Overall, 36 experimental-control comparisons met the inclusion criterion, of which 7 used random assignment to treatments. No studies of reading curricula qualified, but there were 8 studies of CAI, 16 of instructional process programs, 10 of combined CAI and instructional process programs, and 2 of combined curriculum and instructional process programs. Effect sizes were averaged across studies, weighting by sample size.
Computer Assisted Instruction (CAI). Studies of CAI find minimal achievement outcomes. Mean weighted effect size across 8 studies: +0.10.
Instructional Process Strategies (IP). The largest number of high-quality studies evaluated instructional process programs, especially forms of cooperative learning (ES= +0.28 in 7 studies). Mean weighted effect size across 14 studies: +0.21.
Combined CAI and Instructional Process Programs (CAI + IP). Positive effects were found for READ 180. Mean weighted effect size across 9 studies: +0.22.
Combined Curriculum and Instructional Process Programs (Curr + IP). A randomized study of REACH found an effect size of 0.00, and the same study found an effect size of +0.17 for RISE.