Advancing understanding of learning in interaction: An exploration of how ways of participating can influence joint performance and individual learning

Success in joint endeavors is often attributed to the knowledge-base resources that individuals uniquely bring to the collaborative situation. While individual resources are essential, recent research suggests that attention must also be paid to the ecology of relations that develops within interactions that make it possible (or not) for knowledge and other cognitive resources to be accessed and functionally expressed. An enduring issue for CSCL researchers involves developing methods of assessing collaborative interaction and tracing the quality of collaboration to learning outcomes. One critical question for research is whether we can identify relational and/or interactional resources that are important for generative collaboration. In this presentation we will share research that examined the relationship between qualities of student interaction during collaborative problem solving, the accuracy of the problem solution generated and individual student performance on the same and a related problem administered post-collaboration. As part of a previous study (Barron, 2000a), sixth graders of high mathematical ability were randomly assigned to 12 groups of 3 students each. Triads of same-gender students participated by solving the problem together and then solving the same problem and a transfer problem individually. Their interactions were videotaped and transcribed, and their conversations were analyzed. In all groups correct proposals for solution were presented to the group. In high scoring groups these were taken up and documented. However, in low scoring groups these solutions were not accepted and documented and were either rejected or ignored. Previous research (Barron, 2000b; Sears, Barron, & Strobel, 2001) found that successful groups showed higher quality collaboration than unsuccessful groups. In the present research the level of group performance was used as analytic contrast to compare learning outcomes and other dimensions of interactional quality. We found that individuals who participated in successful groups performed significantly better on the mastery and transfer tests. These findings suggest that higher quality collaboration is associated with better group performance and greater individual learning. n