Recent research on Reading Recovery.

There are many forms of early intervention, as well as alternatives to retention and social promotion. These include multiage grouping, looping, afternoon or weekend academies, summer school, and Reading Recovery. One intervention that has a robust research base is Reading Recovery.

Kathleen Martin, Assistant Professor of Early Childhood and Elementary Education at the University of Alabama at Birmingham (UAB), has chosen and reviewed the following five research articles on Reading Recovery. Kathleen is also the principal investigator for the Early Reading First program at UAB.--JA

Current mandates in American education challenge educators to provide experiences that promote success in literacy learning for all children. The No Child Left Behind (NCLB) legislation requires school districts to document that all groups of children, including racial/ethnic minorities, English language learners, and children from low-income families, demonstrate acceptable levels of literacy achievement. At the same time, school district administrators are faced with the mushrooming enrollment of children in special education services. Traditionally, American education has invested in remediation services for children who have failed at literacy learning. An alternative approach is to address the challenge with a focus on prevention. Several recent studies suggest that one early intervention approach, Reading Recovery, may be effective in addressing current mandates and concerns.

Reading Recovery is an early literacy intervention designed to identify and support 1st-grade children struggling to learn to read and write. The goals of Reading Recovery are to reduce the number of failures in literacy learning and consequently reduce related financial costs (Clay, 1993b). The majority of children who participate in this one-to-one intervention achieve average levels of literacy learning after 12-20 weeks, and their lessons are "discontinued." A much smaller group fails to reach average classroom achievement levels and is recommended for additional services or evaluation. The recommended group's failure to achieve average performance levels after approximately 20 weeks of careful, intense, individualized instruction indicates that these children are, indeed, in need of long-term supports. Reading Recovery may prevent special education referrals for children who simply need short-term individual instruction.

READING RECOVERY AND DESCUBRIENDO LA LECTURA NATIONAL REPORT 2001-2002.

National Data Evaluation Center, Columbus, OH. A significant feature of the Reading Recovery intervention is the comprehensive data collection and evaluation included in its design. Data are collected and reported for all children who participate in the intervention for even a single day. Additionally, data are collected for a random sample of 1st-grade children in each elementary school served by Reading Recovery. The annual evaluation compares the progress of Reading Recovery children to the comparison sample, which represents average 1st-grade achievement levels. National annual research reports of Reading Recovery are publicly avail able through the National Data Evaluation Center at Ohio State University (www.ndec.us/ Documentation.asp).

The May 2003 report, containing results for the 2001-2002 school year, documents that the majority of children served in Reading Recovery (60 percent) successfully completed instruction. Since it began in 1984, Reading Recovery has served 1.15 million children in the United States. During the 2001-2002 school year, "146,000 children were served by 17,557 teachers working in 10,451 schools in 3,238... school districts" (Gomez-Bellenge, Rodgers, & Fullerton, 2003, p. 5). Annual data for 2001-2002 demonstrate that Reading Recovery children who received a full series of lessons, or who were successfully "discontinued," reached much the same levels of achievement as did the random sample group. Children who did not complete a full series of lessons, or who were not successfully "discontinued," made some progress, but did not meet the same levels of achievement as the comparison group.

Gomez-Bellenge et al. (2003) report that while outcomes did vary by ethnicity and socioeconomic status, the differences were not as great as expected. Sixty percent of all children successfully discontinued lessons; rates for Asian children were higher (67 percent), and rates for African American and Hispanic/Latino children were lower (55 percent and 54 percent, respectively). As Gomez-Bellenge et al. point out, "Although these differences are larger than one would like, they pale in comparison to many other race-based outcome comparisons" (p. 31). The outcome differences, when measured by the price that a student could afford to pay for lunch, were larger than those differences for race and ethnicity. Almost 68 percent of children who paid regular lunch prices were successfully discontinued from lessons, while 60.4 percent of children who paid reduced prices, and 52.8 percent of children who received free lunches, were successfully discontinued. Gomez-Bellenge et al. noted that a further analysis of these data revealed a compounding variable. Children who received free lunches were more likely to have incomplete or interrupted programs than children who received regular price lunches. When considering only the children who received a full series of lessons, the 15 percentage point difference between the groups was reduced to 8 percentage points. Children receiving free lunch who completed a full series of lessons were successfully discontinued at a higher rate (73.1 percent), as were full-program children paying reduced lunch prices (76.6 percent) and paying full lunch prices (81.4 percent).

CLOSING THE ACHIEVEMENT GAP IN OHIO WITH READING RECOVERY. Rodgers, E. M., & Gomez-Bellenge, F., Journal of Reading Recovery, 2003, Fall, 65-74. Previous independent research (Center, Wheldall, Freeman, Outhred, & McNaught, 1995; Iversen & Tunmer, 1993; Shanahan & Barr, 1995; Sylva & Hurry, 1996; Wasik & Slavin, 1993) has documented that Reading Recovery is effective in helping children who struggle with reading and writing to reach acceptable achievement levels. Rodgers and Gomez Bellenge investigated the potential of Reading Recovery to close achievement gaps along racial/ethnic and economic lines. They examined Reading Recovery data collected in the 2001-2002 school year to determine if: 1) gaps in achievement were present along racial/ethnic and economic lines, 2) receiving a full program (20 weeks) of Reading Recovery lessons closes existing gaps, and 3) successfully discontinued Reading Recovery children closed any existing gaps. Data were collected on 7,282 children who received Reading Recovery services, and on 1,915 first-grade students who did not receive services. Two measures were considered: Text Reading Level and Hearing and Recording Sounds in Words (a phonemic representation task). Analysis of the scores from the fall revealed similar raw scores and identical stanine scores for both African American and white comparison group children. However, the analysis of spring raw scores and stanine scores for the comparison group revealed that African American children scored, on average, one stanine level lower on both the text reading and phonemic representation task. Along racial lines, a gap in achievement was emerging. A similar gap emerged in the spring along economic lines in the comparison group.

Rodgers and Gomez-Bellenge then examined results for children who received a full series of Reading Recovery lessons, whether or not they were successfully "discontinued." When the entire treatment group was considered along racial/ ethnic lines, the gap between control group white children and African American treatment group children narrowed from a two stanine level difference that favored white children in the fall, to a one stanine level difference in the spring on the phonemic representation task. On the measure of text reading, the gap was narrowed from a three stanine difference favoring white control group students in the fall, to a one stanine difference in the spring. When socioeconomic levels were considered, a similar trend was evident. Free lunch Reading Recovery children began the year three stanine levels below the control group regular lunch children on the phonemic representation task. They ended the year two stanines lower. On text reading level, the control group regular lunch group was four stanines higher than the Reading Recovery treatment group in the fall. By spring, this gap had narrowed to a one stanine difference. Although the achievement gap was not entirely closed when the entire treatment group is considered, the gap was narrowed. When the widening gap evident in the control group is considered, this trend is encouraging.

Finally, Rodgers and Gomez-Bellenge examined results for children who had successfully completed (discontinued) their series of Reading Recovery lessons. On the phonemic representation tasks, no gaps were evident. On the measure of text reading, a gap of one stanine was still evident along both racial/ethnic and economic lines. Rodgers and Gomez-Bellenge note that all successfully discontinued Reading Recovery children were reading "the levels of texts that most students would be expected to read at the end of first grade" (p. 73). In addition, at the beginning of 1st grade, "no measurable differences were evident on the two measures," and the gaps emerged after a year in school. Rodgers and Gomez-Bellenge conclude that this finding "underscores the need for educators and policy makers to investigate complex responses to a complex problem [and that] Reading Recovery plays a significant role as part of a comprehensive literacy program" (p. 73).

DOES READING RECOVERY WORK IN KANSAS? A Retrospective Longitudinal Study of Sustained Effects. Briggs, C., & Young, B. K., Journal of Reading Recovery, 2003, Fall, 59-64. The potential of any curriculum intervention must be judged not only by the immediate success of participants, but also by whether the initial gains are maintained over time. Previous follow-up studies (e.g., Brown, Denton, Kelly, & Neal, 1999; Pinnell, 1989; Rowe, 1995; Schmitt & Gregory, 2001) have documented that most Reading Recovery participants continue to perform at average or above levels during subsequent years. Two recent studies (Askew et al., 2003; Briggs & Young, 2003) support this growing body of evidence. Briggs and Young (2003) conducted a retrospective longitudinal study of the sustained effects of Reading Recovery in Kansas. During the 2001-2002 school year, they identified 295 Kansas children who had received the Reading Recovery intervention during the 1998-1999 school year and who were still attending the same school districts in the 4th grade. Fifty-six of these children were randomly selected to form a Reading Recovery group, which included: 1) children who had successfully completed the program (discontinued); 2) children who did not have time to complete their program because the school year ended; and 3) children who finished a full program of lessons, but were not considered successful and were recommended for additional evaluation. A comparison group of 79 children were randomly selected from schools identified as similar (size, socioeconomic status, race/ethnic makeup) that did not offer the Reading Recovery intervention during the 1998-1999 school year. Both groups were given the Gates-MacGinitie Reading Test. This test reports vocabulary, comprehension, and total reading scores and is a widely accepted standardized reading assessment instrument. Data analysis revealed that when results for the successfully discontinued Reading Recovery children and the Reading Recovery children who did not have a complete program were contrasted with the results of children in the comparison group, no statistical differences were observed. Significant statistical differences were observed when the Reading Recovery children who were recommended for further evaluation/service were contrasted with the comparison group. These robust results three years after the intervention suggest that the gains made in Reading Recovery do persist. The results also support Juel's (1988) observation that children who do not reach average reading levels in the 1st grade are not at average levels in the 4th grade. Briggs and Young's results also support the suggestion that Reading Recovery can serve as an effective pre-screening for referral to special education services.

MAKING A CASE FOR PREVENTION IN EDUCATION. Askew, B. J., Kaye, E., Frasier, D. F., Mobasher, M., Anderson, N., & Rodriguez, Y. G. Literacy, Teaching, and Learning: An International Journal of Early Literacy, 2002, 6(2), 43-73. The authors carefully designed a longitudinal study of Texas Reading Recovery children. They built a conceptual framework for their study that was rooted in Caplan's (1961, 1964) and Pianta's (1990) model of prevention on three levels: primary (good classroom instruction), secondary (early intervention before failure), and tertiary (special education, grade retention and long-term remediation). Askew et al. identified 218 Reading Recovery children who successfully discontinued lessons and 244 random sample children in 45 randomly selected Texas schools. Children were evaluated in 1st grade and again in 4th grade, using the required Texas assessment (TAAS), the Gates-MacGinitie Reading Test, the Clay's Observation Survey (which includes a measure of oral text reading), and teacher reports. Due to attrition, these groups were reduced to 116 Reading Recovery children and 129 controls by the time of the final testing in Grade 4.

Initial testing confirmed that the groups were significantly different at the beginning of 1st grade, as would be expected. Children selected for Reading Recovery are typically from the bottom 20 percent of classroom literacy performers. At the end of 4th grade, the average text reading level for both groups was well within Grade 4 levels. Average stanine levels for the Gates-MacGinitie subtests were 5 for the random sample group and 4 for the Reading Recovery group. Mean scores on the TAAS writing sample slightly favored the control group (2.7) over the Reading Recovery group (2.3). One significant finding in this study was that in Grade 4, the distribution of Reading Recovery scores "moved to include more average and some high stanine scores as compared with low scores with little variation at the beginning of Grade 1" (p. 58). Askew et al. conclude that their study supports the argument that Reading Recovery is an effective secondary prevention and that it plays a promising role in "(a) closing the literacy achievement gap at the outset of schooling, (b) reducing the need for tertiary prevention and freeing up those services for those who really need them, and (c) creating a systemic plan for prevention" (pp. 60-1).

THE EFFECT OF READING RECOVERY ON SPECIAL EDUCATION REFERRALS AND PLACEMENTS. O'Connor, E. A., & Simic, 0., Psychology in the Schools, 2002,39(6),635-646. O'Connor and Simic (2002) found further evidence that Reading Recovery can play an important role in reducing the demand for tertiary level services. They compared the rates of special education referrals in 11 New York City school districts for Reading Recovery children (both successfully discontinued and recommended for further services) to those of a comparison group of children experiencing literacy difficulties, but who initially scored at slightly higher levels than the Rea ding Recovery children. The groups included 2,354 students who had a complete series of Reading Recovery lessons (RR-discontinued=1,862; RR-recommended=492) and 1,770 students in a comparison group. O'Connor and Simic found that Reading Recovery significantly reduced referrals and placements in special education. The rate of referral for the total Reading Recovery group was 9 percent, compared to 14 percent for the comparison group. When the Reading Recovery group was disaggregated (successfully discontinued; recommended), O'Connor and Simic observed that only 4 percent of the discontinued group received special education referrals, while 28 percent of those children recommended for further evaluation received referrals. They also noted that fewer Reading Recovery children were placed in special education after they were evaluated, and they concluded, "Our results confirm the diagnostic value of Reading Recovery for the RR-recommended groups, as well as demonstrate that the majority of initially poor achieving students can avoid special education as a result of early intervention" (p. 643).

Conclusions

The value of an education program is often very difficult to quantify objectively. Presently, the predominant practice in American education is to wait until a child who struggles with literacy learning is two years behind peers. Low-achieving 1st-grade readers who continue to fall behind frequently become labeled "learning-disabled in reading" and are referred to long-term special education services, or they are in remedial reading programs for their entire school careers. Disproportionately large numbers of minority culture and low-income children are represented in these groups. At first, an intense one-to-one tutoring program such as Reading Recovery may seem more expensive; when contrasted with long-term special education or remedial services, however, the total costs may compare favorably. Additionally, effective early intervention, such as that provided by Reading Recovery, has the benefit of preventing failure and the corresponding loss of confidence for the majority of its students.

References

Askew, B. J., Kaye, E., Frasier, D. F., Mobasher, M., Anderson, N., & Rodriguez, Y. G. (2003). Making a case for prevention in education. In S. Forbes & C. Briggs (Eds.), Research in Reading Recovery (Vol. 2). Portsmouth, NH: Heinemann.

Briggs, C., & Young, B. K. (2003). Does Reading Recovery work in Kansas? A retrospective longitudinal study of sustained effects. Journal of Reading Recovery, Fall 2003, 59-64.

Brown, W., Denton, E., Kelly, P., & Neal, J. (1999). Reading Recovery effectiveness: A five-year success story in San Luis Coastal Unified School District. ERS Spectrum: Journal of School Research and Information, 17(1), 3-12.

Caplan, G. (Ed.). (1961). Prevention of mental disorders in children. New York: Basic Books.

Caplan, G. (1964). Principles of preventive psychiatry. New York: Basic Books.

Center, Y., Wheldall, K., Freeman, L., Outhred, L., & McNaught, M. (1995). An experimental evaluation of Reading Recovery. Reading, Research Quarterly, 30, 240-263.

Clay, M. M. (1993a). An observation survey of early literacy achievement. Portsmouth, NH: Heinemann.

Clay, M. M. (1993b). Reading Recovery: A guidebook for teachers in training. Portsmouth, NH: Heinemann.

Gomez-Bellenge, F., Rodgers, E., & Fullerton, S. K. (2003). Reading Recovery and Descubriendo la Lectura national report 2001-2002. Columbus, OH: National Data Evaluation Center.

Hummel-Rossi, B., & Ashdown, J. (2002). The state of cost-benefit and cost-effectiveness analyses in education. Review of Educational Research, 72(1), 1-30.

Iversen, S. J., & Tunmer, W. E. (1993). Phonological processing skills and the Read in a Recovery program. Journal of Educational Psychology, 85, 112-126.

Juel, C. (1988). Learning to read and write: A longitudinal study of 54 children from first through fourth grades. Journal of Educational Psychology, 80(4), 437-447.

Pianta, R. C. (1990). Widening the debate on educational reform: Prevention as a viable alternative. Exceptional Children, 56(4), 306-313.

Rowe, K.J. (1995). Factors affecting students' progress in reading: Key findings from a longitudinal study. Literacy, Teaching and Learning: An International Journal of Early Literacy, 1(2), 57-110.

Schmitt, M. C., & Gregory, A. E. (2001, December). The impact of early interventions: Where are the children now? Paper presented at the annual meeting of the National Reading Conference, San Antonio, TX.

Shanahan, T., & Barr, R. (1995). A synthesis of research on Reading Recovery. Reading Research Quarterly, 30, 958-996.

Sylva, K., & Hurry, J. (1996). Early intervention in children with reading difficulties: An evaluation of Reading Recovery and a phonological training. Literacy, Teaching, and Learning: An International Journal of Early Literacy, 2(2), 49-68.

Wasik, B. A., & Slavin, R. E. (1993). Preventing early reading failure with one-to-one tutoring: A review of five programs. Reading Research Quarterly, 28, 179-200.