Logistic Regression Using SAS®: Theory and Application

Book description

If you are a researcher or student with experience in multiple linear regression and want to learn about logistic regression, this book is for you! Informal and nontechnical, this book both explains the theory behind logistic regression and looks at all the practical details involved in its implementation using the SAS System. Several social science real-world examples are included in full detail. This book also explains the differences and similarities among the many generalizations of the logistic regression model. The following topics are covered: binary logit analysis, logit analysis of contingency tables, multinomial logit analysis, ordered logit analysis, discrete-choice analysis with the PHREG procedure, and Poisson regression. Other highlights include discussions on how to use the GENMOD procedure to do loglinear analysis and GEE estimation for longitudinal binary data. Only basic knowledge of the SAS DATA step is assumed.

Supports releases 6.12 and higher of SAS software.

Table of contents

  1. Copyright
  2. Acknowledgments
  3. Introduction
    1. What This Book Is About
    2. What This Book Is Not About
    3. What You Need to Know
    4. Computing
    5. References
  4. Binary Logit Analysis: Basics
    1. Introduction
    2. Dichotomous Dependent Variables: Example
    3. Problems with Ordinary Linear Regression
    4. Odds and Odds Ratios
    5. The Logit Model
    6. Estimation of the Logit Model: General Principles
    7. Maximum Likelihood Estimation with PROC LOGISTIC
    8. Maximum Likelihood Estimation with PROC GENMOD
    9. Interpreting Coefficients
  5. Binary Logit Analysis: Details and Options
    1. Introduction
    2. Confidence Intervals
    3. Details of Maximum Likelihood Estimation
    4. Convergence Problems
    5. Multicollinearity
    6. Goodness-of-Fit Statistics
    7. Statistics Measuring Predictive Power
    8. Predicted Values, Residuals, and Influence Statistics
    9. Latent Variables and Standardized Coefficients
    10. Probit and Complementary Log-Log Models
    11. Unobserved Heterogeneity
    12. Sampling on the Dependent Variable
  6. Logit Analysis of Contingency Tables
    1. Introduction
    2. A Logit Model for a 2 × 2 Table
    3. A Three-Way Table
    4. A Four-Way Table
    5. A Four-Way Table with Ordinal Explanatory Variables
    6. Overdispersion
  7. Multinomial Logit Analysis
    1. Introduction
    2. Example
    3. A Model for Three Categories
    4. Estimation with CATMOD
    5. Estimation with a Binary Logit Procedure
    6. General Form of the Model
    7. Contingency Table Analysis
    8. CATMOD Coding of Categorical Variables
    9. Problems of Interpretation
  8. Logit Analysis for Ordered Categories
    1. Introduction
    2. Cumulative Logit Model: Example
    3. Cumulative Logit Model: Explanation
    4. Cumulative Logit Model: Practical Considerations
    5. Cumulative Logit Model: Contingency Tables
    6. Adjacent Categories Model
    7. Continuation Ratio Model
  9. Discrete Choice Analysis
    1. Introduction
    2. Chocolate Example
    3. Model and Estimation
    4. Travel Example
    5. Other Applications
    6. Ranked Data
  10. Logit Analysis of Longitudinal and Other Clustered Data
    1. Introduction
    2. Longitudinal Example
    3. GEE Estimation
    4. Fixed-Effects with Conditional Logit Analysis
    5. Postdoctoral Training Example
    6. Matching
    7. Mixed Logit Models
    8. Comparison of Methods
    9. A Hybrid Method
  11. Poisson Regression
    1. Introduction
    2. The Poisson Regression Model
    3. Scientific Productivity Example
    4. Overdispersion
    5. Negative Binomial Regression
    6. Adjustment for Varying Time Spans
  12. Loglinear Analysis of Contingency Tables
    1. Introduction
    2. A Loglinear Model for a 2 × 2 Table
    3. Loglinear Models for a Four-Way Table
    4. Fitting the Adjacent Categories Model as a Loglinear Model
    5. Loglinear Models for Square, Ordered Tables
    6. Marginal Tables
    7. The Problem of Zeros
    8. GENMOD versus CATMOD
  13. Appendix
    1. The ROBUST Macro
    2. The PENALTY Data Set: Outcomes for Murder Defendants
    3. The WALLET Data Set: Altruistic Behavior by College Students
    4. The TRAVEL Data Set: Transportation Choices to Australian Cities
    5. The JUDGERNK Data Set: Rankings of Seriousness of Murder Cases
    6. The PTSD Data Set: Psychological Distress among Fire Victims
    7. The POSTDOC Data Set: Postdoctoral Training among Biochemists
    8. The CASECONT Data Set: Women in Homeless Shelters
    9. The PROGNOSI Data Set: Physicians’ Utterances about Prognosis
  14. Books Available from SAS® Press
    1. JMP® Books
  15. Wiley Series in Probability and Statistics
  16. Wiley Series in Probability and Statistics
  17. References
  18. Index

Product information

  • Title: Logistic Regression Using SAS®: Theory and Application
  • Author(s): Paul D. Allison
  • Release date: March 1999
  • Publisher(s): SAS Institute
  • ISBN: 9781580253529