Model Based Systems and Qualitative Reasoning for Intelligent Tutoring Systems

An experimental study is presented, which investigates the process of building qualitative simulation models using HOMER, a tool for building qualitative models of systems. HOMER consists of a number of dedicated editors aiming at decomposing the complexity of the model building process into more manageable subtasks. The aims of the present study are a) validating the original task analysis underlying the implemented system and the resulting task decomposition, b) identifying problems and/or misconceptions users encounter when building simulation models, c) assessing the tool’s usability in supporting the model building process, and (d) defining possible improvements and faults of the present version.