Publications & Resources
An Exploratory Study Examining the Feasibility of Using Bayesian Networks to Predict Circuit Analysis Understanding
Gregory K. W. K. Chung, Gary B. Dionne, and William J. Kaiser
Renewed interest in individualizing instruction, particularly with the use of technology, has resulted in a search for methods that can accurately diagnose student knowledge gaps and prescribe appropriate remediation. In this study we gathered validity evidence for the use of a Bayesian network to model students’ understanding of circuit analysis concepts. Thirty-four undergraduate students completed tasks designed to measure conceptual knowledge, procedural knowledge, and problem-solving skills. Results suggested that the Bayesian network was generally working as intended. When high- and low-performing groups were formed on the basis of posterior probabilities, significant group differences were found favoring the high-performing group with respect to circuit definitions and circuit analysis problems, for both actual and self-assessments, and higher major GPA. The Bayesian network also predicted participants’ performance on problem-solving items on average 75% of the time. The findings of this study are promising for developing scalable and feasible online automated reasoning techniques to diagnose student knowledge gaps.
Chung, G. K. W. K., Dionne, G. B., & Kaiser, W. J. (2016). An exploratory study examining the feasibility of using Bayesian networks to predict circuit analysis understanding (CRESST Report 850). Los Angeles: University of California, Los Angeles, National Center for Research on Evaluation, Standards, and Student Testing (CRESST).