10-09-2013 03:52 PM
I am SAS illiterate and trying to finish my thesis (our departmental statistician retired). Having trouble collecting all necessary data for a strip-split plot. I need to analyze all of the main factors and their interactions. Previously, I was getting a minimum significant difference. Can anyone tell me if I am fully analyzing the data?
Whole plot factor: treatment
Based on information I found online and previous codes, this is what I've come up with:
proc glm data = visual;
class rep timing trt mowing;
model visual = timing|trt|mowing;
lsmeans timing|trt|mowing / pdiff adjust=BON cl lines;
10-10-2013 08:39 AM
Stop. PROC GLM will not correctly calculate standard errors for split plot models. Do whatever it takes to get a copy of Littell et al.'s SAS for Mixed Models, 2nd ed., and go to chapter 4 on split-plot models. Also check out Chapter 16, which has case studies, especially 16.4 A Split-Plot Experiment with Correlated Whole Plots, 16.5 A Complex Split Plot: Whole Plot Conducted as an Incomplete Latin Square, and 16.6 A Complex Strip-Split-Split Plot Example.
Try rewriting your model using PROC MIXED. If you have problems, please post your code and the problems you are having. It would probably help to post any follow-up in the SAS Statistical Procedures forum, where the participants tend to be more focused on statistical issues. Good luck.