MIXED APPROACH FOR ITEM SELECTION IN E-TESTING

Abstract

Learning and instruction can be enhanced with the information provided from assess- ment. Computerized adaptive testing is an effective mode of e-testing as it can be oper- ated in a continual and dynamic mode. It can also cater for individual needs by providing tailor-made tests. In this paper, we will examine a new item selection approach which would allow more flexibility in balancing the two competitive goals of accuracy and item pool utilization in computerized adaptive testing. Moreover, a new index for comparing a weighted balance of these two goals is proposed.