ABSTRACT
We briefly review in this chapter a number of popular parametric and semiparametric models in longitudinal analysis. Because of their simple mathematical structures and interpretations, these models are often the first set of analytical tools to be used in a real study. The estimation and inference methods developed for these models form the foundation of longitudinal analysis. Since the nonparametric models to be presented in this book are flexible extensions of these parametric and semiparametric models, their local and global smoothing estimation methods with different modeling structures are motivated and generalized from the estimation methods in this chapter. We note that, because the topics of parametric and semiparametric longitudinal analysis have been extensively studied in the literature, the models and estimation methods reviewed here only represent a fraction of the available approaches. A more complete account of the most commonly used parametric and semiparametric methods in longitudinal analysis can be found in Fitzmaurice et al. (2009).
