Still use **mpg** data, and say we wish to test whether mean of highway mpg (**hwy**) between the cars with different number of cylinders (**cyl**). There are multiple procedures in SAS to use, from PROC GLM to PROC GLIMMIX. So let use proc glm and proc glimmix as example:

proc glm data = mpg; class cyl; model hwy = cyl; means cyl; run;

proc glm data = mpg; class cyl; model hwy = cyl; means cyl; run;

The mean of the dependent variable differs significantly among the levels of **cyl**. The glimmix procedure also gives pair-wise comparison result.

*See Also:*