- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Posted 04-17-2020 12:24 AM
(1032 views)
Hello,
I have a non-normal data and I think I should use Friedman non-parametric test.
Question is to determine by var1 the effect of var2 & var3 on var4.
I tried the Friedman test using the following codes but I am not sure I am doing the correct thing as regards the question.
/*using proc rank &glm*/
proc rank
data = dataname;
by var2;
var var4;
proc glm
data = dataname;
class var2 var3;
model var4= var2 var3;
by var1;
run;
/*using proq freq*/
proc freq
data=dataname;
tables var2*var3*var4 / cmh2 scores=rank noprint;
by var1;
run;