Generalized Linear Models: An Applied Approach
Author:
|
John P. Hoffmann |
| Publisher: |
Pearson |
| Copyright: |
2004 |
| ISBN-13: |
978-0-205-37793-0 |
| Pages: |
204; paperback |
| Price: |
$69.75 |
|
|
|
|
Comment from the Stata technical group
Generalized Linear Models: An Applied Approach, by John Hoffmann,
presents the reader with an applied tour through the world of generalized
linear models. Using real-world datasets, the author discusses a wide class
of models, organizing the material according to what is to be assumed about
the dependent variable, whether it be continuous, discrete, categorical,
ordered, count, or time to failure. As such, this book is ideal for
researchers wishing to apply these models without having to endure the
detailed discussions of statistical theory or computational algorithms
usually associated with GLIMs. The author focuses instead on the
statistical reasoning behind the different models and in the interpretation
of computer output, which for the most part is obtained using Stata.
After a brief review of the simplest of GLIMs, the linear regression model,
and a review of GLIM terminology, the text moves on to covering the various
other models, including logistic/probit, ordered response, multinomial
logit, count data, and survival or time to failure. A final chapter
discussing advanced issues, such as sample selection and endogeneity, is
also included.
Table of contents
Preface
1 A Review of the Linear Regression Model
Issues of Interest
How to Estimate a Linear Regression Model
A Detailed Example of an OLS Regression Model
The Assumptions of the OLS (Linear) Regression Model
Interaction Terms in the OLS (Linear) Regression Model
Conclusion
Exercises
2 Introduction to Generalized Linear Models
The Role of the Link Function
The Binomial Distribution
The Multinomial Distribution
The Poisson Distribution
The Negative Binomial Distribution
How Do We Estimate Regression Models Based on These Distributions?
How to Check the Significance of Coefficients and the "Fit" of the Model
Conclusion
Exercises
3 Logistic and Probit Regression Models
What Are the Alternatives to the Linear Regression Model?
The Logistic Regression Model
What about a More Sophisticated Model?
The Probit Regression Model
Diagnostic Tests for the Logistic Regression Model
Conclusion
Exercises
4 Ordered Logistic and Ordered Probit Regression Models
Alternative Models for Ordinal Dependent Variables
The Ordered Logistic Regression Model
Testing the Proportional Odds Assumption
The Ordered Probit Regression Model
Introducing Multiple Independent Variables
Conclusion
Exercises
5 The Multinomial Logistic Regression Model
Introducing Multiple Independent Variables
Diagnostic Tests for the Multinomial Logistic Regression Model
Alternatives to the Multinomial Logistic Regression Model
Conclusion
Exercises
6 Poisson and Negative Binomial Regression Models
The Poisson Regression Model
The Overdispersed Poisson Regression Model
The Negative Binomial Regression Model
Diagnostic Tests for the Poisson Regression Model
Other Models for Count Variables
Conclusion
Exercises
7 Event History and Survival Models
Continuous- versus Discrete-Time Models
Censoring and Time-Dependent Covariates
The Basics: Survivor and Hazard Functions and Curves
Parametric Event History Models
The Cox Proportional Hazards Model
Discrete-Time Event History Models
Conclusion
Exercises
8 Where Do We Go from Here?
Sample Selection
Endogeneity
Longitudinal Data
Multilevel Models
Nonparametric Regression
Conclusion
Appendix
SPSS, SAS, and Stata Programs for Examples in Chapters
References
Index
|
){kind=link}
){kind=link}