The study took place in an unspecified number of elementary schools in 14 states.
The study was designed to examine the effect of Reading Recovery® on the outcomes of firstgrade
students. Forty-seven Reading Recovery® teachers each identified two students7 eligible
for Reading Recovery® based on their low scores on the Observation Survey of Early Literacy
Achievement and their own judgment. These 94 students were randomly assigned to enter the
Reading Recovery® program during either the first or second half of the school year. [Note: The
study also included two additional comparison groups of 47 low-average and 47 high-average
readers from the same classrooms as the Reading Recovery® students who were not expected
to participate in the Reading Recovery® program. Analysis involving these comparison groups
was not eligible for WWC review because the WWC considers only comparisons of students
with similar achievement backgrounds in assessing the effectiveness of an intervention.]
Because of missing test data, the author’s final analytic sample included 74 students distributed
across 37 teachers.
Students participated in the one-on-one daily 30-minute tutoring program for up to 20 weeks
or until they were judged by their teacher to have met the criteria for termination of the program
by reaching average levels of literacy performance. The length of program participation
ranged from 12 to 20 weeks. Originally, participants were taught by 47 Reading Recovery®
teachers who had volunteered to be part of the study, but because of missing test data, data
from only 37 teachers and 37 students were included in the author’s final analysis. The intervention
group was 61% male, 47% Black, 38% White, 12% Hispanic, and 3% Asian. About
60% of the group received free or reduced-price lunch.
The comparison group included students who were randomly assigned to receive Reading
Recovery® during the second half of the year. Thus, these participants served as a comparison
group only during the first part of the year when they received instruction in their regular
classroom but no additional supplemental services. The final analysis included data from 37
teachers and 37 students. The comparison group was 41% male, 47% White, 38% Black, and
15% Hispanic. Approximately 57% of the group received free or reduced-price lunch.
The study author reported outcomes on ten literacy measures, all of which were included in
the WWC review and ratings of effectiveness. Six reported subtests of the Observation Survey
were included in the WWC review of this study: two in the alphabetics domain, including Letter
Identification and Word Recognition; one in the fluency domain (Text Reading Level); and three in
the general reading achievement domain, including Concepts About Print, Dictation, and Writing
Vocabulary. The study author also reported two additional outcome measures that fall into the
alphabetics domain, Phoneme Segmentation and Deletion task, one additional outcome in the
fluency domain, Slosson Oral Reading Test–Revised, and one outcome in the comprehension
domain, Degrees of Reading Power. For a more detailed description of the included outcome
measures, see Appendix B.
Support for implementation
Although the study provided no information about training provided to participating teachers,
Reading Recovery® teachers typically must complete a year-long certification program.