Antedependence Models for Longitudinal Data

Antedependence models are useful statistical models of the covariance structure of continuous multivariate data with some natural labeling order, i.e., longitudinal data. Although underutilized, they are more flexible than standard models as they allow for heterogeneous variances and nonstationary correlations. This book presents a standard framework for antedependence models that highlights their usefulness and encourages their application. It brings together important properties of these models and associated statistical methods that are scattered throughout the literature. The methodology is supported by a range of illustrative examples. Data sets and computer code are available on the web.