Hal Stern

Overview

Hal Stern (Harold S. Stern) is a statistician at the University of California, Irvine (Department of Statistics). His research covers Bayesian modeling, model assessment, forensic statistics, and sports statistics. He is a co-author of Bayesian Data Analysis (with Gelman, Carlin, and Rubin).

Key Contributions

• Gelman-Meng-Stern (1996): Co-developed the posterior predictive check framework for Bayesian model assessment via realized discrepancies. Contributed the latent class mixture example (93 infants, Stern et al. 1995 temperament data). See Posterior Predictive Check.
• Stern-Arcus-Kagan-Rubin-Snidman (1995): Applied three-class mixture model to infant temperament data from the Kagan longitudinal study; the resulting dataset was the third example in GMS (1996).

Related

• Andrew Gelman
• Xiao-Li Meng
• Donald Rubin
• Posterior Predictive Check
• Mixture of Normals