Generalized Linear Models for Longitudinal Data

Abstract The analysis of longitudinal data needs to take into account the correlations between the repeated measures of the response variable to draw valid and efficient inferences about parameters of scientific interest. There are a variety of approaches that can be taken depending on what is the target of the estimation process. This article outlines marginal models, random effects models and transition models.