Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Adam Olszewski <adam.olszewski@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: sts test on pweight-ed data |
Date | Tue, 31 Jul 2012 00:06:08 -0400 |
Dear list, I am performing a survival analysis on a pweighted dataset. I am trying to test a variable which does not hold a PH assumption, so the Cox PH model (without time-varying covariate at least) does not hold. On the other hand, standard logrank test is also not feasible. Stata uses a "Cox" variation of the test. My question is - can such a test be relied on in a setting of PH assumption violation? My experience is that the result is basically the same P-value as from a "flat" univariate -stcox- command. I understand there are other options (regression with tvc), but I would like a simple non-parametric test to compare the survival function without worrying about the model assumption. Best, Adam Olszewski * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/