What Are They Talking About? Lessons Learned from a Study of Peer Instruction

Abstract

Electronic classroom response systems (CRSs) have been in use in large college lectures for over three decades. Such systems are designed to provide instructors and students with immediate statistical analyses of student electronic responses to multiple-choice questions posed to the class by the instructor. The technique known as peer instruction uses a CRS to tabulate student responses after students have had an opportunity to discuss ideas with seat partners. In this study, we investigate recorded peer CRS conversations collected in two introductory astronomy courses over two semesters. Findings suggest that when instructors adopt a high-stakes grading incentive that assigns little credit for incorrect CRS responses rather than providing forums for the spontaneous exploration of nascent ideas, conversations tend to become dominated by a single partner as students attempt to earn maximum credit for a correct answer. We also present a comparison of two methods for studying conversation bias in peer instruction discourse that could be valuable to other researchers interested in studying peer discourse.