Information-based Item Selection with Blocking Strategy based on a Bayesian network

Abstract

With the rapid development of computer technology information theory has been implemented for searching optimal adaptive item sequence in computerised adaptive test systems based on Bayesian network. Information theory such as entropy between dichotomous concepts and test items generalise common intuitions about item comparison for heuristic methodology. However, the executive time and the storage space are still open problems in constructing and storing decision item trees. The blocking strategy is proposed for overcoming those problems. Experimental results show that the blocking strategy could overcome both the executive time and storage space problems. Key-Words: Bayesian network, Information, Entropy, Computerised adaptive testing, Blocking strategy.

Topics

9 Figures and Tables

Download Full PDF Version (Non-Commercial Use)