Abstract
Despite great expectations in the educational community, Cooperative Computer-Based Instruction (C-CBI) has not been shown to increase students's learning over individual CBI (I-CBI). These lack of findings may be due to the varied implementation of C-CBI. This meta-analysis concentrates on the presence of two factors, cooperative learning training and problem-solving CBI, in studies which compare C-CBI and I-CBI. The results of twenty-three studies were coded and effect sizes were calculated. Effect sizes were analyzed to determine if a greater difference exists between the cooperative and individual conditions when students were trained and used a problem-solving CBI. Results show that studies that included these elements have greater mean effect sizes (es = .413) than the mean of all studies (es = .251). This study provides more evidence that cooperative learning and problem-solving CBIs are important factors in increasing achievement, group interaction, and elaboration in C-CBI.