Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

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

st: RE: st: xtoverid and reporting robust variance-c​ovariance matrix after xthtaylor

From   "Schaffer, Mark E" <[email protected]>
To   <[email protected]>
Subject   st: RE: st: xtoverid and reporting robust variance-c​ovariance matrix after xthtaylor
Date   Thu, 11 Aug 2011 10:59:44 +0100


If all you need is the robust SEs in order to do your robust Wald tests, you can get these displayed by -xtoverid- by using the (undocumented) -noisily- option.  The output of the internal reestimation uses temporary variable names, but you should be able to figure out which are which since the coefficient estimates will be the same as those reported by -xthtaylor-.

Austin Nichols has a modified version of -xtoverid- called -xtoverid2- that saves the robust VCV in r() results.  You can use this to do Wald tests by hand.  Below is an example that installs it and displays the un-robust and robust VCVs.


net install xtoverid2, from(
webuse abdata, clear
xtivreg ys (k =  emp  wage  cap), re
ereturn list e(V)
xtoverid2, noisily cl(id)
return list r(V)

> -----Original Message-----
> From: [email protected] 
> [mailto:[email protected]] On Behalf Of Kwansoo Kim
> Sent: 11 August 2011 04:43
> To: [email protected]
> Subject: st: st: xtoverid and reporting robust variance-c​
> ovariance matrix after xthtaylor
> Hi all,
> "xtoverid" command has been very helpful in generating robust 
> standard errors after xthtaylor in the presence of 
> heteroscedasticity...
> However, it is not clear how to produce robust 
> variance-covariance matrix (the one used to generate robust 
> standard errors in xtoverid).
> I would need this matrix for Wald test, etc. If you could 
> help me on this, it would be great.
> Thanks a lot.
> Kwansoo
> *
> *   For searches and help try:
> *
> *
> *

Heriot-Watt University is a Scottish charity
registered under charity number SC000278.

*   For searches and help try:

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