Categorical Data Analysis Using SAS, Third Edition, 3rd Edition

Chapter 9

Logistic Regression II: Polytomous Response

Contents

9.1   Introduction

9.2   Ordinal Response: Proportional Odds Model

9.2.1   Methodology

9.2.2   Fitting the Proportional Odds Model with PROC LOGISTIC

9.2.3   Multiple Qualitative Explanatory Variables

9.2.4   Partial Proportional Odds Model

9.3   Nominal Response: Generalized Logits Model

9.3.1   Methodology

9.3.2   Fitting Models to Generalized Logits with PROC LOGISTIC

9.3.3   Generalized Logit Model with Continuous Explanatory Variable

9.3.4   Exact Methods for Generalized Logits Model

9.1 Introduction

Logistic regression most often involves modeling a dichotomous outcome, but it also applies to multilevel responses. The response might be ordinal (no pain, slight pain, substantial ...