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

Re: st: Zeroes and signrank


From   Richard Goldstein <richgold@ix.netcom.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Zeroes and signrank
Date   Sat, 09 Nov 2002 14:42:25 -0500

While I personally don't have strong feeling about the various
methods, I note that other people do.  Discussions can be found
in each of the following books:

Marden, JI (1995), _Analyzing and Modeling Rank Data_, Chapman & Hall

Gibbons, JD (1985), _Nonparametric Stat'l Inference, 2nd edition_,
Marcel Dekker

Lehmann, EL (1975), _Nonparametrics_, Holden-Day

Good luck,

Rich

Marcello Pagano wrote:
> 
> Dear Listers,
> 
> The signrank command deals with the Wilcoxon signed rank test.
> The question I have has to do with ties, or zeroes.
> One approach would be to reduce the sample size by the ties,
> and proceed.
> Another approach is the one taken by Stata.
> The difference is in the calculation of the variance: Stata's
> approach is more conservative because it yields a larger variance.
> Any other reasons for preferring the one approach to the other?
> 
> m.p.
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index