A pilot study of the efficacy of two adult basic literacy programs for incarcerated males.

A Pilot Study of the Efficacy of Two Adult Basic Literacy Programs for Incarcerated Males

In the United States at the end of 2005, more than 7 million adults were in prison, in jail, on probation, or on parole (Bureau of Justice Statistics, 2007). Over 1.4 million are incarcerated in state and federal prisons at any given time (Hrabowski & Robbi, 2002). According to Vacca (2004), more than half of the adults incarcerated in the U.S. can neither read nor write and have less than an 8th grade education. In Alabama, where the study took place, approximately 75% of all state inmates read below the 5th grade level (J. Hopper, personal communication, August 22, 2006) and 60% have not completed high school (ADOC, 2006). With these staggering statistics illustrating the failed educational experiences of incarcerated adults in Alabama and the clear link between increased education and reduced recidivism, implementing effective programs to improve literacy is imperative (Jancic, 1998: Vacca).

According to the Bureau of Justice Statistics (2007), 41% of all inmates in federal and state prison did not complete high school as compared to 18% of the general population. Additionally, over one third of incarcerated school dropouts reported that they quit primarily due to behavioral and academic problems in school (Hrabowski & Robbi, 2002). Women who are incarcerated are more likely to have completed high school than incarcerated men, 30% versus 25%, respectively (Bureau of Justice Statistics).

Literacy and Reading

Literacy has been defined as the ability to read, write, spell, listen, and speak in ways that enable communication, promote understanding of ideas, and enrich lives (Moats, 2000; Glaeser, Lenz, Gildroy, & McKnight, 2000). Literacy may the most important educational goal for which educators are responsible. Academic success, employment, and personal health depend upon an individual's ability to use and understand their culture's language system. Moreover, the National Institute of Student Health and Human Development views literacy and reading failure in the United States as a national health crisis (Lyon, 1999).

Literacy development is the combination of an individual's developmental processes and language and life experiences (Glaeser et al., 2000). Literacy and language skills develop over many years from birth through adulthood. While most individuals acquire language in a natural, developmental manner, their ability to acquire basic reading skills is not a natural process (Lyon, 1999; Moats, 2000). More importantly, the total act of reading is affected if students are weak in just one reading skill (Chall, Jacobs, & Baldwin, 1991).

Limited reading practice limits a person's vocabulary knowledge and comprehension that results in poor academic achievement and limited literacy skills. Stanovich (1986) borrowed the biblical concept, "Matthew effects," that describes a rich-get-richer and poor-get-poorer phenomenon when applied to reading and literacy. In other words, those individuals who have limited access to the meaning of printed words fall behind their peers with average to above reading skills, resulting in an increasingly widening performance gap as they progress through school and life. Thus, reading holds cognitive consequences that limit one's quality of life. Therefore, the purpose of this study was to examine the differential effects of two adult literacy programs for incarcerated males.

Method

Setting and Participants

This study took place in a medium security male prison in Alabama. Medium security prisons house inmates that (a) have not committed capital crimes, (b) are not on death row, and (c) are eligible for parole. Inmates enrolled in an adult basic literacy program (n = 60) were the original sample participating in this study. Due to attrition from release, transfers, and dropping out of the program, a final sample of participants (n = 27) included fourteen students and thirteen tutors. Reading grade equivalent pre test scores of final sample of participants ranged from 1.2 to 16.9, with a mean pre test reading grade equivalent of 7.9. The mean pre-test short scale grade equivalent for students was 3.7 and the mean for tutors was 13.8.

Sixty percent of the participants were African American and 40% European American. Participants' age range was 25 to 64 with a mean age of 38. Forty percent of the participants were using the Laubach Literacy program and sixty percent were using the Direct Instruction Reading Decoding Program. All participation in the study was voluntary.

Materials and Procedures

Each participant completed a pre and post individually administered standardized reading assessment, the Woodcock Reading Mastery Tests Revised/NU (Woodcock, 1998) (WRMT-R/NU). The WRMT-R/NU is a battery of standardized norm referenced tests designed to measure the reading achievement level of individuals age 5 to 75. The WRMT-R/NU has two alternate forms (G & H). According to the examiner's manual, the standard scores yielded by the WRMTR/ NU are based on a mean of 100 and a standard deviation of 15. The WRMTR/ NU consists of six tests including (a) Visual-Auditory Learning, (b) Letter Identification, (c) Word Identification, (d) Word Attack, (e) Word Comprehension, and (f) Passage Comprehension. Administration of all six subtests yields a Total Reading-Full Scale score.

The WRMT-R/NU also has a short scale consisting of three subtests which are (a) Word Identification, (b) Word Attack, and (c) Passage Comprehension. These subtests yield a Basic Skills Cluster score (e.g., Word Identification and Word Attack) and a Total Reading-Short Scale score (e.g., Word Identification and Passage Comprehension). Only these three subtests for the Total Reading-Short Scale were used in this study to yield an overall reading score for each participant. The WRMT-R/NU was administered in the visitation room at the facility. The test was administered by five trained examiners as prescribed in the examiner's manual and was counterbalanced. After completion of preintervention WRMT-R/NU assessment, students were randomly assigned to one of two treatments.

Reading Program Descriptions

Direct Instruction (DI) Corrective Reading Decoding (Englemann & Carnine, 1982) and Laubach Literacy (Pro Literacy America, 2007) were the two treatments used in this study. These programs are designed for adolescents and adults who have difficulty in basic reading decoding and reading comprehension skills. Each program takes an explicit approach to teaching reading meaning that the tutor explicitly teaches sounds of letters, rules of grammar, and comprehension skills to the student.

The DI model emphasizes fast-paced, scripted, well-sequenced, rule-based, and highly focused lessons (Swanson, Hoskyn, & Lee, 1999). Students in DI classes are usually instructed in small groups and given several opportunities to respond in unison and individually, with immediate feedback using a specific correction procedure. Tutors using DI generally employ a three step instructional sequence. They model (provide the correct response), lead (have student say the correct answer with the teacher), and test (give immediate feedback and a delayed probe on the task initially attempted) (Engelmann & Carnine, 1982).

The Laubach Literacy method of teaching reading also uses a scripted and explicit approach, but initially uses illustrations that are faded over time (Pro Literacy America, 2007). This method of adult basic literacy was developed by Dr. Frank Laubach in the early 1900's and used to combat illiteracy worldwide as part of his evangelical Christian mission (Wikipedia, 2007). Laubach coined the phrase "each one teach one" to promote the cause for literacy.

Student/Tutor Selection and Training

The Chaplain at the prison initially identified 25 DI and 25 Laubach volunteer tutors from a pool of applications submitted by inmates at the facility. These volunteer tutors were assessed using the WRMT-R/NU to determine if they were reading above the 7th grade level as criteria for tutoring. Students volunteers were also assessed using the WRMT-R/NU to determine if they were reading below the 5th grade level. After determining the tutor and student reading level, then they were randomly assigned to DI or Laubach. Tutors were trained to deliver reading instruction using DI or Laubach method based on the method to which they were assigned.

The author trained 25 tutors to use the Corrective Reading Decoding Program in one six hour session. Training was provided for all 4 types of exercises within the program (e.g., word attack, story reading, reading checkouts, and workbook. Four critical teaching behaviors were emphasized throughout the training. These teaching behaviors included (1) following the script, (2) error corrections, (3) firming up, and (4) pacing. During the training session, the author modeled each of the four types of exercises and after each model the trainers asked the trainees to model the exercise. If the tutor trainee modeled the exercise incorrectly, the trainer immediately corrected the error and modeled the correct teacher behavior.

The Executive Director of the Central Alabama Laubach Literacy Council conducted training for the 25 Laubach Literacy tutors. Training was conducted in one day and as with the DI training tutors had to demonstrate mastery of the Laubach method before completing the training. During each month of the six month implementation, the author observed tutoring sessions once a month and held three discussion groups with participants to gain feedback on the progress of the reading program. Also during the first two months, tutors received feedback on the fidelity of the program implementation.

Description of Tutoring Sessions

Over a six month period, DI and Laubach one-to-one tutoring sessions took place twice a week, every week, for 1.5 hours per session. For each group, DI or Laubach, this is a total of 12 hours of instruction per month. Tutoring took place in the Chapel of the facility under the oversight of the Chaplain and his trustees. Inmates reported during the discussion groups with the author while the intervention was being implemented, that tutoring was taking place on a regular basis as designed. The Chaplain confirmed this consistency as well.

Data Analysis and Results

Descriptive and analytical data were collected. Descriptive data include mean reading pre and post scores in word identification, word attack, reading comprehension, basic skills cluster (a combination score of the word attack and word identification subtests) as well as the overall reading grade equivalents yielded in the form of a Reading Short Scale score. Over one year of grade equivalent growth was made in all areas assessed after a six month intervention. See table 1 for mean descriptive grade equivalent growth data.

A paired samples t test and effect size calculations were conducted to determine if significant gains had been made in overall reading skills and sub skills, with the exception of word attack skills, from pretest to posttest test after the six month intervention, irrespective of program type. The results indicate that neither program was superior, but that participants made significant grains in reading grade equivalent growth and neared moderate effect size. The t test was highly significant for the total reading short scale, t(26) = 3.21, p < .01, which is the comprehensive measure of reading skills. The effect size overall neared medium effect, d = .23. Effect size is provided below for each area measured.

The t test results for the overall group indicated a significant mean grade equivalent growth for comprehension skills (M = 9.1. SD = 5.18) with a medium effect size d = .30, a non-significant mean grade equivalent growth for word attack skills (M = 10.3, SD = 8.8) with a medium effect size d = .30, a significant mean grade equivalent growth for word identification skills (M = 9.4, SD = 5.6) with a near medium effect size d = .25, a significant mean grade equivalent growth for the basic skills cluster (a combination of word attack and word identification) (M = 9.4, SD = 5.8) with a near medium effect size d = .24, and a significant mean grade equivalent growth for total reading skills (M = 9.2, SD = 5.4). The only area that did not show significant mean grade equivalent growth was word attack skills.

A second and third t test was conducted to determine growth within the student group and within the tutor group. The only significant area of growth for the tutors was word identification (M = 14.5, SD = 2.9), t (12) = 2.92, p < .05, with a large effect size d = .72. All other areas were non-significant for tutors. For students, significant growth was noted in every area except word attack. A significant mean grade equivalent growth for comprehension skills (M = 2.0. SD = 1.9) with a large effect size d = .74, a non-significant mean grade equivalent growth for word attack skills (M = 4.1, SD = 10.6)) with a medium effect size d = .36, a significant mean grade equivalent growth for word identification skills (M = .75, SD = .71) with a medium effect size d = .35, a significant mean grade equivalent growth for the basic skills cluster (a combination of word attack and word identification) (M = 1.3, SD = 2.4) with a small effect size d = .20, and a significant mean grade equivalent growth for total reading skills (M = 1.1, SD = .94) with a large effect size d = .57. Regardless of tutor or student status, the intervention demonstrated significant growth in at least one area for all participants. See tables 2 and 3, on the following page, for tutor and student grade equivalent gains.

Discussion

The purpose of this pilot study was to compare the effects of two adult literacy programs for incarcerated males. Results indicate that all participants made progress in at least one area of reading skill, irrespective of program or student/tutor status. With such a small sample these findings must be interpreted with caution, but nevertheless, they may be helpful in planning future studies in the literacy skills of incarcerated individuals.

This study demonstrates that with effective intervention, low literate individuals can acquire reading skills within a structured program. Other researchers (Hrabowski & Robbi, 2002; Jancic, 1998; Vacca, 2004) have noted the clear link between effective correctional education programs and reduced recidivism rates. By addressing basic literacy skills to improve the likelihood that incarcerated individuals will pursue adult education programs leading to a GED, may be a missing link in the prison education curriculum. In this study, the student participants clearly benefited from the reading instruction. In fact, anecdotally, four of students reached the sixth grade reading level and will be entering the GED program at the facility.

After release from prison, the negative impact of poor reading achievement as an adult is well documented. Kirsch, Jungeblut, Jenkins, and Kolstad (1993) report that those with lower reading levels are more likely to be unemployed, work fewer weeks during the year, and earn lower wages as compared to those with higher reading levels. Further findings by these authors indicate that almost half of all adults with the lowest level of reading literacy live in poverty. Additionally, individuals with lower reading levels have reduced economic bargaining power, make less money, and have fewer career choices. Combine these grim outcomes with a history of incarceration and one need not wonder why recidivism is a chronic issue.

People Power as Human Capital

The data presented in this pilot study also speak to the power of using the "human capital" of literate inmates to increase the reading level of low literate inmates. The opportunity to serve others while incarcerated may hold significant social implications. That is, being in a place of confinement and being able to give to others, while juxtaposed, is according to the participants in this study a "powerful" opportunity. Many of the tutors were anxious to learn of the growth of their students. These data have been shared with the institution and the participants and more inmates have been recruited into the basic literacy program.

The data presented in this study also indicate that both tutors and students gained reading skills. After a six month intervention students gained a mean of over one year in reading grade equivalent as did tutors. However, as indicated in Table 3, students in this study on average still function below the 6th grade level which precludes them from entering the GED program. Development and implementation of tutor/student literacy programs will support the growth of low literate prisoners.

The final area of human capital that must be noted as vital in the completion of this reading tutoring program was the unwavering support of the Warden and Chaplain at the facility where the study took place. Their collect active involvement was critical to the collection of data and implementation of the program. Leadership is a key factor in success of adult education programs within the corrections system.

Limitations

This study is not without limitations. The small sample size, the fact that the data are from only a single facility implementing each model, the heterogeneity of the sample, and the relatively short time frames for data collection may limit the generalization of the findings. Clearly, more data are needed in the area of promoting literacy skills for individuals who are incarcerated. This pilot study provides a starting point for the context of adult basic literacy skills in prisons, but leaves much room for improved inquiry.

Conclusion and Future Research

This pilot study needs to be replicated on a larger scale and over a longer period of time. The author believes that differences between the reading programs may be able to be "culled out" with a larger sample and possibly a comparison at more that one institution. However, the current study demonstrates that literacy growth can and does occur when opportunities to learn are provided. Corrections must begin to address the basic literacy skills of those for whom they are responsible, so that after release, these individuals will have an increased opportunity to succeed. With these staggering statistics related to failed educational experiences of incarcerated adults, finding methods to improve any educational experience is worthwhile, especially for these populations who have historically poor outcomes.

References

Alabama Department of Corrections (2006). Statistical report from ADOC research and planning office. Montgomery, AL.

Bureau of Justice Statistics (2007). Corrections Statistics. Retrieved from http://www.ojp.usdoj.gov/bjs/correct.htm on January 5, 2007.

Chall, J. S., Jacobs, V. A., & Baldwin, L. E. (1991). The reading crisis: Why poor students fall behind. Cambridge, MA: Harvard University Press.

Engelmann, S., & Carnine, D. (1982). Theory of instruction: Principles and applications. New York: Irvington.

Glaeser, B. J., Lenz, B. K., Gildroy, P. G., & McKnight, M. (2000). The development of literacy: As reading instruction begins (Module 1, Lesson 1). In Beginning word reading [Online]. Lawrence, KS: University of Kansas, Center for Research on Learning. Available: www.onlineacademy.org

Hbrabowski, F. A., & Robbi, J. (2002). The benefits of correctional education. The Journal of Correctional Education, 53(3), 96-99.

Jancic, M. (1998). Does correctional education have an effect on recidivism? The Journal of Correctional Education, 49(4), 152-161.

Kirsch, I.S., Jungeblut, A., Jenkins, L., & Kolstad, A. (1993). Executive summary from adult literacy in America: A first look at the results of the National Adult Literacy Survey. Washington, DC: U.S. Department of Education.

Lyon, R. G. (1999). The NICHD research program in reading development, reading disorders and reading instruction. Retrieved August 25, 1999 from http://www.ncld.org/summit99/keys99-nichd.htm

Moats, L. C. (2000). Speech to print: Language essentials for teachers. Baltimore, MD: Paul H. Brookes Publishing.

Pro Literacy America (2007). Laubach Literacy. Retrieved from http://www.proliteracy.org/about/founders.asp, July 12, 2007.

Stanovich, K. E. (1986). Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quarterly, 21, 361-406

Swanson, H., Hoskyn, M., & Lee, C. (1999). Interventions for students with learning disabilities: A meta-analysis of treatment outcomes. New York, NY: Gilford Press.

Vacca, J. S. (2004). Educated prisoners are less likely to return to prison. The Journal of Correctional Education, 55(4), 297-305.

Wikepedia (2007). Retrieved from http://en.wikipedia.org/wiki/Frank_Laubach. July 15, 2007.

Woodcock, R. W. (1998). Woodcock Reading Mastery Tests-Revised/Normative Update: Examiner's manual. Circle Pines, MN: American Guidance Services.

Table 1

Overall grade equivalent (GE) descriptive data for gain scores
and effect size

Area Assessed by            Pre GE   Post GE   GE Gain   Effect
WRMT-R/NU                    Mean     Mean      Mean      Size

Comprehension                7.5       9.1       1.6      .30
Word Attack                  7.6      10.3       2.6      .30
Word Identification          7.9       9.4       1.5      .25
Basic Skills Cluster         8.0       9.4       1.4      .24
Total Reading Short Scale    7.9       9.2       2.2      .23

Table 2
Grade equivalent (GE) descriptive data for gain scores of tutors

Area Assessed by            Pre GE   Post GE   GE Gain
WRMT-R/NU                    Mean     Mean      Mean

Comprehension                11.8     12.9       1.1
Word Attack                  12.0     13.3       1.3
Word Identification          12.4     14.5       2.1
Basic Skills Cluster         12.8     14.1       1.2
Total Reading Short Scale    12.4     13.8       1.4

Table 3

Grade equivalent (GE) descriptive data for gain scores of students

Area Assessed by            Pre GE   Post GE   GE Gain   Effect
WRMT-R/NU                    Mean     Mean      Mean      Size

Comprehension                3.6       5.6       2.0      .74
Word Attack                  3.5       7.6       4.1      .36
Word Identification          3.9       4.7       .8       .35
Basic Skills Cluster         3.6       4.9       1.3      .20
Total Reading Short Scale    3.7       4.8       1.1      .57