Bookmark and Share

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: Robust Regression


From   Jorge Eduardo Pérez Pérez <[email protected]>
To   "[email protected]" <[email protected]>
Subject   Re: st: Robust Regression
Date   Thu, 6 Mar 2014 12:29:22 -0500

If you are estimating a random effects model,you could try regression
through the median, which can be implemented with clustered s.e. using
the -qreg2- package from SSC.

But I would not advise using this if you are estimating a fixed
effects model. Quantile regression with fixed effects is not a simple
matter, see

http://dx.doi.org/10.1016/j.jmva.2004.05.006


--------------------------------------------
Jorge Eduardo Pérez Pérez
Graduate Student
Department of Economics
Brown University


On Thu, Mar 6, 2014 at 11:09 AM, Nick Cox <[email protected]> wrote:
> I guess the short answer is No.
>
> Oddly, or not, two distinct cultures are emerging in which "robust" in
> statistical science means variously (a) robust to outliers (b)
> "robust" standard errors, and not many try to do both. Indeed it's not
> especially clear that the two approaches can co-exist easily.
>
> Nick
> [email protected]
>
>
> On 6 March 2014 15:54, Robert Davidson <[email protected]> wrote:
>> Hello Statalist,
>>
>> I am trying to estimate a model in Stata 13 in which the data appears
>> to be heavily influenced by outliers.  I have reviewed graphs,
>> calculated the leverage of the variables, and reviewed the data for
>> accuracy and everything points to the influence of outliers.  I have
>> looked into robust estimators including mmregress (Vervardi and Croux)
>> and rreg to deal with this and both report similar results.  The data
>> is a panel though, with the same firm (my unit of observation)
>> appearing in multiple years and thus the standard errors are not
>> i.i.d.  Does Stata have a robust estimator to deal with outliers that
>> can also allow standard errors to be clustered at the unit (firm)
>> level?
>>
>> Thank you,
>> Robert Davidson
>> *
>> *   For searches and help try:
>> *   http://www.stata.com/help.cgi?search
>> *   http://www.stata.com/support/faqs/resources/statalist-faq/
>> *   http://www.ats.ucla.edu/stat/stata/
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index