Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: stcox: weighted regression changes the # of observations


From   Michael McCulloch <mm@pinest.org>
To   Statalist <statalist@hsphsun2.harvard.edu>
Subject   st: stcox: weighted regression changes the # of observations
Date   Thu, 30 Nov 2006 17:04:19 -0800

(I'm sorry for the re-send, my last note had poor column alignment.)
I'd appreciate your consideration of this question.
I'm conducting cox regression on a dataset measuring survival after treatment. I'm puzzled as to why weighted regression (using -iweight- after -stcox-) would show a different # of observations in the output. Here are results of the ways I've done the analysis. In each case, I start with only 33 subjects.

WEIGHTED STCOX
*bootstrap, reps(1000): stcox tcm
HR=0.72 (se 0.26), ci95(0.36, 1.48) "No subjects" = 65.62

UN-WEIGHTED STCOX
stcox tcm
HR = 0.77 (se 0.30), ci95 (0.36, 1.66) "No subjects" = 33

parmby "stcox tcm "
HR = 0.77 (se 0.30), ci95(0.36, 1.66) "No subjects" = 33

parmby "bootstrap, reps(1000): stcox tcm
HR = 0.77 (se 0.45), ci95(0.40, 1.49) "No subjects" = 33

bootstrap, reps(1000): stcox tcm
HR = 0.77 (se 0.26), ci95(0.40, 1.50) "No subjects" = 33

*Note: I'm using inverse probability of treatment weighting, with the following command immediately preceding -stcox-:
stset datedied [iweight=wt_unst_II]

Another anomaly is why the bootstrap option yields a narrower ci95.






*
* 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