A Study of Framing Effects in a New Risk Aversion Experiment

Abstract

Risk aversion experiments such as those by Holt and Laury (2002 and 2005) measure risk aversion by examining responses of experimental subjects who are confronted with single-sheet paper displays of probability-ordered arrays of choices in which “real” money is at risk. As an alternative to this approach, the findings reported in this paper were obtained using a modified adventure-type video game to offer the choices presented by the HL experiment embedded in a more realistic scenario. The decisions are confronted first by our experimental subjects in a sequential and unordered manner. Then, later in the experiment, subjects are instructed to examine the results of their decisions in an array that shares the simultaneous and probability-ordered characteristics of the standard laboratory protocol. Subjects then had the option of altering their decisions before their payment was determined. The results indicate that decisions made in a sequential and unordered manner exhibit less risk aversion and higher degrees of decision inconsistency.