Intelligent tutoring systems achieve much of their success by adapting to individual students. One potential avenue for personalization is feedback generality. This paper presents two evaluation studies that measure the effects of modifying feedback generality in a web-based Intelligent Tutoring System (ITS) based on the analysis of student models. The object of the experiments was to measure the effectiveness of varying feedback generality, and to determine whether this could be performed en masse or if personalization is needed. In an initial trial with a web-based ITS it appeared that it is feasible to use a mass approach to select appropriate concepts for generalizing feedback. A second study gave conflicting results and showed a relationship between generality and ability, highlighting the need for feedback to be personalized to individual students' needs.
展开▼
