Managers’ susceptibility to biased learning and experience: An experimental investigation

Abstract

This paper contributes evidence on a fundamental source of biased learning — that which arises from decision-makers only having outcome information for alternatives that are selected and implemented, but not for those that are rejected. In a multi-period experiment with experienced Taiwanese managers, the subjects’ inferences and decisions are found to be adversely affected by incomplete outcome information. This result holds even for managers who believe themselves to be highly aware of the potential bias being investigated. Data from the experiment further illuminate how the subjects had been led by the incomplete outcome feedback to erroneous inferences and in turn, how these inferences had adversely affected their decisions. The subjects’ responses to open-ended questions yield further insight into the extent and effects of the bias in real world settings, and possible ways to reduce their severity.