Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: Hetero and AR1


From   "Schaffer, Mark E" <[email protected]>
To   <[email protected]>
Subject   st: RE: Hetero and AR1
Date   Tue, 4 Apr 2006 19:00:26 +0100

Matthieu, 

> -----Original Message-----
> From: [email protected] 
> [mailto:[email protected]] On Behalf Of 
> Matthieu Boussichas
> Sent: 04 April 2006 18:17
> To: [email protected]
> Subject: st: Hetero and AR1
> 
> Hi everybody,
> 
> i would like to make a estimation with panel data 
> (unbalanced), with heteroskedasticity and autocorelation.
> Should i use xtgls, xtregar, prais, prais with corc, or other 
> ... ?

You can add xtreg with cluster to the list (to make it robust to
arbitrary autocorrelation), and xtivreg2 with bw (to use kernel-based
robust SEs, a la "Newey-West").

> What is the best ?

Here you have to decide.  If there was a "best", we'd all use it and
never use anything else!  One obvious trade-off is efficiency vs.
robustness: the more structure you add (e.g., form of autocorrelation),
the more efficient (precise) your estimates, but the less robust they
are to violations of your assumptions.  But there are many other
considerations ... hence the wide range of choice you have.

Cheers,
Mark

> Thanks
> 
> boussichas matthieu
> 
> *
> *   For searches and help try:
> *   http://www.stata.com/support/faqs/res/findit.html
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
> 
> 

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index