Wilcoxon paired

[Zainab Mahmood]

How to perform wilcoxon paired as a post hoc after Friedman test in long data type in stata

[Joseph Coveney]

I don't know about the "as a post hoc" bit, but signrank is described in its help file as "Wilcoxon matched-pairs signed-rank test".

help signrank

[Zainab Mahmood]

Thank you that's really helpful.. do you know the syntax in long data type?

[Joseph Coveney]

do you know the syntax in long data type?

It doesn't have one. You'll need to reshape first.

[Zainab Mahmood]

Thanks for confirming. that's what I ve concluded finally. Now in the middle of reshape commands. I used this one reshape wide varname, i(PatientID) j(visit) but it did not come out correct. My data is for two different groups ( Treated vs Placebo) where outcomes (many outcomes/ columns measured over three timepoint over a year). I need the group (Treated vs Placebo) to remain as long. but all other columns/ outcomes to be wide visit1, visit2, visit 3 instead of having a variable named visit and all in long format. Thanks in advance