Programming the statistical procedures from SAS

Split plot desig

New Contributor
Posts: 2

Split plot desig

I am having problem to write SAS code for Mean comparisons among subplot levels within a common main plot in a split plot design.


This is what I have so far, But I think it is wrong


proc sort data=test
by ir;
proc glm data=test;
class rep ir nrate;
model ndvi ndre ci = rep ir rep*ir nrate ir*nrate;
test h=ir e=rep*ir;
lsmeans nrate/lsd;
by ir;

Respected Advisor
Posts: 2,055

Re: Split plot desig

Give us a hint.


Why do you think it is wrong?

Paige Miller
New Contributor
Posts: 2

Re: Split plot desig

Posted in reply to PaigeMiller

The mean comparison has to be separate by factor A, in this case, irrigation.

And I also discover that the data is not normallly distributed so I will have to use non-parametric anlysis... I have no idea how to do that

Occasional Contributor
Posts: 18

Re: Split plot desig

Since you are not farmiliar with the Non-parametric test transform your data so that you can continue with the Proc glm? The proc glm will only run data that their residuals are normally distributed. Or it will but the results will be wrong. Another alternative i can suggest is to run your data using the Proc glimmix since it will allow you to define the distribution that fit your data best and link it to the normal dist using any link function that is suitable.


Hope it helps.

Ask a Question
Discussion stats
  • 3 replies
  • 3 in conversation