Overview
P. Richard Hahn is a statistician at the University of Chicago Booth School of Business, working on Bayesian nonparametrics, variable selection, and high-dimensional factor models. He is a co-developer of the sparse factor analytic probit model for high-dimensional binary outcomes and its application to congressional voting patterns.
Key Contributions / Features
- Co-developed the sparse factor analytic probit model with spike-and-slab priors on factor loadings and automatic multiplicity correction via Beta(1,1) inclusion probabilities (Hahn, Carvalho, Scott 2012)
- Model: z_i ~ N(α + λγ_i + Bf_i, I_p); Σ = BVB^T + I with B lower triangular, free factor covariance V fosters sparse B
- Applied model to US Senate roll-call votes 1949–2009, documenting rising partisanship from 1979 onward
Related