A rigorous, systematic presentation of modern longitudinal analysis Longitudinal studies, employing repeated measurement of subjects over time, play a prominent role in the health and medical sciences as well as in pharmaceutical studies. An important strategy in modern clinical research, they provide valuable insights into both the development and persistence of disease and those factors that can alter the course of disease development. Written at a technical level suitable for researchers and graduate students, Applied Longitudinal Analysis provides a rigorous and comprehensive description of modern methods for analyzing longitudinal data. Focusing on General Linear and Mixed Effects Models for continuous responses, and extensions of Generalized Linear Models for discrete responses, the authors discuss in detail the relationships among these different models, including their underlying assumptions and relative merits. The book features: * A focus on practical applications, utilizing a wide range of examples drawn from real-world studies * Coverage of modern methods of regression analysis for correlated data * Analyses utilizing SAS(r) * Multiple exercises and "homework" problems for review An accompanying Web site features twenty-five real data sets used throughout the text, in addition to programming statements and selected computer output for the examples.
So technically I only read about 2/3 of this book since that was what was assigned for class. Its pretty difficult reading, but I still liked the approach for the most part. Even though reading it helped me realized that I don't have the data to be doing longitudinal analysis right now, it demonstrated very clearly how several other approaches (MLM, TSCS) are related to or build on the ideas and math of longitudinal. I'm definitely going to need to come back to this book to work through the math and probably use some other texts that have political science examples (since this is exclusively health sciences based), but I got more out of it on the first go round than I normally do with stats text books.