ABSTRACT
Data collected in psychiatric longitudinal studies are complex because outcomes are rarely directly observed, there are multiple correlated repeated measures within individuals, and there is natural heterogeneity in treatment responses and other characteristics in the populations. This conclusion presents some closing thoughts on the concepts covered in the preceding chapters of this book. The book discusses simple statistical methods that do not work well with such data. To facilitate understanding and increase appreciation of the versatility of these methods, the book presents, at a non-technical level, several approaches for the analysis of correlated data, namely, mixed-effects and generalized estimating equation (GEE) models, mixture models for longitudinal data, and non-parametric methods for repeated measures studies. The book focuses on several important topics in the analysis of longitudinal and clustered data that deserve special attention.
