I need to test that the distribution of a variable is equal across groups. If the groups were two, I would use the - ksmirnov - command, but I have 5 groups. I know that generalized versions of the Kolgomorov-Smirnov test have been proposed (e.g. http://projecteuclid.org/DPubS?servi...oms/1177706261 and http://epub.wu.ac.at/2960/1/Report105.pdf), and I was wondering if any of these (or some other versions) have been implemented in Stata.
Is there a way to compare the distribution of a variable (not just its mean, which I could do with a Kruskal Wallis test) across n>2 groups in Stata?
Also, my data are survey data. Therefore, something that supports - svy: - or, at least, - pweights - would be preferable.
Thanks
Is there a way to compare the distribution of a variable (not just its mean, which I could do with a Kruskal Wallis test) across n>2 groups in Stata?
Also, my data are survey data. Therefore, something that supports - svy: - or, at least, - pweights - would be preferable.
Thanks
