Lay Perceptions About Machine and Human Judgments

Jiani Xue, PhD Candidate in Marketing, The Wharton School and Stefano Puntoni, Marketing, The Wharton School

Abstract: In today’s world, consumers and businesses frequently require expert advice to make accurate predictions. Much of this advice stems from algorithms, such as recommendation systems and AI prediction models. The goal of this project is to explore how lay people perceive the accuracy of machine predictions. We aim to answer the following questions: 1) How do lay people perceive the accuracy of ensemble models (sets of machine predictions)? 2) Where do these lay perceptions come from? 3) Are these lay perceptions accurate? 4) How can we improve perceptions to encourage more adoption of algorithmic predictions and decisions?