The Use of Sensor Data for the Multimodal Analysis of Mathematical Proof Activity in a CSCL Environment

Abstract

Recent technological developments have greatly enhanced the power of multimodal learning analytics to help researchers better understand how students learn in a CSCL environment in a scalable manner. In this study, we try to use depth sensor data to identify the temporal transition of learners’ behaviors and investigate the relationship between it and the progress of their mathematical thinking. Human-coded behavioral data are compared with the log data of learner movements derived from a depth sensor to evaluate the extent to which the sensor data can be used to infer the progress of learners’ thinking during mathematical proof activity in a CSCL environment.