WP 2019-01 Statistical Power and the Individual Level Estimation of Risk Preferences

AUTHORS: Brian Albert Monroe

ABSTRACT: Accurately estimating risk preferences is of critical importance when evaluating data from many economic experiments or behavioral interactions. I conduct power analyses over two lottery batteries designed to classify individual subjects as one of a number of alternative specifications of risk preference models. I propose a conservative case in which there are only two possible alternatives for classification and find that the statistical methods employed to conduct this classification result in type I and type II errors at rates far beyond traditionally acceptable levels. Following a Bayesian approach, I additionally find that the proportion of agents in a population that employ each model critically informs the probability that subjects are correctly classified.