Building a Predictive Model of the Problem-Solving Performance of Students Using a Computer-Based Tutor

• In Event: Poster Session 11
  In Poster Session: Machines and Digital Agents

Abstract

With the rapid development of computer and telecommunication technologies, e-learning is gaining popularity especially in higher education. One of the unrecognized benefits of an e-learning environment is that it can provide a new avenue for understanding student learning without interrupting their learning processes. Most of e-learning environments create log files capturing detailed descriptions of students’ learning activities (e.g., how much time they spent on a particular learning content). This study investigate how information stored in the log files of a computer-based physics tutoring environment can be used to predict students’ problem solving performance, which can be used as a guideline for determining appropriate moments to provide instructional scaffolding.

Author

• Young-Jin Lee, The University of Kansas