Proposal for Group Composing Methods Using Social Network Analysis and Effects of Cooperative Learning

Abstract

Active learning has been focused as an active study method in which a student discovers a problem and actively finds the solution. Cooperative learning, in which students work cooperatively in small groups, is one methods of active learning. Working together in a small group to accomplish shared goals is an effective approach to mutual learning and the development of good friendships. In recent years, group composition methods for cooperative learning have drawn the attention of researchers. Although most grouping methods need a preliminary survey to discover information about students, in cases where many students take a lecture at the same time, such as at a university, it is difficult for teachers to know a student's character and ability in advance. In this paper, we propose new grouping strategies for cooperative learning not requiring special surveys. Our proposed methods use a friendship network for students as prior information. We can easily obtain information about a student's friendship network by estimating their friendship from lecture attendance data. We apply our proposed methods to actual cooperative learning activities and observe their effects. Questionnaire results show that satisfaction with the lecture of the students grouped by the proposed methods were higher than that of the students grouped by the conventional method. An analysis of the resulting friendship networks show that students grouped by the proposed methods made more friends than students grouped by a conventional method.