Roger Newson has an approach that works for any quantile, accounts for
clustering and weighting, but ignores stratification. See:
http://www.stata.com/statalist/archive/2008-12/msg00925.html
One problem with a bootstrap approach is that the weights for each
replicate must be re-normalized so that they add to the population
total.
Steve
On Mon, Mar 29, 2010 at 5:17 PM, Lachenbruch, Peter
<Peter.Lachenbruch@oregonstate.edu> wrote:
> If you can't find anything else, you can bootstrap the percentiles 1000 times and get approximate confidence intervals.
>
> Tony
>
From: Liliana Olarte
> Sent: Monday, March 29, 2010 1:44 PM
> To: statalist@hsphsun2.harvard.edu
> Subject: st: How do I obtain confidence intervals for percentiles with survey data ?
>
> Since I have survey data, I have to use the command pctile but Stata
> reports the percentile not the conffidence interval, ¿what can I do to
> obtain the confidence interval?
Steven Samuels
sjsamuels@gmail.com
18 Cantine's Island
Saugerties NY 12477
USA
Voice: 845-246-0774
Fax: 206-202-4783
*
* 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/
