Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

st: RE: clustered bootstrapped standard errors for quantile regressions


From   "Feiveson, Alan H. (JSC-SK311)" <alan.h.feiveson@nasa.gov>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   st: RE: clustered bootstrapped standard errors for quantile regressions
Date   Fri, 19 Mar 2010 08:23:01 -0500

Ulrich - I have asked this question before. Apparently one must write one's own bootstrap program with clustering to do it. (see below)

al Feiveson


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Jeff Pitblado, StataCorp LP
Sent: Wednesday, September 03, 2008 1:36 PM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: bootstrap fails with qreg

Alan Feiveson <alan.h.feiveson@nasa.gov> is trying to use the -bootstrap- prefix command with -qreg-:

> Hi - I tried to get bootstrapped standard errors with median 
> regression by using the -bs- command. The regression appears to run 
> through all the iterations, but at the end ,there seems to be a 
> problem in collecting all the estimates to get the bootstrapped 
> standard error. I notice that Bill Gould has provided code in STB-9
> (sg11_1.ado) to do this, but this is from some years ago and the
> -qreg- version that it uses might not be the one we now have with 
> Stata 10.1. So my questions are
> 1) Is sg11-1.ado still the best way to do this ,other than 
> bootstrapping by hand with -bsample-?
> 2) Is there a workaround using the current Stata 10.1 cabilities? 
> 3) Can this be done with cluster sampling? 
> 4) I'm curious - what is the reason for the failure (see below)?
> 
> 
> AL Feiveson
> 
> . bs: qreg z4 zrs0
> (running qreg on estimation sample)
> 
> Bootstrap replications (50)
> ----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5
> ..................................................    50
> 
> Median regression                                    Number of obs =
> 1860
>   Raw sum of deviations 1217.215 (about -.99425226)
>   Min sum of deviations 1217.215                     Pseudo R2     =
> 0.0000
> 
> last estimates not found
> r(301);

The -bsqreg- estimation command fits quantile regression with bootstrap standard errors; however, it does not support cluster sampling.

The -qreg- estimation command does not support the -bootstrap- prefix.  This is mostly due to programming details and allocation of development resources.


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