Computer-Assisted Decision Making: Performance, Beliefs, and the Illusion of Control

Abstract

Recent empirical evidence indicates that computer-assisted what-if analysis does not predictably improve decision making. Why then is what-if analysis so widely used by decision makers? We argue that what-if analysis creates an "illusion of control" which leads decision makers to overestimate its effectiveness. A between-subjects experiment was conducted using a production planning task to test this conjecture. As hypothesized, subjects falsely believed that what-if analysis improved their decision making. In fact, what-if analysis users expressed inflated confidence beliefs yet post hoc analysis revealed that they actually performed significantly worse than nonusers in the trials immediately preceding belief measurement. In light of other research linking user acceptance of computer-based technologies to users' performance perceptions, these results forewarn of sustained, but dysfunctional, use of what-if analysis.