While different approaches were developed to create computerized adaptive practices for e-learning systems, we show that exploiting Rasch models to create adaptive practices can be a new promising approach. Rasch analysis enables us to find a mathematical model to analyze students’ answers to exam questions by representing students’ abilities and questions difficulty levels on the same scale. In this paper, we introduce a novel algorithm to generate adaptive practices based on the Rasch analysis of students’ performance in an initial assessment. This approach enables us to generate adaptive practices that consider not only the student’s ability and his previous performance but also the difficulty level of each question. We also present results from a preliminary field experiment that we have conducted using an online learning system that implements this algorithm. The potential advantages of this approach and the practical contributions are discussed.
Zaqoot, Wisam; Oh, Lih-Bin; Koh, Elizabeth; Seah, Lay Hoon; and Teo, Hock-Hai, "The Use of Rasch Model to Create Adaptive Practices in e-Learning Systems" (2021). ACIS 2021 Proceedings. 69.