From Measurements to Measures: Learning Risk Preferences under Different Risk Elicitation Methods

This study explores how people learn and adapt their risk preferences using different elicitation methods, challenging the neoclassical theory that suggests preferences are fixed. Instead, we show that preferences can change. However, we aim to explain whether the observed changes are due to a real change in the measure, i.e. individuals' risk preferences, or if they are attributable to the limitations of the measurement tool, i.e. the specific risk elicitation method employed. We use a detailed experimental design to examine the stability and consistency of risk preferences using a hands-on learning experience. Our main goals are to assess how consistent risk choices are, understand how preferences remain stable or change over time, and evaluate the effectiveness of different elicitation methods like the Multiple Price List and Ordered Lottery Selection ones. On the one hand, results demonstrate that risk preferences are variable and adaptable, and this can be partly due to the role of experience-based learning. On the other hand, we observe how Multiple Price List methods, even if more complex, are more accurate in identifying risk preferences and then in improving measurement stability and accuracy.