| Date: | Tue, 21 May 1996 16:23:16 GMT |
|---|
| Reply-To: | "Dwight H. Galster" <galster@RRNET.COM> |
|---|
| Sender: | "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU> |
|---|
| From: | "Dwight H. Galster" <galster@RRNET.COM> |
|---|
| Organization: | Red River Net - Internet Communications (701-232-3322/guest) |
|---|
| Subject: | help with split plot |
|---|
We have come across a situation which the textbooks don't seem to cover...
Maybe somebody out there can help.
We have a split plot design with 3 factors...but not a split-split plot.
Here's the set-up.
There are 6 Replicates. Each replicate is divided into two whole plots.
(The whole plots represent the first factor, Yes/No)
Within each whole plot we have a 2x4 factorial design...these two factors
being the ones of main interest for testing purposes.
This is different from both the straight split-plot and the split-split plot
examples we have. Can anyone help with the model and sas commands for this?
We may need to correctly identify expected mean squares for the discussion
as well.
Thanks for any help.
Dwight Galster
Statistics Consulting
North Dakota State University
| 
