Exploration of behavioral markers to support adaptive learning

Abstract

In designing and developing adaptive learning systems, it is desirable to incorporate as much information about the learner as possible to better tailor in instructional experience. Behavioral markers exhibited by the learner offer a source of information with the potential to shape instructional content. In the case of computer-based training environments, this source of information may include behaviors ranging from mouse cursor movement, key stroke dynamics, or eye tracking. We present methods for analyzing the mouse behavior of a learner using kinematic data in situations where knowledge of areas of interest on the screen are not known by the system a priori, as well as in multiple-choice scenarios to analyze the amount of attention spent by the user on various response items. The outcome of this work is to help inform and to influence future studies in adaptive learning which may seek to incorporate such sources of learner information.