Carnegie Mellon University
About the Webinar
In this webinar, Dr. Jay Kadane proposes a method to ascertain posterior distributions for the possibly non-independent sensitivity and specificity of binary tests without an assumed “gold standard” test. The key is to administer simultaneously more than one test to each patient. The goal is to offer a technique to analyze in vivo data from COVID-19 testing, particularly aiming at antigen tests, which give results quickly but do not always give results accurately. The proposed algorithm is a Gibbs sampler. It also returns a posterior distribution on the prevalence of the disease in the population studied. Missing data can be augmented within the algorithm if needed. Dr. Kadane exemplifies this algorithm by analysis of a chlamydia data set.
About Jay Kadane
Dr. Joseph B. ("Jay") Kadane is the Leonard J. Savage Professor of Statistics and Social Sciences at Carnegie Mellon University. Dr. Kadane is known for his contributions to Bayesian theory, econometrics, and a wide variety of fields of application. He is a cross-disciplinary statistician, having worked in economics, law, medicine, political science, sociology, computer science, archaeology, environmental science forensics, physics, and chemistry. He is a Fellow of the American Academy of Arts and Sciences, the American Statistical Association, the Institute of Mathematical Statistics, and the International Society for Bayesian Analysis.