- Hardcover: 752 pages
- Publisher: Wiley; 3 edition (December 3, 2012)
- Language: English
- ISBN-10: 9780470463635
- ISBN-13: 978-0470463635
- ASIN: 0470463635
- Product Dimensions: 7.4 x 1.6 x 10.4 inches
- Shipping Weight: 3.2 pounds (View shipping rates and policies)
- Average Customer Review: 26 customer reviews
- Amazon Best Sellers Rank: #89,434 in Books (See Top 100 in Books)
Enter your mobile number or email address below and we'll send you a link to download the free Kindle App. Then you can start reading Kindle books on your smartphone, tablet, or computer - no Kindle device required.
To get the free app, enter your mobile phone number.
Categorical Data Analysis 3rd Edition
Use the Amazon App to scan ISBNs and compare prices.
Frequently bought together
Customers who bought this item also bought
From the Back Cover
Praise for the Second Edition
"A must-have book for anyone expecting to do research and/orapplications in categorical data analysis."
—Statistics in Medicine
"It is a total delight reading this book."
"If you do any analysis of categorical data, this is anessential desktop reference."
The use of statistical methods for analyzing categorical datahas increased dramatically, particularly in the biomedical, socialsciences, and financial industries. Responding to new developments,this book offers a comprehensive treatment of the most importantmethods for categorical data analysis.
Categorical Data Analysis, Third Edition summarizes thelatest methods for univariate and correlated multivariatecategorical responses. Readers will find a unified generalizedlinear models approach that connects logistic regression andPoisson and negative binomial loglinear models for discrete datawith normal regression for continuous data. This edition alsofeatures:
- An emphasis on logistic and probit regression methods forbinary, ordinal, and nominal responses for independent observationsand for clustered data with marginal models and random effectsmodels
- Two new chapters on alternative methods for binary responsedata, including smoothing and regularization methods,classification methods such as linear discriminant analysis andclassification trees, and cluster analysis
- New sections introducing the Bayesian approach for methods inthat chapter
- More than 100 analyses of data sets and over 600 exercises
- Notes at the end of each chapter that provide references torecent research and topics not covered in the text, linked to abibliography of more than 1,200 sources
- A supplementary website showing how to use R and SAS; for allexamples in the text, with information also about SPSS and Stataand with exercise solutions
Categorical Data Analysis, Third Edition is an invaluabletool for statisticians and methodologists, such as biostatisticiansand researchers in the social and behavioral sciences, medicine andpublic health, marketing, education, finance, biological andagricultural sciences, and industrial quality control.
About the Author
ALAN AGRESTI is Distinguished Professor Emeritus in the Department of Statistics at the University of Florida. He has presented short courses on categorical data methods in thirty countries. He is the author of five other books, including An Introduction to Categorical Data Analysis, Second Edition and Analysis of Ordinal Categorical Data, Second Edition, both published by Wiley.
Try the Kindle edition and experience these great reading features:
Showing 1-6 of 26 reviews
There was a problem filtering reviews right now. Please try again later.
First off, this book is another must have for those who will not only practice statistical modeling of categorical data, but who also need to be able understand the mathematical underpinnings of the GLM and the Generalized Linear Mixed Model (GLMM). Chapter 4, in particular, reviews the components of the GLM - systematic and random components, and the link function - and then goes on to explain what how these three differ depending on the type of your response ("y"): continuous data, binary data, count data, and so fourth. The next few chapters then go into detail on more of the application of the GLM for each of these type of data, followed by chapters discussing different models for GLMM - when your response variable is correlated. Finally, the book reviews asymptotic theory (chapter 16) which is vital for understanding how sample size affects the results for these models. For me, this book is does a better job at explaining these kind of topics (particularly, the Delta Method) than Casella, which is also an important book. Despite the emphasis on theory, there are applied examples and problems and the author does provide code for R, SAS, and Stata on his website.
The main differences between this version and the second version are threefold. First, there are two additional chapters: one chapter devoted to other ways to model binary data and one based on other model approaches to classifying and clustering data. For the former, Agresti expanded coverage on probit regression, which is like a logistic regression but using a different distribution assumption. It also discusses topics that have become of interest in the last ten to fifteen years such as high dimensional analysis. The other new chapter discusses methods such as decision trees and linear discriminant analysis. Next, some chapters also include a discussion on bayesian approaches to the GLM, including the alternative approach to modeling binary data that I mentioned. Finally, the problem sets have been partly changed. I feel like the classification and clustering topics are better covered in a book like Johnson's Multivariate Data Analysis, and much of the new material feels more like an overview than a rigorous discussion of the math like other parts of the book. On the positive side, I thought that the section on probit regression was good because it provided insight on its importance in latent modeling.
Despite my critiques this is still an excellent book to own. If you already own the second edition, I can't recommend purchasing this book but if you don't then you definitely should.
For me, how Agresti wrote it is the most efficient way to get the needed information into my brain. Best stats book I've read.
Frist, if you are a self-learner, you'd better choose another textbook with a more smooth learning curve. In some of the chapters, you might encounter some classical or even deemed "trivial" examples to illustrate some basic concepts, then delve into the technical part where has nothing but the mathematical proofs. Though proofs in this book are not assuming the readers have mastery of measure theory, many of them require readers to have a grasp of exponential family, canonical link, which are often introduced in graduate-level statistical classes. Another textbook " An introduction to categorical data analysis", by the same author, will be a better choice.
Also, this textbook is too oriented towards the application of statistical methods to bio-industries, which I find it not very enjoyable. I admit that similar approaches can also be applied to other fields as long as you have a deep understanding of them. But for people who just want learn methods that they can immediately put them into practice. This book doesn't serve well for them.
Last but not least, the practice problems don't relate to the textbook materials well. For sure, you can easily crack the first few practice problems without even reading that chapter. As you move on, many of the questions are becoming quite formidable. One reason is that some of them were drawing from Agresti's research, which assumes a high-level mastery of statistical methods and probability.
I'd like to see a few changes, as I mentioned above could be incorporated into the fourth edition. If so, this book is an unquestionable 5 stars textbook.