A Contemporary Computer Adaptive Mastery Assessment using Sequential Probability Ratio and Clustering

DOI: 10.5176/2251-1814_EeL13.44

Authors: Kavitha Rajamani and Vijaya Kathiravan

Abstract:
The advancement in information technology and varied learner group made the e-learning as popular. Hence computer based assessment become a prevalent method of administering the tests. Randomization of test items here may produce unfair effect on test takers. There is a need to develop the System that assigns intelligent question depending on the student’s response in the testing session. Computer adaptive testing (CAT) is theoretically sound and efficient to tackle the situation. Using Norm Referencing, Item difficulty and Item discrimination is computed. Both examinees and test items are clustered into 3 groups using k-means in order to choose the specific item according to the ability. The proposed adaptive mastery assessment system overcomes the item maintenance issues in IRT (Item Response Theory) and less adaptive nature of SPRT (Sequential Probability Ratio Test). The system considers the critical item parameters to provide ability based assessment from IRT and content mastery assessment from SPRT. Hence the proposed assessment works in an intelligent way without compromising the accuracy in terms of ability and item.

Keywords: Adaptive Mastery Assessment, Intelligent Testing, Computer Adaptive Testing