JMP 13 Online Documentation (English)
Discovering JMP
Using JMP
Basic Analysis
Essential Graphing
Profilers
Design of Experiments Guide
Fitting Linear Models
Predictive and Specialized Modeling
Multivariate Methods
Quality and Process Methods
Reliability and Survival Methods
Consumer Research
Scripting Guide
JSL Syntax Reference
JMP iPad Help
JMP Interactive HTML
Capabilities Index
JMP 12 Online Documentation
Fitting Linear Models
•
Generalized Linear Models
• The Generalized Linear Model Personality
Previous
•
Next
The Generalized Linear Model Personality
Generalized linear models are fit as a personality of the Fit Model launch window. After selecting
Analyze > Fit Model
, select
Generalized Linear Model
from the drop-down menu before or after assigning the effects to the model.
Generalized Linear Model Launch Window
When you specify that you are fitting a generalized linear model, the Fit Model launch window changes to allow you to select a Distribution and a Link Function. Options for Overdispersion Tests and Intervals and for Firth Bias-adjusted Estimates also appear. In addition, an Offset button is added to the Fit Model window.
Description of Fit Model options for Generalized Linear Models
Distribution
Assigns the appropriate response probability distribution to the model. See
Examples of Generalized Linear Models
.
Link Function
A link function relates the linear model to the response variable. See
Examples of Generalized Linear Models
.
Overdispersion Tests and Intervals
Fits a model that includes an overdispersion parameter.
Firth Bias-adjusted Estimates
This MLE method has been shown to produce better estimates and tests than MLEs without bias correction. In addition, bias-corrected MLEs ameliorate separation problems that tend to occur in logistic-type models. Refer to Heinze and Schemper (2002) for a discussion of the separation problem in logistic regression.
Offset
Often used in Poisson regression with the log link function to account for exposure.