Generalized Linear Models

Fit Models for Nonnormal Response Distributions

Generalized linear models provide a unified way to fit responses that do not fit the usual requirements of traditional linear models. For example, frequency counts are often characterized as having a Poisson distribution and fit using a generalized linear model.

The Generalized Linear Model personality of the Fit Model platform enables you to fit generalized linear models for responses with binomial, normal, Poisson, or exponential distributions. The platform provides reports similar to those that are provided for traditional linear models. The platform also accommodates separation in logistic regression models using the Firth correction.

Figure 12.1 Example of a Generalized Linear Model Fit 

Contents

Overview of the Generalized Linear Model Personality

Example of a Generalized Linear Model

Launch the Generalized Linear Model Personality

Generalized Linear Model Fit Report

Whole Model Test

Generalized Linear Model Fit Report Options

Additional Examples of the Generalized Linear Models Personality

Using Contrasts to Compare Differences in the Levels of a Variable

Poisson Regression with Offset

Normal Regression with a Log Link

Statistical Details for the Generalized Linear Model Personality

Model Selection and Deviance