Abstract

This study quantitatively synthesized the literature on the effects of social context (i.e., small group versus individual) when learning with computer technology (CT) on cognitive, process, and affective outcomes. In total, 404 independent findings were extracted from 103 studies involving kindergarten to adult learners. The results indicate that, on average, small group learning had small to moderate positive effects ( p < .05) on group task performance d + = +0.39), individual achievement ( d + = +0.16), and several other outcomes. No significant social context differences were found on amount of task attempted, academic self-concept, and attitudes toward computers, subject, and teachers. Significant heterogeneity ( p < .05), however, was found for several of the outcomes analyzed. Study features analyses were, therefore, performed first univariately and then in multiple weighted least squares regressions to account for the variability in the findings. In the group task performance outcome, type of tasks, feedback, and presence of others accounted for the significant variability in the findings. When tasks involved factual learning, the effects were consistently positive. When tasks involved more complex problem solving, the superiority of group performance over individual performance was stronger when programs provided no or minimal feedback and when no other peers were working close by. Different from the results on group task performance, the study features that accounted for the significant variability on individual achievement include group work experience or instruction, group size, subject, type of programs, gender, and relative ability. Small group learning with CT was more effective when: (a) students had experience or specific instruction for effective group learning than when no such experience or instruction was reported; (b) working in pairs than in 3-5 member groups; (c) working with tutorial or drill-and-practice programs than with exploratory or tool programs; (d) learning social sciences or computer skills than mathematics, science, reading, and language arts; (e) students were female than male; and (f) students were of relatively low or high ability rather than medium ability. When all the positive study features were in place, the social context effect was d + = +0.95. When none of the positive study features were in place, the social context effect was d + = -0.22. These results and others suggest that social context plays an important role in student learning with CT. However, the role of social context differs between group task performance and individual achievement. While the former is a result of distributed cognition, the latter relies on whether each learner is actively involved in constructing his/her individual knowledge and whether group members assist each other in doing so