The study took place in a large urban school district. There were a total of 120 3rd and 5th-grade students in 13 schools across the three study groups. The third-grade students were from 8 schools and 28 classrooms, and the fifth-grade students were from 5 schools and 22 classrooms.(pp. 5,7)
Students in the analytic sample were about evenly split between males and females, with 49 percent of the sample being female. Black students made up 39 percent of the sample, and white students were 29 percent of the sample. Race is not observed for the 24 percent of the sample that is Hispanic, and is not reported for 8 percent of the sample whose race is "other." The racial and ethnic characteristics of the grade 3 subgroup was quite different from the grade 5 sample. For example, in the grade 3 sample, 46 percent of students were Hispanic as compared to 2 percent in the grade 5 sample. Nearly all students in the sample, 94 percent, were eligible for free or reduced price lunch, and 6 percent had individualized education plans. The study does not present sample characteristics separately by intervention condition. (Table 2, p. 31)
The study examined the effects of a reading comprehension intervention alone (COMP) and in combination with a working memory training component ([WM]COMP). For each of 14 weeks, the [WM]COMP group received tutoring three times per week. The first two tutoring sessions of each week lasted 45 minutes and were delivered to pairs of students. The third session of each week lasted 20 minutes and was delivered individually. All sessions were scripted.
The intervention group received instruction in "before-reading strategies," (which included, for example, identifying the title, headings, pictures, and charts; identifying and looking up bolded vocabulary words; and considering background information relevant to the passage) and "after-reading strategies" (which included retelling important facts from the passage, identifying the most important person or thing and the most important information about that person or thing; and constructing a main idea statement). Each tutoring session also began with a "speeded cloze activity" to enhance fluency and comprehension. This activity involved reading a paragraph summarizing the previous lesson's text, but with words omitted, and students had to choose which word from a pair belonged in the blank. (pp. 8-11) Finally, students in the [WM]COMP group participated in activities to strengthen their working memory. For example, after completing the cloze activities students in the [WM]COMP group were asked to recall, in order, all the words they had selected to fill in the blanks.
Forty-seven percent of students in the [WM]COMP condition missed some general reading instruction, including activities such as centers, independent work, Response to Intervention (RTI) or other intervention services (unrelated to COMP or [WM]COMP), and skills practice. (pp. 8-11)
In the comparison condition, students received the business as usual reading instruction from their regular classroom teachers. All students (intervention and comparison) may have also received additional supports generally available to all students, including after school tutoring, English language tutoring, other small group tutoring, occupational therapy, or other available services. (p. 33)
Support for implementation
Tutors received three half-day training sessions, at which they learned the underlying rationale for each intervention component, saw demonstrations of instructional activities, and participated in role-playing exercises. After the training sessions, tutors practiced the COMP and [WM]COMP protocols with another tutor for six hours. They also "tutored" a project coordinator, who acted as a student, and had to show 90 percent adherence to both intervention protocols. (pp. 13-14)