Applied Longitudinal Data Analysis for Medical Science

Essential for non-statisticians and researchers working with longitudinal data from medical studies, this updated new edition discusses the most important techniques available for analysing data of this type. Using non-technical language, the book explores simple methods such as the paired t-test and summary statistics as well as more sophisticated regression-based methods, including mixed model analysis. The emphasis of the discussion lies in the interpretation of the results of these different methods, covering data analysis with continuous, dichotomous, categorical and other outcome variables. Datasets used throughout the book are provided, enabling readers to re-analyse the examples as they make their way through chapters and improve their understanding of the material. Finally, an extensive and practical overview of, and comparison between, different software packages is provided. Readers will be able to use this book as a practical manual in their everyday work without needing a strong background in statistics.

• An applied approach to longitudinal data analysis and the use of non-technical language allows readers to use the book as a practical manual in their everyday work, without the need for a strong background in statistics
• The text is supplemented with figures and examples of computer software outputs to aid reader understanding of the techniques presented and allow them to compare available software packages
• Datasets used throughout the book are provided, enabling readers to re-analyse the examples as they make their way through chapters and improve their understanding of the material
• Material has been updated to ensure explanations are as clear as possible and several new chapters are included, discussing the use of hybrid models, causality in observational longitudinal studies and the analysis of outcome variables with floor or ceiling effects