Bayesian Networks and Their Applications in Educational Assessment

Abstract

This research introduces the Bayesian Networks (BNs) as one of the modern models in psychological and educational measurement and assessment that are based on probability theory and graph models. It also reviews the concepts of Bayesian Networks (BNs) and their theoretical foundations, their advantages, in addition to their components and their relation to evidence centered design -ECD and assessments that include multidimensional proficiency models, fit indices, estimation methods, and their important statistical programs. The research also addresses the steps of their construction, and their applications, how to use them in the field of educational assessment, some studies that applied them in the area of assessment, the important criticisms to them, and their future directions. The research concluded that it is important to raise the awareness of Bayesian network (BN) models and their uses in constructing assessment tools for the purposes of diagnosis, prediction, and learning monitoring and improvement, and to conduct studies on their psychometric properties in the contexts of psychological and educational measurement and assessment.