Overview of the Generalized Linear Model

A generalized linear model is an extension of a traditional linear model that allows the population mean to depend on a linear predictor through a nonlinear link function. A generalized linear model requires that you specify a distribution and a link function. The distribution should match the distribution of the response variable. The link function is used to relate the response variable to the effect variables.
The generalized linear model requires a measure response variable and at least one effect variable or interaction term. The distribution imposes range requirements on the measure response variable. These requirements are provided in the following table:
Distribution
Range Requirements
Beta
Values must be between 0 and 1, exclusive
Binary
Exponential
Nonnegative real values
Gamma
Nonnegative real values
Geometric
Positive integers
Inverse Gaussian
Positive real values
Negative Binomial
Nonnegative integers
Normal
Real values
Poisson
Nonnegative integers
Last updated: January 8, 2019