Cooperative Learning in Introductory Statistics: Assessing Students’ Perceptions, Performance, and Learning in Heterogeneous and Homogeneous Groups

Abstract

Past research has shown the benefit of cooperative learning in organizing instruction to promote students’ engagement and achievement. However, it remains unclear whether students gain more from peers at relatively higher or similar levels of prior knowledge. Relatedly, it is important to distinguish collective performance from individual learning. Lastly, students’ perceived value of cooperative learning, group dynamics, and self-concept could elucidate the importance of group composition. We conducted a mixed-methods experiment in five introductory statistics classrooms with 139 university students of relatively low (LPK) or high (HPK) prior statistics knowledge. Students were randomly assigned within-class to homogeneous (all LPKs) or heterogenous (at least one HPK) groups, with whom they completed exercises throughout the term. Without a more knowledgeable peer, LPKs in homogeneous groups nonetheless were able to clarify much of the conceptual confusion among themselves and eventually reach the same level of understanding as LPKs in heterogeneous groups. Group composition did not influence LPKs’ individual learning on average. Within heterogeneous groups, LPKs had greater performance gains and similar individual learning when compared with HPKs. Regardless of prior knowledge or group composition, students found cooperative learning valuable, as it promoted opportunities for positive interactions, comparing work, and revealing misunderstandings for all. Supplementary materials for this article are available online.