If you are going to model only the existing data, then PROC GLM is a good starting place. See this: https://support.sas.com/documentation/onlinedoc/stat/141/glm.pdf . Now look at the section called Getting Started. The first example is PROC GLM for Unbalanced ANOVA. Once you go through that example, you could try it on your data.
SteveDenham
... View more