Knowledge Structures Based Adaptive Testing Model

Abstract

How to know the students' overall cognitive situation quickly and accurately by testing has been a hot issue in teaching. Based on the relevant research at home and abroad over the adaptive testing, and in view of the current adaptive testing is not intelligent enough, this paper presents a personalized intelligent selection method which combines the traditional dichotomy and knowledge state boundary method. In test error correction process, this paper proposed the recent knowledge state compatible set theory, by using this theory, this paper tries to use probabilistic approach to rationalize the irrational state of knowledge to achieve the maximum "real" of the level of output. Examples of verification found that the proposed adaptive test model could quickly and accurately measure the true level of students' cognitive structure, and greatly improve test efficiency.