Marginal additive models for population‐averaged inference in longitudinal and cluster‐correlated data

Abstract

AbstractWe propose a novel marginal additive model (MAM) for modeling cluster‐correlated data with nonlinear population‐averaged associations. The proposed MAM is a unified framework for estimation and uncertainty quantification of a marginal mean model, combined with inference for between‐cluster variability and cluster‐specific prediction. We propose a fitting algorithm that enables efficient computation of standard errors and corrects for estimation of penalty terms. We demonstrate the proposed methods in simulations and in application to (a) a longitudinal study of beaver foraging behavior and (b) a spatial analysis of Loa loa infection in West Africa.