Success in Reading Recovery Depends on the Development of Phonological Processing Skills

James W. Chapman, William E. Tunmer and Jane E. Prochnow, Massey University, New Zealand

James W. Chapman, William E. Tunmer and Jane E. Prochnow
Massey University, New Zealand
August 1999
Revised Research Report for Phase Three of Contract ER35/199/5,
submitted to the Ministry of Education (New Zealand), August 1999.

ABSTRACT

This study examined the relationship between the development of phonological processing skills and the effectiveness of Reading Recovery (RR) in a whole language instructional context. The participants were part of an original cohort of 5-year old school entrants taking part in a New Zealand-based longituninal study of beginning literacy achievement. Four groups were formed as follows: RR group, comprising 26 children who successfully completed RR; Referred On group, comprising 6 children who failed to complete RR and were referred on for further assistance; Poor Reader Comparison group, comprising 20 poor readers who did not receive RR; and Normally Developing Group, comprising 80 average to above average readers. Reading Recovery and Poor Reader Comparison children had deficiencies in phonological processing skills during the year preceding their participation in RR, and participation in the programme did not eliminate or reduce these deficiencies. Success in RR was closely associated with phonological processing skills: children who derived modest benefits from RR significantly outperformed those who derived minimal benefits. No signs of accelerated reading performance were evident for the RR group after completion of the programme, and one year later their reading was around one year below age-appropriate levels. The RR children showed declines in reading self-concept following RR, and reported more negative perceptions of ability in reading and spelling, and in general academic self-concept, 6 and 12 months following RR. Teachers of the RR children rated them as having more classroom behaviour problems than the Normally Developing readers, and fewer adaptive functioning behaviours. We conclude that RR can be more effective in a whole language instructional context if greater emphasis is placed on the development and use of word-level skills and strategies involving phonological information.

AUTHOR’S NOTES

This study is part of the project An Investigation of Language-Related and Cognitive-Motivational Factors in Beginning Literacy Achievement, funded by the New Zealand Ministry of Education, Contract ER35/299/5, the Educational Research and Development Centre, Massey University (now the Massey University College of Education Institute for Professional Development and Research), and the Massey University Research Fund.

We are grateful to the principals and teachers of the participating schools for their willingness to participate in the study and their cooperation over the six year duration of the project. We are also grateful to Julie Russell, Debbie Bennett, Rosemary Manning, and Johanna Can Laar-veth for their professionalism and competence in collecting the data.

Correspondence concerning this report should be addressed to James W. Chapman, Department of Learning and teaching, Massey University, Private Bag 11222, Palmerston North, New Zealand. Electronic mail may be sent via Internet to J.Chapman@massey.ac.nz.

EXECUTIVE SUMMARY

Reading Recovery (RR) was developed by Marie Clay as a preventive early intervention programme designed for young children who have failed to benefit from formal reading instruction after 12 months in school. The general aim of RR is to substantially reduce the incidence of reading failure in a school system by accelerating to average levels of performance the progress of 6-year-old children who show early signs of reading difficulty (normally children whose reading progress falls in the lowest 20% of the enrolment cohort in any given school). The RR programme is in addition to the regular classroom reading programme, and in New Zealand (where the programme was developed) RR complements the whole language approach to beginning literacy instruction.

Clay (1993a) argues that children need to acquire the ability to use multiple cues to solve problems while reading and to engage in the specific strategies of cross-checking, predicting, and confirming. However, she specifically states that children should be discouraged from relying too heavily on word-level cues, and warns that children’s attention may be diverted from comprehension and an understanding of the language structure “when instruction directs students to conscious manipulations of letters, sounds, or single words” (p.321). Several researchers have rejected this approach and the theoretical model of reading upon which it is based. Indeed, it is well established that using spelling-to-sound patterns in identifying unfamiliar words provides the basic mechanism for acquiring word-specific knowledge, including knowledge of irregularly spelled words. In view of these findings, a potential criticism of the instructional philosophy of RR is that it stresses the importance of using information from many sources without recognising that skills and strategies involving phonological information are of primary importance in beginning literacy development. The lack of emphasis on the development of phonological processing skills in RR may be especially problematic in the context of classroom literacy programmes that place little emphasis on the development of these skills, such as those that follow a more naturalistic, informal “whole language” approach to reading instruction. This may be particularly true in New Zealand, where beginning readers are urged to use sentence context as the primary strategy for identifying unfamiliar words in text.

The purpose of the present study was to examine the relationship between the development of phonological processing skills and the effectiveness of the RR programme in a whole language instructional context. We hypothesized that children with deficiencies in phonological skills at school entry would encounter difficulties in learning to read, and that unless these deficiencies were overcome during regular classroom instruction or RR, the reading difficulties would continue. We further hypothesized that as a result of repeated learning failures, poor readers whose reading difficulties were not remediated would develop negative reading-related self-perceptions and, possibly as a consequence, would show higher levels of classroom behaviour problems than normally developing readers. Specifically, this study was designed to answer the following questions: (1) Do children selected from RR from a whole language instructional context show deficiencies in phonological processing skills during the year preceding entry into the programme? (2) If so, does participation in RR reduce or eliminate these deficiencies?and, (3) Is there a relationship between the development of phonological processing skills and the immediate and long-term effectiveness of the RR programme with respect to literacy development, reading-related self-perceptions, and classroom behaviour?

The children who participated in the present study were part of an original cohort of 152 new school entrants who were taking part in a larger longitudinal study of beginning literacy achievement. The children were enrolled in 16 primary schools located in Palmerston North. At the start of the study, the target children had recently turned 5 years of age (M =5 yrs 1 mo), and commenced school for the first time at the start of the 1993 school year (February). Four groups were formed as follows: RR group, comprising 26 children who successfully completed RR; Referred On group, comprising 6 children who failed to complete RR and were referred on for further assistance; Poor Reader comparison group, comprising 20 poor readers who did not receive RR; and Normally Developing Group, comprising 80 average to above average readers.

The children in the four groups were assessed on seven occasions: the beginning, middle, and end of Year 1; the middle and end of Year 2; and the middle and end of Year 3. Several reading and reading-related measures were taken, each at developmentally appropriate testing times.

The results of the study are presented in relation to each of the three research questions. In regard to the first question, the results clearly showed that both the Reading Recovery and Poor Reader Comparison children had deficiencies in phonological processing skills during the year preceding the RR programme. Findings relating to the second question showed that participation in the RR programme did not eliminate or reduce these phonological processing deficiencies. Although the RR children showed some improvement in phoneme segmentation skill, this development was considerably delayed and was not matched by significant improvements in other phonological processing skills. Rather, the RR children continued to make progress in the phonological domain that was similar in trend but persistently well below that of the ND children. results that addressed the third research question indicate that success in RR is closely associated with phonological processing skills: the children who derived modest benefits from the RR programme significantly outperformed those who derived minimal benefits. Further, RR failed to significantly improve the literacy development of children considered to have succeeded in the programme: RR children showed no signs of accelerated reading performance, and one year after completion of the programme, they were performing at around one year below age-appropriate levels. Moreover, the RR children showed declines in reading self-concept following RR, and they held more negative perceptions of ability in reading and spelling, and in general academic self-concept six and 12 months following RR. Finally, teachers of the RR children rated them as having a greater number of classroom behaviour problems relative to the ND children, and fewer adaptive functioning behaviours in the areas most closely associated with learning.

These results clearly indicate that the RR programme did not meet the goal of accelerating to average levels of reading performance the progress of 6-year old children who showed early signs of reading difficulty. Additional evidence for the failure of RR to bring children up to average reading levels was found following an examination of the reading book level data. Attaining a book level of at least 16 has been recommended before students are discontinued from the RR programme. In accordance with this recommendation, the mean book level reported by the RR teachers for the discontinued children was 16.6. However, the classroom teachers reported a mean book level of only 9.0 for the same children, with only one child attaining a level of 16. We have no explanation for this marked discrepancy.

The failure of RR to significantly improve the reading performances of children in the present study is most likely due to the instructional philosophy and practices of RR. Reading Recovery stresses the importance of using information from many sources without recognizing that skills and strategies involving phonological information are of primary importance in beginning literacy development. Yet, children in the present study who were placed in RR were clearly deficient in phonological processing skills during the year preceding entry into the programme, and during the 18 months following the programme. The failure of the RR programme to eliminate the phonological processing deficiencies of the discontinued children is not surprising because systematic instruction in word-level strategies is not a central component of RR programmes. Rather, RR instruction stresses the importance of using sentence context cues for identifying unfamiliar words in text with as little as possible letter-sound information being used to confirm language predictions. Even with the emphasis on the development of language prediction skills, however, the discontinued RR children performed poorly.

We conclude that the findings of the study do not support the philosophy of RR that reading instruction should focus on teaching children to use sentence context cues as a major strategy for identifying unfamiliar words in text. Rather, our results provide further support for Pressley’s (1998) claim that reliance on the use of semantic or syntactic cues in recognizing unfamiliar words is a highly ineffective strategy. For Reading Recovery to be more effective in a whole language intsructional context, the results of this study strongly support the view that children in RR programmes should be encouraged to look for familiar spelling patterns first, and use context to confirm hypotheses about what unfamiliar words might be based on available word-level information.