Student model accuracy using inspectable bayesian student models

Abstract. Inspectable student models (ISMs) have been used in a variety of applications. In order to create and fully deploy learning environments based on ISMs, issues such as how accurate are the student models when students and teachers interact with them should be further explored. This paper presents a framework for learning environments based on inspectable student models. The learning game, a learning environment based on this framework, has been implemented and used in a study focused on student model accuracy using inspectable Bayesian student models. Results of a study with the learning game suggest that Bayesian student models successfully integrate evidence from the student and the system/teacher producing an accurate aggregate view of the model. 1