Congratulations to Jake Maronge for winning the JSM Travel Award from the Biometrics Section of the American Statistical Association. Jack won this award for a paper titled “Generalized case-control sampling under generalized linear models,” one of the main points of his dissertation research.
Jake’s committee includes Paul Rathouz, Guanhua Chen, Rick Chappell, Menggang Yu, and Jun Zhu.
We provide a novel approach for analyzing data arising from a generalized case-control (GCC) study. A GCC study is a type of two-phase study closely related to the well-known case-control study, wherein we allow for non-binary response distributions. We propose a method using a semiparametric extension of generalized linear models to analyze data from generalized case-control studies in the same way that logistic regression is used for case-control studies. Our methodology is developed over the course of three steps. First, we develop a method for estimation for studies with biased sampling and show possible efficiency gains from GCC studies. Second, we develop asymptotically correct inference. Finally, we extend these results to perform analysis for fixed sample size GCC studies. At the end, we demonstrate our method using data from the AHEAD study. Our approach allows the experimenter to analyze these complicated data in a way that will seem more natural for those familiar with practical analysis of traditional case-control data.