Identifying peer interaction patterns and related variables in community-based learning

Abstract

The purpose of this study was to investigate patterns of peer interactions and to identify the relationships of peer interactions with learner characteristics and learning outcome in community-based learning. The participants were 24 middle school students joined in an online learning community for a week. Two patterns of peer interactions such as in-degree and out-degree centrality were identified. Students with high intra-personal or verbal-linguistic intelligence were related to high in-degree centrality, while students with high interpersonal intelligence or prior knowledge were related to high out-degree centrality. That means "self smart" or "word smart" students were popular and played knowledge broker-s roles in their community. On the other hand, "social smart" students or "high prior knowledge" students were open and friendly activators and delivered vast information. Moreover, higher peer interactions were related to better learning outcome. These results indicated that peer interactions were important intervening variables to enhance learning effect. © 2005 International Society of the Learning Sciences.