This three-year longitudinal study involved the use of hypermedia computer assisted instruction (CAI) to provide individualized reading instruction within a general classroom environment for students in grades K-3. The software provided reading lessons in a hypermedia format designed as supplementary material for a basal reader series. Because this project was interested in the maintenance of students with mild disabilities in the general education setting, particular attention was given to the reading progress of low-achieving students, as they appeared to be the most at risk for being referred to special education classes. Analysis of the achievement test data indicates: Year One. Entire-class significance was shown with experimental classes outperforming control classes at kindergarten, second grade, and third grade. When the classrooms were broken down into ability groups, low students in the kindergarten, second grade, and third grade experimental classes achieved significantly higher total test scores than their control peers. Year Two. No entire class significance was shown at the first, second, or third grade levels. When the classrooms were broken down into ability groups, at the first, second, and third grade levels, the low group students in the experimental classrooms had significantly higher total test scores than the students in the control classroom low groups. Year Three. When comparing entire classes on total test scores, no entire class significance was shown at the kindergarten, second, or third grade levels. When classrooms were divided into ability groups, low students in the kindergarten experimental class achieved significantly higher total test scores than the control students.
