Class Variables

Specify one or more variables whose levels form distinct categories in the model, in this field.

List all variables whose levels form distinct categories in the model in the Class Variables field. Indicator variables are created for each level of each variable. If you do not specify an effect, then the numerical values of the variables are used directly in the model as in linear regression.

Note: The order in which you specify class variables determines how LSMeans and/or interaction effects are sorted.

This option is available only when the Response Type is an incidence density.

To Specify Class Variables:

Make sure that either Incidence density based on first event only or Incidence density based on all events is selected as the Response Type.

Click to open the Add window (shown below) that lists each variable contained in ADSL or the domain and SUPPxx (if merged) specific to this report.

Select the variables you want to include in the analysis and click OK to add those variables to the text box.

In this example, SEX was selected and added.

Refer to the SAS PROC GLIMMIX documentation for more information.

Caution: Specifying a very large number of variables or using a data set with a large number of variables can prohibitively increase the time needed to run this report.