Using Data Mining of Learner Behaviors to Inform Design of Role-Play in Virtual Reality–Based Teacher Training

• In Event: Studies and Applications of Virtual and Augmented Reality

Abstract

This study recruited 33 graduate teaching assistants in STEM disciplines to participate in teacher training that incorporated VR-supported role-play. The participants’ behaviors were screen recorded and coded using a systematic scheme. The coding results were further analyzed using data sequential pattern mining and association rule mining approaches to investigate the participants’ behavioral patterns. The on-going data analysis yielded preliminary findings. The participants tended to actively react to prompting but were less so in response to the feedback they received during training. The GTAs also rarely took spontaneous or voluntary teaching acts without being prompted. The current findings provide insights that help understand learner behaviors during VR-supported role-play and shed light on designing effective role-play for teacher training in VR environments.

Authors

• Zhaihuan Dai, University of South Florida

• Fengfeng Ke, Florida State University

• Chih-Pu Dai, Florida State University

• Mariya Pachman, Florida State University

• Luke West, Florida State University

• Saptarshi Bhowmik, Florida State University

• Xin Yuan, Florida State University