Pinnell, G. S., Lyons, C. A., DeFord, D. E., Bryk, A., & Seltzer, N. (1994). Reading Research Quarterly, 29, 8-39.

Background
Pinnell, Lyons, DeFord, Bryk, and Seltzer’s study systematically compared Reading Recovery to three other instructional models of early intervention. In this study (N = 324) the lowest-achieving first-grade students from 40 different schools in 10 different school districts were randomly assigned within schools to one of five groups.

1. Reading Recovery
2. a Reading Recovery-like intervention with partially trained teachers
3. a skills-based individual intervention
4. small group instruction offered by Reading Recovery teachers
5. a control group

Measures included those used in Reading Recovery as well as generally known reading tests (Gates-MacGinitie Reading Test and Woodcock Reading Mastery). The study employed a formal experimental design that used split plots to control effects that may result from differing cultures among school districts or individual schools. The difficulty of small standard errors in analysis of data at the student level was addressed by using the Hierarchical Linear Model for data analysis.

Researchers at the University of Chicago independently analyzed the data. In addition, a renowned national panel of researchers not involved in Reading Recovery provided oversight for analyzing results.

Findings
The results of the study were definitive: Reading Recovery subjects performed significantly better than any other treatment and comparison group on all measures. Essential differences were related to individual instruction, the lesson framework (combination of techniques), and teacher training.

For more information:
Download Six Reading Recovery Studies: Meeting the Criteria for Scientifically Based Research (PDF)

This abstract first appeared in What Evidence Says About Reading Recovery (2002). Columbus, OH: Reading Recovery Council of North America.