Search
Program Calendar
Browse By Day
Browse By Time
Browse By Person
Browse By Room
Browse By Unit
Browse By Session Type
Search Tips
Visiting Washington, D.C.
Personal Schedule
Sign In
X (Twitter)
This paper demonstrates the utility of a novel algorithm for predicting student performance within the ASSISTments Intelligent Tutoring System (n = 3684), the Inverse Bayes Filter (IBFi). Prediction accuracy is demonstrated by 1) describing patterns of prediction error and 2) comparing the performance of the algorithm against the modified Bayesian Knowledge Tracing model (KTPC) of Wang and Heffernan (2011) when predicting partial knowledge. IBFi tends to over-predict high values of student performance and under-predict low values. However, IBFi is more accurate at predicting student performance than the KTPC model on a per-student basis. These results demonstrate the feasibility of using Inverse Bayes to model partial credit and the implications of the method as an individualized measure are discussed.