _ _ _ _ _ _ _ _ _ _ __
__No, participation in correctional education programs is voluntary
3. Are work assignments currently considered to be part of correctional education
within your state's prison system?
__Yes

116

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

____No
____Don’t know

4. Please indicate which of the following funding sources are used to pay adult
inmates’ post-secondary education or college courses in your state (Mark all that apply):
_____Inmate benefits or welfare funds
_____State funding (e.g., department of corrections budget allocation)
____College or university funding
_____Private funding (e.g., foundations, religious/community group, individual
donation)
_____Personal or family finances
_____Not Applicable, our state does not offer post-secondary/college courses
to adult inmates

We now want to ask you a couple of questions about correctional
education within your state’s correctional facilities.
By adult state correctional facility we mean prison facilities that hold sentenced adult
offenders in state custody. It excludes residential treatment or community programs.
5. What was the total number of adult state correctional facilities offering correctional education
programs in:
Fiscal Year 2009

Fiscal Year 2012

Total Number of Facilities Offering:
Academic Programs*

_______________

117

_______________

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Vocational education/CTE programs _______________

_______________

[*Includes adult basic education (ABE), adult secondary education (ASE), GED
preparation, adult post-secondary education (PSE), and English as a Second Language (ESL)
programs]

II. CORRECTIONAL EDUCATION CAPACITY
6. Please indicate the time period that your state’s fiscal year covers (e.g., January
through December; July through June, or October through September):

From: ________ (month)

To: ________ (month)

Now we are going to ask you to consider the total number of students
in correctional education programs and the number of teachers and
instructors.

7. What was the total number of adult students enrolled in your state’s correctional
education programs in Fiscal Year 2009 and Fiscal Year 2012
Types	
  of	
  Educational	
  
Service	
  

Fiscal	
  Year	
  2009	
  
(Number	
  of	
  
Adult	
  Students)	
  

Fiscal	
  Year	
  
2012	
  
(Number	
  of	
  
Adult	
  Students)	
  

Adult	
  Basic	
  Education	
  
(ABE)	
  
Adult	
  Secondary	
  
Education	
  (ASE)	
  
GED	
  (General	
  
Education	
  Development)	
  
Test	
  preparation	
  
Vocational	
  skills	
  

118

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

training/career	
  technical	
  
education	
  (CTE)	
  
Post-­‐secondary	
  
education/	
  college	
  courses	
  

8. Does your state screen adult inmates for special education needs?
____Yes
____No (please skip to Q5)

8a. If yes, in Fiscal Year 2012 how many adult students were on a formal
Individualized Education Program (IEP) plan within your correctional education system?
Number of IEP students: _____________

9. What was the total number of academic teachers/ instructors and vocational
education/CTE instructors in Fiscal Years 2009 and 2012?
Fiscal Year
2009
(Number)
	
  Academic Programs

Number of teachers that are state
employees (include full-time and
part-time employees in your response)
• Number of contract instructors
Vocational Education/CTE
Programs
•

•

Number of vocational instructors
that are state employees (include
full-time and part-time employees in
your response)
119

Fiscal Year
2012
(number)

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

•

Number of contract instructors

THE NEXT SET OF QUESTIONS ASK ABOUT WHAT IMPACT, IF ANY, OF
BUDGET CUTS OR OTHER FISCAL PRESSURES MAY HAVE HAD ON YOUR
STATE’S CORRECTIONAL EDUCATION PROGRAMS.
10. Between Fiscal Years 2009–2012, did your state’s correctional education
programs (academic and/or vocational education/CTE) experience a decrease in
funding as a result of budget cuts or other fiscal pressures?
_____Yes
_____No (Skip to 17)

11. What changes, if any, were made to staffing levels and capacity in response to
budget cuts or other fiscal pressures during Fiscal Years 2009–2012?

Changes	
  Implemented	
  to	
  Staffing	
  Levels	
  
and	
  Capacity	
  
Hiring	
  freeze(s)	
  of	
  teachers/instructors	
  
were	
  implemented	
  
Staff	
  furloughs	
  of	
  teachers/instructors	
  
were	
  made	
  
Did	
  not	
  fill	
  vacant	
  teaching/instructor	
  
positions	
  	
  
Delayed	
  and/or	
  cancelled	
  pay	
  increases	
  
for	
  teachers/instructors	
  
Reduced	
  salaries	
  and/or	
  benefits	
  for	
  
teachers/instructors	
  
Reduced	
  the	
  number	
  of	
  
teachers/instructors	
  for:	
  	
  

120

Mark	
  all	
  that	
  
Apply	
  

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

•
•

Academic	
  programs
Vocational	
  education/career
technical	
  education	
  (CTE)	
  programs
Reduced	
  or	
  eliminated	
  contracts	
  with	
  
community	
  or	
  technical	
  colleges	
  
Reduced	
  the	
  number	
  of	
  course	
  offerings	
  
for:	
  
•
•

Academic	
  programs
Vocational	
  education/career
technical	
  education	
  (CTE)	
  programs

Other	
  (please	
  specify):	
  
_______________________	
  
None	
  

12. Did your state increase the number of contract teachers/instructors for the
following programs during Fiscal Years 2009–2012?
_____Yes, for academic programs
_____Yes, for vocational education/career technical education (CTE) programs
_____No

13. Did your state increase its use of inmates as staff in the classroom during Fiscal
Years 2009–2012? If so, briefly describe:
_____Yes, we did increase the use of inmates as staff in the classroom due to budget
cuts
or other fiscal pressures
_____Yes, but the increase use of inmates as staff in the classroom was not in direct
response to budget cuts or other fiscal pressures
_____No
13a. If yes, in what ways were inmates used:
_____ As peer tutors to assist students with coursework
_____ As a clerk assisting with administrative tasks
_____ To help oversee a computer lab
121

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

_____ To assist with vocational education/CTE programs
_____ Other (please specify): _______________________________________

14. As a result of budget cuts or other fiscal pressures, did your state change its
policies regarding mandatory participation in correctional education programs during
Fiscal Years 2009–2012?
_____Yes, for certain academic programs participation was changed to voluntary
(Briefly describe): _________________________________________________
_____Yes, for certain inmates participation in academic programs was changed to
voluntary (Briefly describe): ______________________________________________
_____No changes were made to our state’s policies regarding mandatory
participation in
correctional education programs
15. Do you anticipate any additional budget cuts to your state’s correctional
education programs in the upcoming fiscal year (Fiscal Year 2013)?
_____Yes
_____No
_____Don’t Know
We now want to ask you about the last two fiscal years (Fiscal Years
2011–2012).
16. During the past two fiscal years (2011–2012), has your state’s correctional
education programs (academic and vocational education/CTE)) experienced an increase
in funding?
_____Yes
_____No

16a. If yes, how has the increase in funding been used by your correctional education
system? (Mark all that apply)
_____ Increased the number of teachers/instructors for:

122

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

• Academic programs
• Vocational education/career technical education (CTE) programs
_____ Increased the number of contractor teachers/instructors for:
• Academic programs
• Vocational education/career technical education (CTE) programs
_____ Increased the number of vocational programs offered
_____ Increased the capacity of:
• academic programs
• vocational education/career technical education (CTE) programs
_____Reinstated the number of post-secondary or college courses offered
_____Expanded the number of post-secondary or college courses offered
_____ Increased classroom space for:
• academic programs
• vocational education/career technical education (CTE) programs
_____ Increased the number of computer labs
_____ Purchased computer equipment
_____ Other (please specify): _______________________________________

USE OF TECHNOLOGY
We now want to ask you about the use of technology in your state’s
correctional education system. These questions pertain to both academic
and vocational education/CTE programs.
17. How many correctional facilities within your state have a computer lab?
Number of facilities with a computer lab(s): _______________
18. What types of technology hardware and networks does your state correctional
education system use? (Mark all that apply)
_____Local area network(s) (LAN)
_____Statewide or wide area network(s) (WAN)
_____Local area network(s) (LAN)
_____Closed-circuit TV
_____Desktop computers (standalone or networked)
_____Mobile laptops
_____Kindles

123

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

_____iPads
_____Other technology (please specify):
__________________________________________
_______________________________________________________________________
_______________________________________________________________________
19. What means are used to provide instruction for academic programs or vocational
education/CTE courses offered: (Mark all that apply)
_____On-site instruction
_____Video/satellite instruction
_____One-way
_____Interactive
_____Internet-based instruction
_____One-way
_____Interactive
_____Correspondence courses
_____Other (please specify): ______________________________________________
20. In what ways is internet technology being used in your state correctional
education classrooms (academic and vocational education/CTE programs) and/or
libraries? (Mark all that apply)
_____Only teachers/instructors have access to live Internet technology
_____Students have full access to live Internet
_____Students have restricted access to live Internet
_____Students use simulated Internet programs
_____Students do not have access to any Internet technology
_____ Other, please specify________________________________

PREPARATION FOR THE 2014 GED EXAM
In 2014, the new GED exam will be implemented along with computerbased testing. We now would like to ask you about your state’s
preparations for the 2014 GED exam (or another high school equivalency
examination) and for computer-based testing.
21. Is your state planning on implementing the 2014 GED exam?
124

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

_____ Yes
_____ No, our state is exploring other high school equivalency examinations (skip to
Q25)

22. How many correctional facilities within your state are currently set-up or will be by
January 1, 2014 to implement computer-based testing for the 2014 GED exam?
Number of correctional facilities: _______________

23. Is your state planning to use computer workstations or laptops for inmates taking
the GED test (Mark only one)?
_____
_____
_____
_____

Computer workstations only
Laptops only
Combination of computer workstations and laptops
Other (please specify): _______________________________________

24. As part of your state’s preparations, will professional development training be
provided to your correctional teachers/ instructors to prepare them to teach the new GED
exam?
_____ Yes, we are providing professional development training for the new GED
exam

_____ No (skip to Q25)

24a. If yes, what subjects will your correctional education system’s professional
development training address? (Mark all that apply)
_____Training on the administration of the test process
_____Training on the test protocols
_____Training on computer literacy
_____Assistance with instruction development
_____Training on instruction aligned with the common core standards
_____Training on test security requirements
_____Other (please specify) __________________________________________________

125

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

25. In your view, what is the likely effect of the new GED exam and computer-based
testing requirement on your state’s correctional education population:
LIKERT SCALE
1=negative effect
effect
•
•
•

3= no effect

5=positive

On the number of inmates who will be prepared to take the new GED exam
On the length of time it will take to prepare inmates to take the new GED exam
On GED completion rates

26. What concerns, if any, do you have about the forthcoming changes to the 2014
GED exam and the move towards computer-based testing? (Mark all that apply)
_____Cost of purchasing equipment for computer-based testing
_____Fewer students may be ready to take the 2014 GED exam due to length of time
it takes to prepare them for the new exam
_____ Limited access to computers may preclude some students from taking the
GED exam
_____More extensive preparation required for the 2014 GED exam may make it
difficult for some students to complete their test preparations while they are in
prison
_____Security concerns about access to the Internet for the GED exam may make it
more difficult to do testing
_____Teachers may not be prepared to teach the new GED exam
_____Teachers may not be prepared to implement computer-based testing
_____Other (please specify) __________________________________________________
_____No concerns

OUTCOMES/PERFORMANCE INDICATORS
27. Which of the following outcome indicators does your state’s correctional
education system track for academic and vocational education/CTE programs: (Mark all
that apply)
126

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

_____ Gains in reading or math skills
______Number of
• GED tests passed
• GED certificates earned
• High school degrees awarded
• College credits earned
• College degrees earned (e.g., Associate degrees)
• Vocational certificates awarded
• National or industry-recognized certificates awarded
_____Other (please specify): _______________________
28. What post-release indicators does your state’s correctional education system
consider to be important outcome measures? (Mark all that apply)
______Post-release employment
______Job retention
______College attainment
______Degrees awarded
______Enrollment in vocational training programs
______Enrollment in post-secondary education/college courses
______Recidivism
_____Other (please specify): _______________________

29. What national or industry-recognized certifications, if any, does your state’s
correctional education system offer? (Mark all that apply)
_____National Center for Construction Education and Research (NCCER)
_____National Institute for Automotive Service Excellence (ASE)
_____Microsoft Office certification (please specify): __________________________
_____American Welding Society
_____Occupational Safety and Health Administration (OSHA) training programs
_____Apprenticeship cards (e.g., plumbing, electrical)
_____Other (please specify): ______________________________________________
_____Other (please specify): ______________________________________________
_____Other (please specify): ______________________________________________

ORGANIZATION OF CORRECTIONAL EDUCATION WITHIN YOUR STATE

127

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

By program authority we refer to the agency or department with decision-making
authority with regard to correctional education policy and administration for incarcerated
adults.
30. How is correctional education administered within your state’s correctional
institutions? (Mark only one)
_____ The majority of correctional education program authority is vested within
one central state agency
_____ Correctional education program authority is vested among several state
agencies
_____ Other (please specify): ____________________________________________
31. Which of the following is the lead agency(s) for administering adult correctional
education within your state? (Mark all that apply)
_____ Department of Public Safety
_____ Department of Corrections
_____ Department of Education
_____ Department of Adult Education
_____ Department of Labor
_____ Other (please specify): _____________________________________

FUNDING ISSUES
32. What was the total amount of your state’s correctional education budget in Fiscal
Years 2009 and 2012?
Total Correctional Education Budget
$	
  Mil.	
  

Thou.	
  

Dol.	
  

Fiscal	
  Year	
  
2009	
  

000	
  

Fiscal	
  Year	
  
2012	
  

000	
  

33. In
which federal,
state or private grant programs does your state’s correctional education system
currently participate in or receives funding from (Mark all that apply)?
_____ESEA (Elementary and Secondary Education Act), Title I, Part D

128

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

_____ESEA, Title II, Part A
_____Workforce Investment Act, Title II (also known as the Adult Education Family
Literacy Act)
_____Federal Second Chance Act (SCA) grants
_____Office of Juvenile Justice and Delinquency Prevention (OJJDP)
_____Bureau of Justice Assistance (BJA) grant funding (other than Second Chance Act)
_____National Institute of Justice (NIJ)
_____Foundations (e.g. Sunshine Lady) (please specify): _______________________
__________________________________________________________________
__________________________________________________________________
34. Please indicate the amount of funding your correctional education system received
in 2012 from the following three sources:

Amount of Funding Received From:
$ Mil.	
  
Thou.	
  
Dol.	
  

Workforce
Investment Act, Title
II	
  
Perkins Act	
  
State higher
education/aid
resources for postsecondary education
or training	
  

000	
  

000	
  
000	
  

Thank you for participating in this data collection effort. Please provide
in the space below any comments or feedback you may have about it.
____________________________________________________________________	
  
____________________________________________________________________	
  
____________________________________________________________________	
  
____________________________________________________________________	
  

129

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

____________________________________________________________________	
  
____________________________________________________________________	
  
____________________________________________________________________	
  

GLOSSARY OF TERMS USED IN THE SURVEY
The following definitions are provided to assist you in completing this form.	
  
Adult Basic Education (ABE): basic skills instruction in arithmetic, reading, and writing
Adult Secondary Education (ASE): instruction to complete high school or prepare for a certificate
of high school equivalency, such as the General Education Development (GED)
General Education Development (GED): tests that are a group of subject tests which, when
passed, certify that the taker has American or Canadian high school-level academic skills.
Adult Postsecondary Education (PSE): college-level instruction that enables an individual to earn
college credit that may be applied toward a two-year or four-year postsecondary degree
Vocational education or Career Technical Education (CTE): training in general employment skills
and in skills for specific jobs or industries
Adult state correctional facility: prison facilities that hold sentenced adult offenders in state
custody. It excludes residential treatment or community programs.	
  

130

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

References

*Indicates a reference included in the systematic review of juvenile correctional education.
Aizer, Anna, and Joseph J. Doyle, Juvenile Incarceration, Human Capital and Future Crime:
Evidence from Randomly-Assigned Judges, Cambridge, Mass.: Massachusetts Institute of
Technology, 2013. As of February 3, 2014:
http://www.mit.edu/~jjdoyle/aizer_doyle_judges_06242013.pdf
Allen, F., The Decline of the Rehabilitative Ideal: Penal Policy and Social Purpose, New Haven,
Conn.: Yale University Press, 1981.
Allen-DeBoer, Robyn A., Kimber W. Malmgren, and Mary-Elizabeth Glass, “Reading
Instruction for Youth with Emotional and Behavioral Disorders in a Juvenile Correctional
Facility,” Behavioral Disorders, Vol. 32, No. 1, 2006, pp. 18–28.
Aos, S., M. Miller, and E. Drake, “Evidence-Based Adult Corrections Programs: What Works
and What Does Not,” Olympia, Wash.: Washington State Institute for Public Policy, #06-011201, January 2006.
*Archwamety, Teara, and Antonis Katsiyannis, “Academic Remediation, Parole Violations, and
Recidivism Rates Among Delinquent Youths,” Remedial and Special Education, Vol. 21,
No. 3, May/June, 2000, pp. 161–170.
Associated Press, “Daniels Endorses Findings of Criminal Code Review,” December 15, 2010.
As of March 15, 2011:
http://www.ibj.com/daniels-endorses-findings-of-criminal-codereview/PARAMS/article/24023
Association of State Correctional Administrators, Member Survey of Education Programming
and Services Tools and Member Survey of Alternative to GED Testing, 2013.
Bazos, A., and J. Hausman, Correctional Education as a Crime Control Program, Los Angeles,
Calif.: UCLA School of Public Policy and Research, Department of Policy Studies, March
2004.
Becker, Gary S., “Crime and Punishment: An Economic Approach,” Journal of Political
Economy, Vol. 76, 1968, pp. 169–217.
Biggs, Marie C., Susan P. Homan, Robert F. Dedrick, Vanessa Minick, and Timothy Rasinski,
“Using an Interactive Singing Software Program: A Comparative Study of Struggling Middle
School Readers,” Reading Psychology, Vol. 29, 2008, pp. 195–213.

131

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Borenstein, Michael, Larry V. Hedges, Julian P. T. Higgins, and Hannah R. Rothstein,
Introduction to Meta-Analysis, Hoboken, N.J.: Wiley, 2009.
Bozick, Robert and Benjamin Dalton, “Balancing Career and Technical Education with
Academic Coursework: The Consequences for Mathematics Achievement in High School,”
Educational Evaluation and Policy Analysis, Vol. 35, No. 2, 2013, pp. 123–138.
Caldagno, Juan Carlos, and Bridget Terry Long, The Impact of Postsecondary Remediation
Using a Regression Discontinuity Approach: Addressing Endogenous Sorting and
Noncompliance, New York: National Center for Postsecondary Research, 2008.
Calderone, Cynthia, Susan V. Bennett, Susan Homan, Robert F. Dedrick, and Anne Chatfield,
“Reaching the Hard to Reach: A Comparison of Two Reading Interventions with
Incarcerated Youth,” Middle Grades Research Journal, Vol. 4, No. 3, 2009, pp. 61–80.
California Rehabilitation Oversight Board, Biannual Report, Sacramento, Calif., November 7,
2011.
Carnine, D. W., J. Silbert, and E. J. Kameenui, Direct Instruction Reading, 3rd ed., Upper Saddle
River, N.J.: Prentice Hall, 1997.
Carson, E. A., and W. J. Sabol, Prisoners in 2011, Washington, D.C.: Bureau of Justice
Statistics, NCJ 239808, December 5, 2012.
Carson, E.A., and D. Golinelli, Prisoners in 2012—Advance Counts, Washington, D.C.: Bureau
of Justice Statistics, NCJ 242467, July 16, 2013.
Chambers, Jay G., Thomas B. Parrish, and Jenifer J. Harr, What Are We Spending on Special
Education in the United States, 1999–2000? Report 1, Washington, D.C.: American
Institutes for Research Center for Special Education Finance, 2004. As of February 3, 2014:
http://www.csef-air.org/publications/seep/national/AdvRpt1.PDF
Chappell, C. A., “Post-Secondary Correctional Education and Recidivism: A Meta-Analysis of
Research Conducted 1990–1999,” electronic thesis or dissertation, University of Cincinnati,
2003.
Childress, Stacey, “Shared Attributes of Schools Implementing Personalized Learning,” post on
the blog “Thinking Out Loud: Ideas about Next Generation Learning,” Dcember 19, 2013.
As of February 3, 2014:
http://nextgenstacey.com/2013/12/19/shared-attributes-of-schools-implementingpersonalized-learning/
Chlup, D. T., “Chronology of Corrections Education,” Focusing on Basics: Connecting Research
and Practice, Vol. 7, Issue D, Sept. 2005. As of February 3, 2014:
http://www.ncsall.net/index.php@id=865.html

132

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Coley, R .J. and P. E. Barton, Locked Up and Locked Out: An Educational Perspective on the
U.S. Prison Population. Princeton, N.J.: Educational Testing Service, 2006. As of February
3, 2014:
http://www.ets.org/Media/Research/pdf/PIC-LOCKEDUP.pdf
Coulter, Gail, “Using One-to-One Tutoring and Proven Reading Strategies to Improve Reading
Performance with Adjudicated Youth,” Journal of Correctional Education, Vol. 55, No. 4,
2004, pp. 321–333.
Crayton, A., and S. R. Neusteter, “The Current State of Correctional Education,” paper presented
at Reentry Roundtable on Education, New York: John Jay College of Criminal Justice,
Prisoner Reentry Institute, 2008.
Davis, Lois M., Robert Bozick, Jennifer L. Steele, Jessica Saunders, and Jeremy N. V. Miles,
Evaluating the Effectiveness of Correctional Education: A Meta-Analysis of Programs That
Provide Education to Incarcerated Adults, Santa Monica, Calif.: RAND Corporation, RR266-BJA, 2013. As of February 3, 2014:
http://www.rand.org/pubs/research_reports/RR266.html
*DelliCarpini, Margo, “Building a Better Life: Implementing a Career and Technical Education
Program for Incarcerated Youth,” Journal of Correctional Education, Vol. 61, No. 4, 2010,
pp. 283–295.
Division of Criminal Justice Services, Preliminary Impact of 2009 Drug Law Reform (October
2009– September 2010), New York: Division of Criminal Justice Services, 2010.
*Drakeford, William, “The Impact of an Intensive Program to Increase the Literacy Skills of
Youth Confined to Juvenile Corrections,” Journal of Correctional Education, Vol. 53, No. 4,
2002, pp. 139–144.
Engelmann, S., L. Meyer, L. Carnine, W. Becker, J. Eisele, and G. Johnson, Corrective Reading:
Decoding Strategies, Worthington, Ohio: SRA/McGraw-Hill, 1999.
Farrington, D. P., D. C. Gottfredson, L. W. Sherman, and B. C. Welsh, “The Maryland Scientific
Methods Scale,” in L. W. Sherman, D. P. Farrington, B. C. Welsh, and D. L. MacKenzie,
eds., Evidence-Based Crime Prevention, London: Routledge, 2002, pp. 3–21.
Gaes, G. G., “The Impact of Prison Education Programs on Post Release Outcomes,” paper
presented at the Reentry Roundtable on Education, John Jay College of Criminal Justice,
New York, March 31, 2008
Gagnon, Joseph C., Brian R. Barber, Christopher Van Loan, and Peter E. Leone, “Juvenile
Correctional Schools: Characteristics and Approaches to Curriculum,” Education and
Treatment of Children, Vol. 32, No. 4, 2009, pp. 673–696.
GED Testing Service, “About Ged Testing Service: History of the Ged Test,” 2013.
133

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Gellar, A., I. Garfinkel, and B. Western, “The Effects of Incarceration on Employment and
Wages: An Analysis of the Fragile Families Survey,” Princeton, N.J.: Princeton University,
Center for Research on Child Wellbeing, Working Paper # 2006-01-FF, August 2006.
Gehring, T. “The History of Correctional Education,” 2008. As of February 3, 2014:
http://www.ceanational.org/history.htm
Glaze, Lauren E., and Erinn J. Herberman, Correctional Populations in the United States, 2012,
Washington, D.C.: Bureau of Justic Statistics, Office of Justice Programs, U.S. Department
of Education, December, 2013. As of February 3, 2014:
http://www.bjs.gov/content/pub/pdf/cpus12.pdf
Glaze, Lauren E., and Erika Parks, Correctional Populations in the United States, 2011,
Washington, D.C.: Bureau of Justic Statistics, November, 2012. As of February 3, 2014:
http://www.bjs.gov/content/pub/pdf/cpus11.pdf
Gorgol, L. E., and Sponsler, B. A., Unlocking Potential: Results from a National Survey of
Postsecondary Education in State Prisons, Washington, D.C.: Institute for Higher Education
Policy, 2011.
Greenberg, E., E. Dunleavy, and M. Kutner, Literacy Behind Bars: Results from the 2003
National Assessment of Adult Literacy Prison Survey, Washington, D.C.: Institute for
Education Services (IES) National Center for Education Statistics, U.S. Department of
Education, NCES 2007-473, May 2007.
Harlow, C. W., Education and Correctional Populations, Washington, D.C.: Bureau of Justice
Statistics, NCJ 195670, January 2003; revised April 15, 2003.
Hazel, Neal, Cross-National Comparison of Youth Justice, London: Youth Justice Board for
England and Wales, 2008.
Heckman, James J. and Yona Rubinstein, “The Importance of Noncognitive Skills: Lessons from
the Ged Testing Program,” American Economic Review, Vol. 91, No. 2, 2001, pp. 145–149.
Henrichon, C., and R. Delaney, The Price of Prisons: What Incarceration Costs Taxpayers,
Center on Sentencing and Corrections, New York: Vera Institute of Justice, 2012.
Heward, W. L., S. H. McCormick, and Y. Joynes, “Completing Job Applications: Evaluation of
an Instructional Program for Mildly Retarded Juvenile Delinquents,” Behavioral Disorders,
Vol. 5, 1980, pp. 223–234.
Hill, N. C., D. G. Minifie, and E. L Minifie, “Handicapped Adolescent Delinquents: Educational
Diagnosis and Tutoring,” Remedial and Special Education, Vol. 5, No. 5, 1984, pp. 44–45.
Hill, C., “Inmate Education Programs,” Corrections Compendium, Vol. 33, No. 3, May/June
2008.

134

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Holzer, H. J., S. Raphael, and M. A. Stoll, “Employment Barriers Facing Ex-Offenders,” paper
presented at the Urban Institute Re-Entry Roundtable, New York University Law School,
May 19–20, 2003.
Horn, Michael B., and Heather Staker, The Rise of K–12 Blended Learning, San Mateo, Calif.:
Innosight Institute, 2011.
*Houchins, David E., Kristine Jolivette, Mike P. Krezmien, and Heather M. Baltodano, “A
Multi-State Study Examining the Impact of Explicit Reading Instruction with Incarcerated
Students,” Journal of Correctional Education, Vol. 59, No. 1, 2008, pp. 65–85.
*Jeffords, Charles R. and Scott McNitt, The Relationship Between Ged Attainment and
Recidivism: An Evaluation Summary, Austin, Tex.: Texas Youth Commission, 1993.
Kane, B. J., and G. R Alley, “A Peer-Tutored, Instructional Management Program in
Computational Mathematics for Incarcerated, Learning Disabled Juvenile Delinquents,”
Journal of Learning Disabilities, Vol. 13, 1980, pp. 39–42.
Karoly, L., “21st Century at Work: Trends and Implications” briefing, December 2013, RAND
Corporation.
*Katsiyannis, Antonis, and Teara Archwamety, “Academic Remediation/Achievement and Other
Factors Related to Recidivism Rates Among Delinquent Youths,” Behavioral Disorders,
Vol. 24, No. 2, 1999, pp. 93–101.
Kemple, James J., and Cynthia J. Willner, Career Academies: Long-Term Impacts on Labor
Market Outcomes, Educational Attainment, and Transitions to Adulthood, New York:
MDRC, June, 2008.
Kratochwill, T. R., John Hitchcock, R. H. Horner, J. R. Levin, S. L. Odom, D. M. Rindskopf,
and William R. Shadish, Single-Case Design Technical Documentation: Version 1.0,
Washington, D.C.: What Works Clearinghouse, 2010,
http://ies.ed.gov/ncee/wwc/pdf/wwc_scd.pdf
Krezmien, Mike P., and C. A. Mulcahy, “Literacy and Delinquency: Current Status of Reading
Interventions with Detained and Incarcerated Youth,” Reading and Writing Quarterly, Vol.
24, 2008, pp. 219–238.
Kyckelhahn, T., State Corrections Expenditures, FY 1982–2010, Washington, D.C.: Bureau of
Justice Statistics, December 2012, NCJ 239672, 2012.
Leone, Peter E., “Education Services for Youth with Disabilities in a State-Operated Juvenile
Correctional System: Case Study and Analysis,” Journal of Special Education, Vol. 28,
1994, pp. 43–58.

135

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Leone, Peter E., Sheri M. Meisel, and Will Drakeford, “Special Education Programs for Youth
with Disabilities in Juvenile Corrections,” Journal of Correctional Education, Vol. 53, No. 2,
June, 2002, pp. 16–20.
Leslie, Lauren and JoAnne Schudt Caldwell, Qualitative Reading Inventory—III, Boston: Allyn
& Bacon, 2000.
Lillis, J., “Prison Education Programs Reduced,” Corrections Compendium, Vol. 19, No. 3,
1994, pp. 1–4.
Lipsey, Mark W., “The Primary Factors That Characterize Effective Interventions with Juvenile
Offenders: A Meta-Analytic Overview,” Victims and Offenders, Vol. 4, 2009, pp. 124–147.
Lipton, D. S., R. Martinson, and J. Wilks, The Effectiveness of Correctional Treatment: A Survey
of Treatment Valuation Studies, New York: Praeger Press, 1975.
*Loadman, William E., Raeal J. Moore, Weijia Ren, Jing Zhu, Jing Zhao, and Richard G.
Lomax, Striving Readers Year 5 Project Evaluation Report: Ohio: An Addendum to the Year
4 Report, Columbus, Ohio: The Ohio State University, December, 2011.
Lockwood, Susan R., John M. Nally, John Dowdell, Jerry McGlone, Steve Steurer,
“Implementing the 2014 GED Exam and Computer-Based GED Testing in Correctional
Facilities: A Guide for Correctional Educators and Administrators,” Journal of Correctional
Education, Vol. 64, No. 2, May 2013.
MacKenzie, D., What Works in Corrections: Reducing the Criminal Activities of Offenders and
Delinquents, New York: Cambridge University Press, 2006.
MacKenzie, D. L., “Structure and Components of Successful Education Programs,” paper
presented at the Reentry Roundtable on Education, John Jay College of Criminal Justice,
New York, 2008.
Malmgren, K. W., and Peter E. Leone, “Effects of a Short-Term Auxiliary Reading Program on
the Reading Skills of Incarcerated Youth,” Education and Treatment of Children, Vol. 23,
No. 3, 2000, pp. 239–247.
Marks, A. (1997). “One Inmate’s Push to Restore Education Funds for Prisoners,” Christian
Science Monitor, 89, 3.
Martorell, Francisco, and Issac Jr. McFarlin, “Help or Hindrance? The Effects of College
Remediation on Academic and Labor Market Outcomes,” Review of Economics and
Statistics, Vol. 39, No. 2, May 2011.
Maruschak, L. M., HIV in Prisons, 2001–2010, Washington, D.C.: Bureau of Justice Statistics,
NCJ 238877, September 2012.

136

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

*Mayer, E. E. and R. A Hoffman, “A Comparison of the Effects of Individualized Instruction
with Group Instruction on the Academic Achievement and Self-Concept of Youthful
Offenders,” Journal of Correctional Education, Vol. 13, 1982, pp. 11–13.
McGraw Hill Education, “SRA Corrective Reading,” web page, 2013. As of February 11, 2014:
http://www.mcgraw-hill.co.uk/sra/correctivereading.htm
Mears, Daniel P., and Laudan Y. Aron, Addressing the Needs of Youth with Disabilities in the
Juvenile Justice System: The Current State of Knowledge, Washington, D.C.: Urban Institute
Justice Policy Center, November 2003. As of February 3, 2014:
http://www.urban.org/UploadedPDF/410885_youth_with_disabilities.pdf
Meisel, S., P. Leone, K. Henderson, and M. Cohen, Collaborate To Educate: Special Education
in Juvenile Correctional Facilities, College Park, Md.: The National Center on Education,
Disability, and Juvenile Justice, 1998.
Murph, D., and S McCormick, “Evaluation of an Instructional Program Designed to Teach
Minimally Literate Juvenile Delinquents to Read Road Signs,” Education and Treatment of
Children, Vol. 8, 1985, pp. 133–151.
Muse, Frederic M., “A Look at the Benefits of Individualized Instruction in a Juvenile Training
School Setting: How Continuous Progress Accelerates Student Performance,” Journal of
Correctional Education, Vol. 49, No. 2, 1998, pp. 73–80.
*National Council on Crime and Delinquency, In Search of Evidence-Based Practice in Juvenile
Corrections: An Evaluation of Florida’s Avon Park Youth Academy and Street Smart
Program, Madison, Wisc.: National Council on Crime and Delinquency, 2009.
National Governors Association and the National Association of State Budget Officers, The
Fiscal Survey of States: Fall, 2010. Washington, D.C.: National Association of State Budget
Officers, 2010. As of February 3, 2014:
http://www.nga.org/Files/pdf/FSS1012.PDF
National Juvenile Justice Network, “Keep Youth out of Adult Courts, Jails, and Prisons,” 2011.
As of February 3, 2014:
http://www.njjn.org/about-us/keep-youth-out-of-adult-prisons
NBC News and the Hechinger Report, Carpe Diem: Seize the Digital Revolution, 2013. As of
February 3, 2014:
http://www.educationnation.com/casestudies/carpediem/NBCCaseStudy_CarpeDiem.pdf
Neild, Ruth Curran, C. Boccanfuso, and V. Byrnes, The Academic Impacts of Career and
Technical Schools: A Case Study of a Large Urban School District, Baltimore, Md.: Center
for the Social Organization of Schools, Johns Hopkins University, 2013.

137

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

NGA/NSBO—See National Governors Association and the National Association of State Budget
Officers.
Office of Juvenile Justice and Delinquency Prevention, “Statistical Briefing Book,” 2013. As of
February 3, 2014:
http://www.ojjdp.gov/ojstatbb/corrections/
Pager, D., “The Mark of a Criminal Record,” American Journal of Sociology, Vol. 108, No. 5,
2003, pp. 937–975.
Pasternak, Robert, Robert Portillos, and Henry Hoff, “Providing an Appropriate Education to
Adjudicated and Incarcerated Juvenile Delinquents: The Challenge to Correctional Education
Administrators,” Journal of Correctional Education, Vol. 39, No. 4, December, 1988, pp.
154–159.
Pew Center on the States, State of Recidivism: The Revolving Door of America’s Prisons,
Washington, D.C.: The Pew Charitable Trusts, April 2011.
Platt, J. M., and M Beech, “The Effectiveness of Learning Strategies in Improving Performance
and Increasing the Independence of Juvenile Offenders with Learning Problems,” Journal of
Correctional Education, Vol. 45, 1994, pp. 18–29.
President’s Commission on Law Enforcement and the Administration of Justice, The Challenge
of Crime in a Free Society: A Report by the President’s Commission on Law Enforcement
and the Administration of Justice, February 1967. As of February 3, 2014:
https://www.ncjrs.gov/pdffiles1/nij/42.pdf
Priest, Nora, Antonia Rudenstine, Ephraim Weisstein, and Carol Gerwin, Making Mastery Work:
A Close-up View of Competency Education, Quincy, Mass. Nellie Mae Education
Foundation, November, 2012.
Proctor, C. M., Jr., and P. Johnson, Computational Skills Development Kit—Teacher’s Guide,
Chicago: Science Research Associates, Inc., 1965.
Project READ, To Make a Difference, Silver Spring, Md.: National Criminal Justice Reference
Service, 1978. As of February 3, 2014:
https://www.ncjrs.gov/pdffiles1/Digitization/49939NCJRS.pdf
Public Law 89-10., Elementary and Secondary Education Act, 1965.
Public Law 89-329, Higher Education Act of 1965, November 8, 1965.
Public Law 101-476, Individuals with Disabilities Education Act, October 30, 1990.
Public Law 103-322, Violent Crime Control and Law Enforcement Act, September 13, 1994.
Public Law 105-220, Workforce Investment Act, August 7, 1998.

138

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Public Law 110-199, Second Chance Act of 2007, April 9, 2008.
Raphael, S., “The Employment Prospects of Ex-Offenders,” Focus, Vol. 25, No. 2, Fall–Winter
2007–2008.
Ryan, T. A., and K. A. McCabe, “Mandatory vs Voluntary Prison Education and Academic
Achievement,” The Prison Journal, Vol. 74, 1994, pp. 450–461.
*Roos, Leendert E., The Effect of Career Development on Employment and Recidivism Among
Juvenile Offenders, Walden University, 2006.
Rosenbaum, Paul R. and Donald B. Rubin, “The Central Role of the Propensity Score in
Observational Studies for Causal Effects,” Biometrika, Vol. 70, No. 1, 1983, pp. 41–55.
Sander, Janay B., Erika A. Patall, Laura A. Amoscato, Alexandra Fisher, and Catherine Funk, “A
Meta-Analysis of the Effect of Junveile Delinquency Interventions on Academic Outcomes,”
Children and Youth Services Review, Vol. 34, No. 9, 2012, pp. 1695–1708.
*Scarlato, Mary C. and Erin Asahara, “Effects of “Corrective Reading” in a Residential
Treatment Facility for Adjudicated Youth,” Journal of Direct Instruction, Vol. 4, No. 2,
2004, pp. 211–217.
Scholastic, “Scheduling options for Scholastic Read 180 implementations,” n.d. As of January
24:
http://teacher.scholastic.com/products/statehomepages/uploadfile/resource/fl_read180_sched
ulingoptions.pdf
Scientific Learning Corporation, “Improved Reading Skills by Students in the Virginia
Department of Correctional Education Who Used Fast Forword Products,” MAPS for
Learning: Educator Reports, Vol. 8, No. 28, 2004, pp. 1–5.
Scott-Clayton, Judith and Olga Rodriguez, Development, Discouragement, or Diversion? New
Evidence on the Effects of College Remediation, Cambridge, Mass.: National Bureau of
Economic Research, August, 2012. As of February 3, 2014:
http://www.nber.org/papers/w18328.
Sedlak, A. J., and K. S. McPherson, “Youth’s Needs and Services: Findings from the Survey of
Youth in Residential Placement,” in Juvenile Justice Bulletin, Washington, D.C.: Office of
Juvenile Justice and Delinquency Prevention, Office of Justice Programs, U.S. Department of
Justice, 2010.
Shadish, William R., M. H. Clark, and Peter M. Steiner, “Can Nonrandomized Experiments
Yield Accurate Answers? A Randomized Experiment Comparing Random to Nonrandom
Assignment,” Journal of the American Statistical Association, Vol. 103, No. 484, December
2008, pp. 1334–1356.

139

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Sherman, L. W., D. C. Gottfredeson, D. L. MacKenzie, J. Eck, P. Reuter, and S. Bushway,
Preventing Crime: What Works, What Doesn’t, What’s Promising, Washington, D.C.: U.S.
Office of Justice Programs, 1997.
*Shippen, Margaret E., Rhonda Collins Morton, Samuel W. Flynt, David E. Houchins, and
Tracy Smitherman, “Efficacy of a Computer-Based Program on Acquisition of Reading
Skills of Incarcerated Youth,” Remedial and Special Education, Vol. 33, No. 1, 2012, pp.
14–22.
Sickmund, Melissa, Anthony J. Sladky, W. Kang, and Charles Puzzanchera, “Easy Access to the
Census of Juveniles in Residential Placement: 1997–2011,” Office of Juvenile Justice and
Delinquency Prevention, U.S. Department of Justice, 2013. As of February 3, 2014:
http://www.ojjdp.gov/ojstatbb/ezacjrp/asp/selection.asp
*Simpson, S. B., J. M. Swanson, and K Kunkel, “The Impact of an Intensive Multisensory
Reading Program on a Population of Learning Disabled Delinquents,” Annals of Dyslexia,
Vol. 42, 1992, pp. 54–66.
Sinatra, Richard, “Use of Three Writing Tasks at an Adolescent Treatment Center,” Perceptual
and Motor Skills, Vol. 59, 1984, pp. 355–358.
*Skonovd, Norman, Wesley Krause, and L. Armstrong Troy, “The Regional Youth Educational
Facility: A Promising Short-Term Intensive Institutional and Aftercare Program for Juvenile
Court Wards,” 1991, pp. 395–422.
Slavin, R. E.,”Meta-Analysis in Education: How Has It Been Used?” Educational Researcher,
Vol. 13, No. 8, 1984, pp. 6–15.
Slavin, Robert E., Alan Cheung, Cynthia Groff, and Cynthia Lake, “Effective Reading Programs
for Middle and High Schools: A Best-Evidence Synthesis,” Reading Research Quarterly,
Vol. 43, No. 3, July/August/September, 2008, pp. 290–322.
Snyder, Howard N., and Melissa Sickmund, Juvenile Offenders and Victims: 2006 National
Report, Washington, D.C.: Office of Juvenile Justice and Delinquency Prevention, Office of
Justice Programs, U.S. Department of Justice, 2006,
http://files.eric.ed.gov/fulltext/ED495786.pdf.
Soe, K., S. Koki, and J. Chang, “Effect of Computer-Assisted Instruction on Reading
Achievement: A Meta-Analysis,” 2000. As of February 3, 2014:
http://nichcy.org/research/summaries/abstract12
Steele, Jennifer L., Matthew W. Lewis, Lucrecia Santibañez, Mollie Rudnick, Brian M. Stecher,
Laura Hamilton, and Susannah Faxon-Mills, “Proficiency-Based Pathways in Three Pilot
Programs: Examining Implementation and Outcomes,” paper presented at Society for
Research on Educational Effectiveness, Washington, D.C., September 2013.
140

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

Stephan, J. J., Census of State and Federal Correctional Facilities, 2005, Washington, D.C.:
Bureau of Justice Statistics, National Prisoner Statistics Program, NCJ 222182, October
2008.
Steurer, S. J., L. G. Smith, and A. Tracy, Education Reduces Crime: Three-State Recidivism
Study, Lanham, Md.: Correctional Education Association, 2003.
Students at the Center, homepage, 2013. As of February 3, 2014:
http://www.studentsatthecenter.org/
Sturgis, Christine and Susan Patrick, When Success Is the Only Option: Designing CompetencyBased Pathways for Next Generation Learning, Vienna, Va.: International Association for
K–12 Online Learning, 2010.
Sturgis, Christine, Bob Rath, Ephraim Weisstein, and Susan Patrick, Clearing the Path: Creating
Innovation Space for Serving over-Age, under-Credited Students in Competency-Based
Pathways, Vienna, Va.: International Association for K–12 Online Learning, 2010. As of
February 3, 2014:
http://net.educause.edu/ir/library/pdf/CSD6173.pdf
Taylor, J. M., “Alternative Funding Options for Post-Secondary Correctional Education: Part
One,” Journal of Correctional Education, Vol. 56, 2005, pp. 6–17.
Torgesen, J., D. Myers, A. Schirm, E. Stuart, S. Vartivarian, W. Mansfield, F. Stancavage, D.
Durno, R. Javorsky, and C. Haan, National Assessment of Title I Interim Report—Volume II:
Closing the Reading Gap: First Year Findings from a Randomized Trial of Four Reading
Interventions for Striving Readers, Washington, D.C.: Institute of Education Sciences, U.S.
Department of Education, 2006. As of February 3, 2014:
http://www.ed.gov/rschstat/eval/disadv/title1interimreport/index.html
Tyler, John H. and Jeffrey R. Kling, “Prison-Based Education and Re-Entry into the Mainstream
Labor Market,” in Shawn Bushway, Michael Stoll, and David Weiman, eds., Barriers to
Reentry? The Labor Market for Released Prisoners in Post-Industrial America, New York:
Russell Sage Foundation Press, 2007, pp. 227-256.
Tyler, John H., Richard J. Murnane, and John B. Willett, “Estimating the Labor Market
Signaling Value of the Ged,” Quarterly Journal of Economics, Vol. 115, No. 2, May, 2000,
pp. 431–468.
U.S. Census Bureau, “Current Population Survey, Annual Social and Economic Supplement,”
2011. As of February 3, 2014:
https://www.census.gov/hhes/www/poverty/publications/pubs-cps.html
———, “State and County Quick Facts: USA,” 2013. As of February 3, 2014:
http://quickfacts.census.gov/qfd/states/00000.html
141

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

U.S. Department of Education, “U.S. Departments of Education and Justice Release School
Discipline Guidance Package to Enhance School Climate and Improve School Discipline
Policies/Practices,” January 8, 2014. As of February 3, 2014:
http://www.ed.gov/news/press-releases/us-departments-education-and-justice-release-schooldiscipline-guidance-package———, “Building the Legacy: Idea 2004,” 2013a. As of February 3, 2014:
http://idea.ed.gov/explore/home
———, “Competency-Based Assessment at Young Women’s Leadership Charter School,” 2010.
As of February 3, 2014:
http://www.ed.gov/technology/draft-netp-2010/competency-based-assessment
———, “Competency-Based Learning or Personalized Learning,” 2011. As of February 3, 2014:
http://www.ed.gov/oii-news/competency-based-learning-or-personalized-learning
———, National Assessment of Career and Technical Education: Interim Report, Washington,
D.C.: Policy and Program Studies Service, Office of Planning, Evaluation, and Policy
Development, U.S. Department of Education, 2013b. As of February 3, 2014:
http://www2.ed.gov/rschstat/eval/sectech/nacte/career-technical-education/interim-report.pdf
———, Title I, Part D: Neglected, Delinquent, and At-Risk Youth Prevention and Intervention
Programs for Children and Youth Who Are Neglected, Delinquent, or At-Risk (N or D),
Nonregulatory Guidance, Washington, D.C.: Policy and Program Studies Service, Office of
Planning, Evaluation, and Policy Development, U.S. Department of Education, 2006.
Wells, R. E., “Education as Prison Reform: A Meta-Analysis,” unpublished dissertation, Baton
Route, La.: Louisiana State University, 2000.
Wexler, Jade, Nicole Pyle, Andrea Flower, Jacob L. Williams, and Heather Cole, “A Synthesis
of Academic Interventions for Incarcerated Adolescents,” Review of Educational Research,
2013.
What Works Clearinghouse, Corrective Reading (WWC Intervention Report), Washington, D.C.:
Institute of Education Sciences, U.S. Department of Education, 2007. As of February 3,
2014:
http://ies.ed.gov/ncee/wwc/pdf/intervention_reports/WWC_Corrective_Reading_070207.pdf
———, Fast Forword (WWC Intervention Report), Washington, D.C.: Institute of Education
Sciences, U.S. Department of Education, October, 2013. As of February 3, 2014:
http://ies.ed.gov/ncee/wWc/pdf/intervention_reports/wwc_ffw_031913.pdf
———, Procedures and Standards Handbook (Version 2.0), Washington, D.C.: Institute of
Education Sciences, U.S. Department of Education, December, 2008.

142

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.

*Wilson, Paul R., “Recidivism and Vocational Education,” Journal of Correctional Education,
Vol. 45, No. 4, 1994, pp. 158–163.
Wilson, D. B., C. A. Gallagher, and D. L. Mackenzie, “A Meta-Analysis of Corrections-Based
Education, Vocation, and Work Programs for Adult Offenders,” Journal of Research in
Crime and Delinquency, Vol. 37, No. 4, 2000, pp. 347–368.
*Zhu, Jing, William E. Loadman, Richard G. Lomax, and Raeal Moore, Evaluating Intervention
Effects of Scholastic Read 180 on Low-Achieving Incarcerated Youth: Society for Research
on Educational Effectiveness, March, 2010. As of February 3, 2014:
http://search.ebscohost.com/login.aspx?direct=true&db=eric&AN=ED514404&site=ehostlive

143

PRE-PUBLICATION VERSION: This document has been peer-reviewed and edited but has not yet been finalized.
Typeset and fully designed print and online volumes forthcoming.