Computerised Adaptive Testing based on Decision Tree

Abstract

Computerised adaptive testing (CAT) is the process of providing a test experience that suits each examinee’s ability (Weiss, 1983). The focus on using and developing item response theory (IRT) as the most effective method for implementing CAT has been increasing and evolving since 1920. This evolution is interested in reducing the limitations of IRT algorithms. Instead, this research designs, implements, and evaluates the effectiveness of a decision tree (DT) as an alternative CAT algorithm.