Multilevel modelling of health statistics (Q2771722)

This is an introductory text to multilevel modelling aimed at researchers in the health sciences. Therefore, while there are good introductions to multilevel models, and plenty of examples, a discussion of how to estimate these models is absent, except for extensive references to a computer software package called Mlwin. Each chapter has its own author(s). NEWLINENEWLINENEWLINEThe first chapter is an introduction to multilevel modelling in the framework of the ordinary linear model. The following chapters then complement with applications, such as repeated measurements, binomial regression, Poisson regression, multivariate regression, outliers, non-hierarchical models, multinomial regression, institutional performance, and spatial analysis. Most of the discussion is done by examples, and the reader is shown how the analysis is done with the aid of computer software, mostly Mlwin. Then one chapter discusses the analogy between multilevel analysis and multistage sampling, a second follows up with pieces not earlier covered. Finally, a comparison between available (2001) software for multilevel modelling is done in the last chapter.