Model Selection and Goodness-of-t Tests for Longitudinal Data

Abstract: For complex high dimensional longitudinal data the full parametric likelihood function of the model is often dicult to specify. A semiparametric model which is dened by a set of mean zero estimating functions might be useful for estimation. It can also occur that there are more moment conditions than estimable parameters in longitudinal data settings. We will illustrate how to select moment conditions which contribute to parameter of interest estimations, and how to perform hypothesis testing for model selections. References [1] Qu, A., Lindsay, B.G. and Li, B. (2000). Improving Generalised Estimating Equations using Quadratic Inferenc