Multivariate Data and GLM: Generalized Estimating Equations
摘要
Generalized linear models (GLMs) have been extended to handle multivariate data arising from longitudinal or repeated measures using various approaches: Quasi-likelihood, working likelihood, and pseudo-likelihood methods play a crucial role. Among these, quasi-likelihood is commonly used for analyzing longitudinal data. The generalized estimating equations (GEE) approach, a quasi-likelihood method, addresses correlated responses over time. Researchers have proposed alternative procedures to enhance the analysis of multivariate data in this context. Few of these methods are discussed in this chapter.