This work addresses the widening gap between research in intelligent tutoring systems (ITSs) and practical use of this technology by the educational community. In order to ensure that ITSs are effective, teachers must be involved in their design and evaluation. We have followed a user participatory design process to build a set of ITS knowledge acquisition tools that facilitate rapid prototyping and testing of curriculum, and are tailored for usability by teachers. The system (called KAFITS) also serves as a test-bed for experimentation with multiple tutoring strategies. The design includes novel methodologies for tutoring strategy representation (Parameterized Action Networks) and overlay student modeling (a "layered" student model), and incorporates considerations from instructional design theory. It also allows for considerable student control over the content and style of the information presented. Highly interactive graphics-based tools were built to facilitate design, inspection, and modification of curriculum and tutoring strategies, and to monitor the progress of the tutoring session. Evaluation of the system includes a sixteen-month case study of three educators (one being the domain expert) using the system to build a tutor for statics (forty topics representing about four hours of on-line instruction), testing the tutor on a dozen students, and using test results to iteratively improve the tutor. Detailed throughput analysis indicates that the amount of effort to build the statics tutor was, surprisingly, comparable to similar figures for building (non-intelligent) conventional computer aided instructional systems. Few ITS projects focus on educator participation and this work is the first to empirically study knowledge acquisition for ITSs. Results of the study also include: a recommended "design process" for building ITSs with educator participation; guidelines for training educators; recommendations for conducting knowledge acquisition sessions; and design tradeoffs for knowledge representation architectures and knowledge acquisition interfaces.
Identifer | oai:union.ndltd.org:UMASS/oai:scholarworks.umass.edu:dissertations-8500 |
Date | 01 January 1992 |
Creators | Murray, Thomas Joseph |
Publisher | ScholarWorks@UMass Amherst |
Source Sets | University of Massachusetts, Amherst |
Language | English |
Detected Language | English |
Type | text |
Source | Doctoral Dissertations Available from Proquest |
Page generated in 0.0157 seconds