Eight elementary schools in metropolitan Nashville, Tennessee, participated in the study. Four
of the eight study schools were classified as Title I schools; the other four were middle-class,
non-Title I schools.
Before the study began, 323 first-grade students used PALS for seven weeks and were subsequently
tested. The 66 students who scored 0.5 standard deviations or more below average
readers in terms of both level and slope on the curriculum-based measures comprised the
sample for this study. These 66 students were randomly assigned to one of three conditions:
PALS, modified PALS, and adult tutoring, with 22 in each condition.
The intervention was delivered one-on-one by peers in the PALS and modified PALS conditions
and one-on-one by up to eight adult tutors in the adult-tutoring condition. The final analysis
sample consisted of 56 students: 21 in PALS, 15 in modified PALS, and 20 in tutoring. Fifteen
(27%) of the 56 students were English language learner students. Only the PALS vs. tutoring
analysis meets WWC evidence standards, so the analysis sample used in this review includes
41 students in eight schools.
PALS is a peer-tutoring program that emphasizes phonological awareness, decoding, and
fluency. In this study, it was implemented three times a week for 35 minutes each session.
Teachers paired higher performing and lower performing readers who took turns coaching
each other. The intervention group received PALS over the course of 13 weeks.
The comparison group received one-on-one tutoring from trained adult research assistants.
Adult tutoring took place three times a week, 35 minutes each session, for 13 weeks, and
covered the same topics as in the two PALS conditions. The tutoring session was structured
similar to a special education pullout program, with greater attention to skill mastery and the
student’s specific needs. The study viewed PALS as the business-as-usual comparison group,
but the WWC treated the tutoring condition as the comparison for the purposes of this review
Testing was conducted at baseline and at follow-up by two full-time project coordinators and
eight graduate students who were trained to ensure inter-rater agreement of at least 90%.
Students were tested over two one-on-one sessions in a quiet location in their school. Students
were not tested by staff who had tutored them. The baseline Dolch measure, developed
by research staff, was used as the covariate in analyses. For this measure, the score was
recorded as the number of high-frequency words read correctly in one minute. The outcomes
included in this study were Blending, Rapid Letter Naming, Rapid Letter Sound, Segmentation,
Spelling, Word Identification, and Word Attack in the alphabetics domain; Near-Transfer
Fluency and Far-Transfer Fluency in the fluency domain; and the Comprehensive Reading
Assessment Battery: Comprehension measure in the comprehension domain. For a more
detailed description of these outcome measures, see Appendix B.
Support for implementation
Teachers were trained to use PALS in October through a one-day training session before the
start of this study. A research staff member visited each classroom twice weekly over the sevenweek
period of the initial PALS implementation. In January, research staff attended a one-day
workshop to learn the modified PALS and tutoring procedures. Each staff member was then
assigned to implement the tutoring or modified PALS intervention. Intervention fidelity for PALS
was measured at 92% based on classroom checks conducted in December and March.