Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: std errors in sqreg
From
Austin Nichols <[email protected]>
To
[email protected]
Subject
Re: st: std errors in sqreg
Date
Thu, 24 Mar 2011 13:18:55 -0400
Guadalupe, Maria <[email protected]>:
Are you resampling clusters (countries, I guess)?
On Thu, Mar 24, 2011 at 12:04 PM, Guadalupe, Maria <[email protected]> wrote:
> Dear Statalist,
> I am having trouble with the sqreg command and do not quite understand what is going on.
>
> I have a dataset with 930 firm-level observations and I am running sqreg at 5 percentiles and with 55 regressors (mostly year/industry/country dummies). My coefficient of interest only varies at the country level. Now, all is good when I have "few" (up to 500 is ok) repetitions of the bootstrap: sqreg and bsqreg give very similar results, up to sampling variation. But when I increase the number of repetitions to 1000, then sqreg gives "crazy" standard errors for some of the quantiles (column of p-values of 1). The log says that convergence was not achieved in many of the 1000 iterations, but that is true for both commands and for fewer iterations. And the problem is not there with fewer regressors.
>
> The old stata manuals say that sqreg could no run more than 336/#quantiles, and I am getting close to that, but I am running Stata 11 and the online manuals do not indicate such constraint. Is maybe the constraint still there, but not posted on the new manuals?
*
* 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/