A flexible Bayesian panel data model for analyzing binary response data using Pólya-Gamma random variables

Roman Pfeiler, Institute of Applied Statistics JKU Linz

Co-authors: Helga Wagner, Johannes Kepler University Linz

Abstract: We consider a Bayesian panel data model with time-varying parameters for binary response data and shrinkage priors to avoid overfitting. The model is evaluated in a simulation study and then applied to analyze depressive symptoms among adolescents.