Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: non-parametric tests

From   "Nick Cox" <>
To   <>
Subject   st: RE: non-parametric tests
Date   Tue, 28 Jun 2005 13:31:59 +0100

By correct, I guess you mean that e.g. computed P-values
are reliable. The answer is No. The Mann-Whitney 
test depends on mutual independence.  See for example 
Box, Hunter and Hunter: Statistics for experimenters 
Wiley 1978. 



> I have a question for you. I'm comparing the data from two 
> independent samples and I'm running the Mann-Whitney test. 
> However, the observations are independent between the samples 
> but not within  the same sample. Is the Mann-Whitney test 
> still correct? 

*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index