Setting
The study was conducted across 18 schools, 56 third-grade classrooms in Nashville and 24 third-grade classrooms in Houston.
Study sample
For this contrast, three out of four students (75 percent) were eligible for free- or reduced-price lunch, 16 percent were special education, and 9 percent were English learners. About half (48 percent) were female.
Intervention Group
There were four groups in this study; for this contrast, the intervention condition was Fact Retrieval tutoring, which consisted of three activities: computer-assisted instruction (7.5 minutes), flash card practice with corrective feedback (4 minutes), and cumulative review with corrective feedback (4 minutes).
The computer-assisted instruction used Math Flash in which addition and subtraction problems with answers were briefly displayed (“flashed”) on the computer screen for 1.3 seconds, after which the student was asked to recollect and type the math fact from memory. Typing in the math fact populated a number line illustrating the math fact at the top of the screen. The student received feedback to indicate whether they completed the task correctly, after which a new problem was displayed.
The flash card practice involved two types of activities. First, math facts without answers were shown and students were asked to complete the displayed math problem. After students demonstrated sufficient mastery of these types of exercises, they were shown a second type of flash card, which displayed a number line illustrating a math problem. The students were asked to describe the math fact stated by the number line.
Cumulative review was a paper-pencil activity, in which students were asked to complete 15 math problems on paper. The tutors then corrected the math problems aloud while the student observed.
As with the other three groups in the study, students received 3 sessions per week for 15 weeks for a total of 45 sessions.
Comparison Group
For this contrast, the comparison condition was Fact Retrieval and Procedural Estimation (Combined) tutoring. This condition comprised four activities: computer-assisted instruction using Math Flash (part of the Fact retrieval skill intervention: 7.5 minutes), computer-assisted instruction using Magic Math (part of the Procedural/estimation skill intervention: 5-10 minutes), the flash card practice incorporated in the procedural/estimation skill intervention with corrective feedback (4 minutes), and cumulative review with corrective feedback (4 minutes).
The Math Flash portion of the curriculum was exactly the same as the intervention group.
The Magic Math computer-assisted instruction comprised three segments: i) addressing conceptual underpinnings using pictorial representations of ones and tens; ii) teaching procedural steps of two-digit addition and subtraction, relying on the same addition or subtraction problems worked in the first segment; and iii) estimation for which the student worked on a double-digit addition problem that differed from the problems used in segments 1 and 2.
Flash cards practice in this condition included three types of flash card activities: i) two-digit addition or subtraction problems with or without regrouping to which the student responded by stating whether to add or subtract and then whether to regroup or not regroup; ii) cards showing a two-digit addition problem, with or without regrouping for which the student stated whether the sum of the ones column was closest to 0, 10, or 20; and iii) the same set of cards for which the student had to state the estimated answer to each two-digit addition problem.
Cumulative review was a paper-pencil activity, in which students were asked to complete 15 math problems on paper. The tutors then corrected the math problems aloud while the student observed.
As with the other three groups in the study, students received 3 sessions per week for 15 weeks for a total of 45 sessions.
Support for implementation
Tutors were trained over two full days. During the weeks following training, tutors studied the tutoring scripts and practiced implementing the procedures alone and with each other. Tutors then each conducted a session in each study condition with a project coordinator who provided corrective feedback. Research assistants met with project coordinators every two to three weeks to address problems and questions.