DOES PEER DISCUSSION DURING THE USE OF CLICKER QUESTIONS RESULT IN DEEP LEARNING?

ConcepTest questions are conceptual, multiple-choice questions that are asked during lecture to gauge student understanding. In a typical use of clickers during lecture, students are asked a ConcepTest question, they answer it individually using a clicker, they are shown the distribution of responses and discuss their own responses with their peers, and finally they answer the question again using a clicker. Although students more frequently respond correctly to the second asking of the question after peer discussion, it is difficult to assess if that is because students learn the topic during peer discussion or because they change their answer to their peers’ answers without truly understanding.

In this study, ConcepTest questions were used in a small historical geology course at a community college. On three exams, a total of 26 ConcepTest questions were included that had been asked in class on which students had substantially increased the number of correct responses after peer discussion. Student answers were analyzed only if they answered the same ConcepTest question for both clicker questions and the following exam. If students learn the topic through peer discussion, their responses on the exam question should be more similar to the answers after peer discussion compared to their initial answers. Overall, students initially answered the ConcepTest questions 51% correctly on the initial asking, 71% correctly after peer discussion, and 68% correctly on the exam. The response on the exam for all students matched the initial response 48% of the time (37% of the responses were correct), and it matched the response after peer discussion 57% of the time (50% correct). For students who changed their ConcepTest responses from incorrect to correct after peer discussion, 65% of them answered the exam question correctly, 18% answered the exam question incorrectly matching their initial response, and 17% answered the exam question incorrectly with a different response. In comparison, of those students who answered the ConcepTest question incorrectly both times in class, 57% answered the exam question correctly and 28% retained their same incorrect answer on the exam. As a result, preliminary results indicate that, in general, ConcepTest questions likely induce the deep learning we desire through peer discussion.