Authors:
Mariz Awad
;
Jailan Salah
;
Nabila Hamdi
and
Slim Abdennadher
Affiliation:
Faculty of Media Engineering and Technology, German University in Cairo, New Cairo, Cairo, Egypt
Keyword(s):
Intelligent Tutoring Systems, Computer Adaptive Testing, Adaptive Item Selection, e-Learning, Expert Systems, Item-based Learning.
Abstract:
Computer Adaptive Testing (CAT) methods have been widely used by test centres to assess examinees quickly. These methods change question difficulty in response to the performance of the examinee. This work presents a modified framework, which we call Computer Adaptive Learning (CAL). CAL uses the CAT principles to improve exam-training efficiency rather than assessment efficiency. We applied the proposed method to a learning platform and conducted a comparative experiment using 50 participants to investigate the effectiveness of CAL. We evaluated the system in terms of knowledge gain, learning efficiency, and engagement by comparing it to another adaptive method in which the game mechanics and UI adapt to the user’s emotional state. Results confirm that the proposed CAL algorithm exposes the learner to questions more efficiently and improves the learning gain when compared to traditional systems in which difficulty increases sequentially. Engagement, however, did not differ across sy
stems.
(More)