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

st: Sensitivity of non-parametric tests to N


From   siyama@who.int
To   statalist@hsphsun2.harvard.edu
Subject   st: Sensitivity of non-parametric tests to N
Date   Tue, 10 Feb 2004 16:34:09 +0100

Dear stata-listers,

I have a variable T (which is time to failure) for a binary failure event. I
wish to test whether five groups, assumed to be homogenous, have the same
survivor function. So,  I have tried using:

sts test group, logrank 	<assuming proportional hazards>

sts test group, wilcoxon	<assuming hazards vary in ways other than
proportionally and censoring patterns are similar across groups>

sts test group, peto		<<assuming hazards vary in ways other than
proportionally and censoring patterns are different across groups>


All of these tests results indicate lack of equality  (i.e. p-values <
0.05). I was wondering could that in part due to the sizeable groups  I have
(about 120-200 patients per group), or is that real evidence of inequality
in survivor functions.


Many thanks for your help in advance.

Amani
*
*   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