Handleman Professor of Marketing, Professor of Statistics (by courtesy)
He/Him/His
About
My research deals primarily with discrete choice models: using decisions people actually make to extract information about covariates of interest (demographics, product attributes, socioeconomics, etc.) Some of this falls under the umbrella of "Big Data", and the main methodological approach is Bayesian, with a particular emphasis on HB (Hierarchical Bayes) models. I'm interested in both pure / theory-based and applied work in statistics. In terms of the former, recent emphases are on algorithms to allow Bayesian inference on very large data sets; the formulation of dyadic utility theory; finding excellent starting values for high-dimensional MCMC algorithms; and characterizations of optimal search in uncertain environments. Applications usually involve large-scale (marketing-based) choice data, for example, click-throughs on web sites; reactions to "menus" of options; analysis of dating site data; geographic / spatial modeling; and analyzing discrete survey responses. I'm also interested in the intersection of statistical and marketing models with engineering and operations management, particularly in modeling reaction to intangible and aesthetic attributes.
Education
Massachusetts Institute of Technology, Ph.D., Sloan School of Management (1989)
Cornell University, doctoral program in Mathematics (1983-84)
Massachusetts Institute of Technology, S.B. Mathematics; S.B. Philosophy (1983)
Research Areas of Interest
Bayesian estimation and marketing models
Dynamic models, optimal stopping / cutoffs, ordinal statistics and processes
Decision-making under uncertainty, mathematical psychology
Choice theory, dyadic choice
