What is risk preference? Using data-driven ontologies and meta-analyses to inform and achieve expert consensus

Risk preference is a prominent concept in both economics and psychology and it is often employed as explanans (i.e., cause) of choice behavior in financial, health, and political domains. Given its central theoretical role, there is a surprising lack of consensus about the nature and measurement of risk preference and its relation to other constructs (e.g., impulsivity, cognitive control, time preference). This lack of a clear ontology (i.e., a set of concepts, respective properties, and conceptual and empirical inter-relations) requires both an explanation and a resolution. As it stands, the lack of conceptual and empirical clarity leads to multiplication of efforts in the use of similar concepts (e.g, risk preference, impulsivity) and associated operationalizations (e.g., gamble tasks vs. impulsivity questionnaires) that provide conflicting results about the power of such constructs and associated operationalizations to explain and predict real-world outcomes. This proposal aims to help answer the question “What is risk preference?" and thus help resolve these issues in three ways. First, it proposes mapping the potential different conceptual and measurement approaches over historical time and constructing data-driven ontologies that place risk preference in a dynamic historical context. Second, it aims to provide a quantitative summary of central characteristics of various risk preference measures, including their convergent validity and predictive validity using data from the extant literature as well as additional coordinated analysis of several large longitudinal panels. Third, the proposal aims to contribute to the establishment of expert consensus based on a discussion of the new insights provided by the ontologies and meta-analytic summaries provided by the first two components.The proposal promises to provide an overview of historical and current perspectives on risk preference that can be communicated to and discussed by experts in economics and psychology so as to help build a common research agenda. More broadly, the proposed approach provides a novel framework for questioning the conceptual status of extant theoretical constructs and developing interdisciplinary research agendas in the behavioral and social sciences.