Tailoring of feedback in web-based learning systems : certitude-based assessment with online multiple choice questions

Web-Based Learning Systems (WBLS) become a new reality of the modern education. Among the whole WBLS functionality, feedback is the one that gives the student responses from the system. It performs many functions: informing, motivating, correcting, evaluating students, keeping their attention, and providing additional comments and explanations. The increasing number and variety of the users call for studying and developing personalization methods and technologies for WBLS in general and, in particular, tailoring of feedback to individual characteristics of learners and their knowledge, thereby providing access to the most suitable type of feedback. This thesis presents a study of potentially influential factors on the effectiveness of and students’ preferences for different types of feedback during online assessment, and the results of the experimental quantification of the selected relations. Our studies demonstrate in particular that individual learning styles and a combination of response correctness and certitude are important factors that highly influence the feedback preferences of students and the effectiveness of presented feedback. We also develop a feedback personalization mechanism based on our empirical findings and show its feasibility and benefits for learning and interaction processes in certitude-based assessment.