Assessing the Impact of a Proposed Expert System via Simulation

Abstract

A proposed expert system was simulated in order to evaluate its potential impact upon instruction. Undergraduate students ( n = 100) enrolled in a course in deductive logic were assigned to one of three treatment conditions. Students in one group received diagnostic feedback from the instructor as was the current standard in the course. Students in a second group received diagnostic feedback from a computer in ways that simulated the operation of an expert system. Students in a third group received no feedback from either the instructor or the computer so as to provide a baseline for assessing feedback effectiveness. Dependent variables included exam performance, attendance rate, in-class responsiveness, and several attitudes. The most favorable results were found for students receiving feedback from the instructor.