Joint paper with Shuo Liu
Abstract: Surveys that are designed to measure subjective states (e.g., happiness) typically generate ordinal data. A fundamental problem is that methods used to analyse ordinal data (e.g., ordered probit) rely on strong and often unjustified distributional assumptions. In this paper, we propose using survey response times to solve that problem. The key assumption of our approach is that individual response time is decreasing in the distance between the value of the latent variable and an indecision threshold. This assumption is supported by a large body of evidence on chronometric effects in psychology, neuroscience, and economics. We provide conditions under which the expected value of the latent variable (e.g., average happiness) can be compared across groups, even without making distributional assumptions. We apply our method to an online survey experiment and obtain some evidence that happiness follows distributions for which traditional regression analysis is valid.