Prior Divergence: Do Researchers and Participants Share the Same Prior Probability Distributions?

Abstract

Do participants bring their own priors to an experiment? If so, do they share the same priors as the researchers who design the experiment? In this paper, we examine the extent to which self-generated priors conform to experimenters’ expectations by explicitly asking participants to indicate their own priors in estimating the probability of a variety of events. We find in Study 1 that despite being instructed to follow a uniform distribution, participants appear to have used their own priors that deviated from the given instructions. Using subjects’ own priors allows us to account better for their responses rather than merely to test the accuracy of their estimates. Implications for the study of judgment and decision making are discussed.

Keywords: Prior probability distributions, decision making, experimental design