one way compare two three way anova

[deleted_user]

hello all,

I have an experiment with three factors and two levels each one.

When i am running a reduced glm model I am having statistical differences, but in order to get a t groupping i am running the glm model also as one way anova, where there are not statistical differences between the treatments.

Is there any easy way to run post hoc test for three way anova and to get t-grouping?

[Doc_Duke]

You want to do a three-way ANOVA on a two-way design? I don't think that you can.

You can do post-hoc tests in GLM with the TEST statement. I find it a bit more flexible than using the ANOVA procedure.

You may also want to repost this question on the SAS Statistical Procedures forum for a broader viewing. Please include some code samples and more specificity in your question.