Understanding Representation Fidelity: Guidelines for Experimental Evaluation of Conceptual Modeling Techniques

Recently, there has been a resurgence of interest in experimental research on conceptual modeling in information systems analysis and design. There is a need to explicitly identify the objectives of specific experiments in this area, and the role that assumptions play in experimental design. We provide four guidelines for developing materials for experiments aimed at evaluating conceptual modeling techniques, based on the premise that the primary purpose of conceptual modeling is to facilitate communication between analysts and users in validating domain knowledge relevant to an information system. We offer the guidelines as recommendations to assist the development of experiment materials that support meaningful tests of domain semantics, and present empirical evidence to illustrate the value of two of the guidelines. We also evaluate the degree to which a selection of recent experiments on conceptual modeling adheres to the guidelines, and consider implications of that assessment.