•Each treatment should have exactly the same number of observations; every categorical outcome has the same number of observations.
•Caution: If you use PROC ANOVA for analysis of unbalanced data, you must assume responsibility for the validity of the results.
•Use PROC GLM instead.
No comments:
Post a Comment