Statalist


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

Re:Re: st: help for xtlsdvc / Bias corrected FE


From   olfa alouini <olfasgb@hotmail.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   Re:Re: st: help for xtlsdvc / Bias corrected FE
Date   Sun, 3 May 2009 15:17:56 +0000

many thanks for your answer, this works!
 
- I used bias(3) and vcov (1000) as in the example but I actually don't know how to set the bias and vcov parameters properly considering that:
I have a panel with N=15; T= 10 or T=40 (considering whether i'm working on decades or on the whole data set ) 
I have to address heteroskedasticity and autocorrelation (and assuming those elements come into play when choosing bias and vcov settings; I'm sometimes a little lost on the theory in the paper)
 
 
-when I include the lagged dependent variable in my regression (and perform the bias correction, but even without), almost all the other regressors loose significance compared to the regressions ran without the lagged dependent variable (using -xtreg, fe cluster or xtregar,fe).
How should I interpret this , could this mean I should opt for another estimation method?
 
 
Thanks in advance,
 
Olfa
_________________________________________________________________
Hotmail® has a new way to see what's up with your friends.
http://windowslive.com/Tutorial/Hotmail/WhatsNew?ocid=TXT_TAGLM_WL_HM_Tutorial_WhatsNew1_052009
*
*   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/



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