Proc GLM

This procedure is used to do regression, analysis of variance, analysis of covariance, and multivariate analysis of variance. You can use one or several independent and dependent variables, either categorical (discrete) or continuous variables in the model. 

I was writing for myself some notes on a sample data I practiced the GLM procedure. I have attached the note file at the bottom of this page, I thought might be useful for someone.

* The first group in the data used below is control group and then there are two levels of group 1 and two levels of group 2;

proc glm data=dataname; 

class group; 

model var1-var4 = group 

                / nouni 

                  alpha=0.1 

                  clm

                  cli

                  p 

                  noint

                  solution; 

means group 

          / tukey

            LSD

            hovtest

            welch

            alpha=0.1 

            cldiff; 

contrast 'Two levels of group 2' Scenario 0 0 0 -1 1; 

estimate 'Difference of first and second groups' scenario 0 1 1 -1 -1; 

repeated myvariable 4 contrast / mean summary; 

lsmeans scenario / out = means; 

output out = dataname2 

             predicted=PredValues 

             residual=Resvalues 

             stdi=IndivPredSTD 

             stdp=MeanPredSTD 

             stdr=ResSTD; 

run;