Feedback Micro-engineering in EER-Tutor
Konstantin Zakharov
Department of Computer Science
University of Canterbury
Abstract
Effectiveness of one-to-one tutoring is the key motivating factor supporting the development of Intelligent Tutoring Systems. These systems are capable of adjusting learning support to the individual needs of their users. This study presents EER-Tutor, an Intelligent Tutoring System for teaching Enhanced Entityrelationship modelling. We use EER-Tutor for testing a hypothesis, that feedback messages designed with the theory of learning from performance errors in mind are more effective than conventional feedback messages. The principles of the Cognitive Architecture suggest that the proposed feedback design should provide long and short-term learning advantages through revision of faulty knowledge in the context of learners' errors. The overall outcome of this evaluation study is promising. However, a comparison of the experimental and control groups did not reveal a significant statistical difference between the proposed format and conventional format of feedback. At the same time, the analysis based on the Power Law of Learning displayed a consistent advantage of the experimental feedback. The lack of conclusive results suggests the need of further research in this area.
