Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Steve Samuels <sjsamuels@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: subpopulation weights with stcox |
Date | Fri, 27 Dec 2013 10:50:37 -0500 |
Three points: 1. For survey data, standard errors are based on between psu or between- replicate variation. Therefore the only allowed variance method is that defined in the -svyset- statement as an option (before the colon) in the -svy- prefix command. See the Manual entries for -svyset- and -svy-. 2. Your subpopulation option is not specified correctly. You need the subpop() option for -svy- prefix command. See the Manual entry on "subpopulation estimation". 3. As a matter of theory and terminology, there are is no such thing as a "subpopulation weight". You specify the probability weight once for the entire study in a [pweight=] option. There is one time you need to worry about whether the supplied weight is suitable for a subpopulation analysis. That is when the weight was post-stratified, calibrated, or raked, and the weight construction ignored the factors that define the subpopulation. In that case you might want to revert to the strict sampling weights. Steve sjsamuels@gmail.com On Dec 26, 2013, at 12:36 AM, Lisa Marie Yarnell wrote: Hi all, I am having some trouble requesting robust standard errors to account for the survey nature of my data, and in particular, the use of subpopulation weights (males and females). I can run the models without weights, but when I try to apply weights, Stata tells me that my error type for the variance-covariance matrix is incorrect. Yet when I try to request robust standard errors, Stata does not allow it. See below. Does anyone know how to request robust standard errors for subpopulation Cox regressons in Stata? Thank you, Lisa . svy(male): stcox rrisk1_t invalid vcetype r(198); . svy(male): stcox rrisk1_t, vce(robust) option vce() of stcox is not allowed with the svy prefix r(198); . svy: stcox rrisk1_t, vce(robust) option vce() of stcox is not allowed with the svy prefix r(198); -- Lisa M. Yarnell, Ph.D. Postdoctoral Research Associate Department of Psychology University of Southern California 3620 McClintock Ave. Mailroom SGM 501, Office 826B Los Angeles, CA 90089-1061 Mobile, preferred: (617) 548-2893 Office: x0-0850 or (213) 740-0850 Fax: (213) 746-9082 * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/ * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/