Statalist


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

Re: Repost: finding median when using svy?


From   "Austin Nichols" <austinnichols@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: Repost: finding median when using svy?
Date   Thu, 6 Sep 2007 15:31:29 -0400

Stas is right, and you should read the references he cites.
Note that the problem is not one of estimating the median (that's
easy), but of estimating the standard error. To get an incorrect SE,
you can:

webuse nhanes2
prog mymed
 qreg height [aw=finalw]
end
bs, cluster(psu): mymed

but if you don't care about the SE, you can ignore the clustering
induced by your survey design and just

qreg height [aw=finalw]

On 9/6/07, Stas Kolenikov <skolenik@gmail.com> wrote:
> Hey UCB gal,
>
> if you are doing multiple imputations on your own, you won't get a
> correct result from -svy-. Each imputed sample has to undergo all the
> non-response and post-stratification adjustments that only the data
> provider can do properly. Thus in the end you will have understated
> standard errors.
>
> As for your substantive question -- yes, there are some peculiar
> commands that people using survey data are always interested in, but
> Stata does not have them implemented, such as -corr- and all those
> quantile things. In all, getting a median of the complex survey data
> is not such a trivial task. The basic reason is that quantiles,
> including the median, are not smooth functions of data, and thus the
> standard linearization/Taylor series argument fails. The estimates are
> still asymptotically normal, but you need more refined derivations to
> justify that.
>
> Here's a couple of refs:
> http://www.citeulike.org/user/ctacmo/article/1628262;
> http://www.citeulike.org/user/ctacmo/article/1628265;
> http://www.citeulike.org/user/ctacmo/article/1628268.
>
> On 9/6/07, ucb_gal <ucb_gal@yahoo.com> wrote:
> > Hi, I asked this question a while ago, but never received a response that worked.
> >
> > I'm using a large household survey dataset and would like to find median income and wealth, but can't figure out how to use median and the svy prefix.
> >
> > And it'd be even better if it worked with the multiple imputation "mim" prefix.
*
*   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