Longitudinal and Incomplete Data

From Wednesday 26 March 2014 -  09:30am To Thursday 27 March 2014 - 05:00pm
Location Royal Statistical Society 12 Errol Street London EC1Y 8LX

Presented by Geert Verbeke and Geert Molenberghs

 

The course begins with a brief presentation of linear mixed models for continuous hierarchical data. The course focus is from the modeller’s perspective and on applications. Emphasis will be on model formulation, parameter estimation, and hypothesis testing, as well as on the distinction between the random-effects (hierarchical) model and the implied marginal model. Models for non-Gaussian data will be discussed, with a strong emphasis on generalized estimating equations (GEE) and the generalized linear mixed model (GLMM). A brief review of the classical generalized linear modelling framework will be presented. Similarities and differences with the continuous case will be discussed. The differences between marginal models, such as GEE, and random-effects models, such as the GLMM, will be explained in detail. Focus will be primarily on binary outcomes, however, GEE and GLMM model formulations will also be covered. When analysing hierarchical and longitudinal data, one is often confronted with missing observations, i.e. scheduled measurements have not been made, due to a variety of (known or unknown) reasons. It will be shown that, if no appropriate measures are taken, missing data can seriously jeopardize results, and interpretation difficulties are bound to occur. Methods to properly analyse incomplete data, under flexible assumptions, will be presented.

 

All topics will be illustrated with worked examples using SAS. While there are no hands-on practical sessions, the course notes include worked examples with annotated programs and output from SAS, discussed in such a way that they are also of use to non-SAS-users.

 

For more information and to register, visit our courses website