"Allison has done a remarkable job in distilling a vast literature on the analysis of panel data, including continuous and categorical outcomes. For those familiar with these models, Allison provides an invaluable guide to using SAS to estimate these models. For those just beginning to use panel models, Allison provides a sophisticated yet accessible guide to understanding the key issues in the analysis of panel data. Even though the focus is on fixed effects models, the book provides detailed information on random coefficient models as well."
"The power of fixed effects models comes from their ability to control for observed and unobserved time-invariant variables that might confound an analysis. As knowledge of this feature of fixed effects models has diffused, so has the interest in using these methods. One obstacle to further use has been accessible and consolidated information on fixed effects methods in diverse models such as linear regression, categorical and count regression, and event history models. A second obstacle to wider use has been having the knowledge of the software to implement these techniques.
Paul Allison's Fixed Effects Regression Methods for Longitudinal Data Using SAS guide goes a long way toward eliminating both barriers. This is a clear, well-organized, and thoughtful guide to fixed effects models. There are separate chapters devoted to linear regression, categorical response variables, count data, and event history models. These represent the most widely used models in the social sciences. In a brief monograph, Allison is able to present the essentials of fixed effects for each model and the appropriate procedures in SAS that can implement them. Empirical examples and SAS code are included, making it easier for the reader to implement these methods... In sum, Paul Allison has produced a terrific guide to fixed effects models and their estimation using SAS. I highly recommend it."
"Fixed Effects Regression Methods for Longitudinal Data Using SAS represents an excellent piece of work--it is clear, coherent, well-structured, and useful, and it has a sense of logical flow not always found in efforts of this sort. To say that I was impressed with this book would be an understatement.
What I especially liked about the book was how the author is able to fluidly mix clear and accurate explanations of statistical concerns and procedures with specific directions for how to go about these procedures in SAS. It merits observing that even researchers or students not thoroughly versed in the statistical underpinnings or mathematical complexities will be able to analyze and interpret their data using the directions provided. The author even provides sample outputs and takes the reader through a scholarly interpretation of results."
"I found this book to be clear and easy to follow in terms of describing which SAS procedures apply to a given set of data in order to answer a given question. Procedures addressed in the book include REG, GLM, TSCSREG, GENMOD, MIXED, LOGISTIC, PHREG, and CALIS. In cases where multiple procedures give the same answer, examples are given in the book, with an indication of which procedures are quicker and more efficient. The difference between fixed effects models and random effects models is made very clear, and throughout the book examples of how each method would be applied to a given problem are shown, even though, as the title suggests, the major stress is on fixed effects models. Applications for linear regression, categorical response variables, count data, event history analysis, and linear structural equation models with latent variables each have a chapter dedicated to them. Treatment of each subject is very complete, ranging from the underlying mathematical formulae to examples of SAS code and output, yet remarkably clear, requiring only minimal prior experience with either theoretical statistics or SAS practice.
This is a book which I will keep close at hand because I plan to refer to it quite frequently in my work."