An empirical test of expert system explanation facility effects on incidental learning and decision-making
Kathleen Moffitt, James C. Hershauer · 1989
This research examines the incidental learning of the declarative knowledge and decision processes present in expert systems by users of those systems. This is examined in reference to the ability of expert systems to provide explanations of their reasoning. The research draws from three distinct bodies of literature, which are explored and synthesized in reference to the research goals and questions. These are: explanation, learning, and choice shifts in decision making. Cognitive learning is a theme which underlies each and forms the foundation for the research. Incidental learning of declarative knowledge and decision processes is assessed within a laboratory setting in which four forms of explanation provision are manipulated. A comparison of incidental learning among these forms of explanation provision is made. Results indicate that context specific embedded explanations provide for a higher degree of incidental learning, that users differentially rate explanation provision techniques when usefulness as a learning aid is measured, and that users do not distinguish among explanation provision techniques when their usefulness as decision aids is measured.