Kruskal-Wallis (or similar) in a complex sample

[Lara Guedes]

How can I do a KW test (or similar) in a survey sample?

[Carlo Lazzaro]

Lara:
why not considering:

svy: regress

[Lara Guedes]

My data is non linear, so I should to quantile reg, right? But is it possible in svy?

[Rich Goldstein]

a list of the estimation commands compatible with svy can be found at:

help svy_estimation

without more information (please read and follow the advice in the FAQ), no one can tell you whether quantile regression or something else is "best" in your situation