Co-Principal Investigator: Jones, Akisha

Purpose: The purpose of this project is to test the sustained efficacy of Reading Recovery, an intensive one-on-one reading instruction program for the lowest-achieving first grade students, on state test scores in third and fourth grade. Many first grade students struggle with reading and for some of these students, low literacy achievement in first grade can set them up for continued difficulty in literacy throughout elementary school and beyond. The Reading Recovery program is based on the idea that individualized, short-term, and highly-responsive instruction delivered by an expert teacher can disrupt this trajectory and allow the lowest achieving students to catch up to their peers. This study is a follow-up study to an i3-funded scale-up study of the Reading Recovery program.

Project Activities: Reading Recovery is a fully-developed program that includes 12- to 20-week cycles of daily one-to-one lessons provided by a specially-trained teacher. This study is a follow-up study to an i3-funded scale-up study designed to examine the efficacy of the Reading Recovery program. In this study, the researchers will obtain state test scores for students who participated in the original i3 study to see if the impact of Reading Recovery is sustained through third and fourth grades.

Products: The products of this project include evidence of the sustained efficacy of Reading Recovery through third and fourth grades, and peer reviewed publications.

Structured Abstract

Setting: This project will take place in more than 30 states.

Sample: Data from approximately 20,000 students from more than 1,200 schools will be collected for this study.

Intervention: Reading Recovery is a fully-developed program that uses intensive one-on-one reading instruction for the lowest-achieving first grade students in a school. These students receive 12- to 20-week cycles of daily, 30-minute, one-on-one lessons from specially trained Reading Recovery expert teachers. Lessons target phonemic awareness, phonics, vocabulary, fluency, and comprehension. Reading Recovery teachers receive specialized training that prepares them to tailor lessons to an individual student's strengths and needs. Each lesson begins with re-reading familiar books, followed by word and letter work, story composition, assembling a cut-up sentence, and previewing and reading a new book. Reading Recovery also relies on continuous collection of data to gauge student progress.

Research Design and Methods: This project will extend an analysis of student outcomes that was conducted for the original i3 scale-up grant, using a Regression Discontinuity Design (RDD). The RDD used students' performance on the Observation Survey of Early Literacy (OS) as the forcing variable. Teachers administered the OS to students in their classroom, then ranked students in order of OS scores. Beginning at the bottom of the ranked list, teachers assigned those students with the lowest scores to Reading Recovery until all open Reading Recovery slots were filled. For the current study, state test data will be collected on students in both the Reading Recovery and business-as-usual control condition.

Control Condition: In the control condition, students received standard classroom practices in place at the school.

Key Measures: Third and fourth grade state achievement tests in reading or English language arts will serve as the primary outcome variable. The research team will standardize scores on each state test to a common z-score scale. Student demographic data has already been collected during the previous i3 study and will be used for the current study as well. Finally, Reading Recovery teachers, teacher leaders, or school administrators will complete surveys for all students who participated in the original RDD study to document: the student's performance in reading in grades 2 through 4; the student's participation in supplemental reading programs and interventions; and instructional programs and curricula in first, second, and third grade reading classrooms at each school.

Data Analytic Strategy: The research team will use multilevel modeling to compare students below and above the cutoff OS score. Researchers will use separate analyses to examine third and fourth grade outcomes. The team will include a random treatment effect to examine variation between schools in the sustained impact of Reading Recovery.