The Regenwetter Decision Making Lab tackles a three-pronged challenge in decision research.
1) Over the past half-century, behavioral decision research has struggled with a profound conceptual problem in measurement: How can we bridge the gap between the deterministic nature of algebraic utility theory and the probabilistic nature of random sample choice data?
2) There is little dialogue and cross-fertilization between the research community in individual decision making and that in social choice, even though behavioral decision researchers routinely aggregate individual choice data before they analyze them.
3) Few social choice theorists have studied consensus methods from a descriptive, behavioral perspective and few behavioral decision researchers have submitted mathematical properties of social choice procedures to the empirical test.
RECONCILING THEORY AND DATA. The research team systematically applies a broad range of solutions that mathematically link algebraic theories to empirical random sample data. We achieve this through a multitude of alternative probabilistic specifications of deterministic theories, many of them new. All of these specifications interface with recent breakthrough developments in order constrained statistical inference. The team aims to raise the methodological sophistication of future behavioral decision research by eliminating important sources of artifacts caused by unsound data analysis.
RECONCILING INDIVIDUAL AND SOCIAL CHOICE. Our systematic program of theory comparison integrates quantitative models of individual and collective choice.
BEHAVIORAL SOCIAL CHOICE. The lab is engaged in a long term effort to complement social choice theory with behavioral analyses of consensus methods. Our past work has questioned the scope of the famous Concorcet paradox’s empirical and policy relevance. The team is evaluating how well the famed mutual theoretical incompatibility of consensus methods is empirically validated by survey, ballot, and/or experimental choice data. Prior work on this important question has been extremely sparse.