Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: AW: logistic regressions including the bootstraps command and weights


From   Ben Jann <ben.jann@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: AW: logistic regressions including the bootstraps command and weights
Date   Wed, 16 Sep 2009 16:44:32 +0200

Although there seems to be no reason for such a restriction. This came
up before. Form the relevant literature I infer that the bootstrap has
problems if there are few PSUs per Stratum, but sampling weights
should not be a problem. Simply doing a plain bootstrap and apply
weights within each bootstrap sample should yield valid results. I
also ran some simulations on a handful of problems and bootstrap with
weights always performed well. If anyone has a different opinion on
this or has references to work evaluating this, I'd be interested.
ben

On Wed, Sep 16, 2009 at 4:20 PM, Martin Weiss <martin.weiss1@gmx.de> wrote:
>
> <>
>
>
>
> The [R] manual, p. 209, is very clear on the question: "bootstrap is not
> meant to be used with weighted calculations."...
>
>
>
> HTH
> Martin
>
>
> -----Ursprüngliche Nachricht-----
> Von: owner-statalist@hsphsun2.harvard.edu
> [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Sergi VIDAL
> TORRE
> Gesendet: Mittwoch, 16. September 2009 16:18
> An: statalist@hsphsun2.harvard.edu
> Cc: kmaul@empas.uni-bremen.de
> Betreff: st: logistic regressions including the bootstraps command and
> weights
>
> Hi can anybody help me with this:
>
>
> I do logistic regressions including the bootstraps command (example:
> logistic dependent independent if alter < 39 , vce(boot, reps (100)) and now
> I want to apply a weight. Is there a way I can do weighting and
> bootstrapping at the same time?
>
> Thanks,
> S
>

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index