Barber, L.; Carbo, M.; Thomasson, R. (1998). Bloomington, IN: Phi Delta Kappa.
This tightly controlled study involved thousands of students in grades 1-9 from 10 school districts. It compared the effectiveness of numerous different programs of teaching reading to the Carbo Reading Styles Program. It was found that, when CRS is implemented to the 85% level or higher, it is significantly more effective than extant site programs used in the regular classrooms. Of the 45 reading subtests, 39 produced positive effects in favor of the experimental (CRS) students, and six produced negative effects favoring the control group over a period of two years. This study compared matched teacher pairs. Both the teachers and students were matched. Insufficient matches or low CRS implementation eliminated four of the 10 participating districts, leaving 561 subjects and 15 matched teacher pairs from grades 1-6 in the following six school districts: Bledsoe (TN), Clark County (NV), Detroit (MI), Dillon (SC), Lindberg (MO) and New Bern (NC). This research used a multi-site simultaneous replications design.