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]

st: RE: How is pooled OLS corrected for heteroschedasticity different from the FE model?


From   "Jacobs, David" <[email protected]>
To   "'[email protected]'" <[email protected]>
Subject   st: RE: How is pooled OLS corrected for heteroschedasticity different from the FE model?
Date   Tue, 27 Sep 2011 20:22:58 +0000

The main reason fixed-effects is considered superior is that it automatically controls for effects that don't change even if the researcher cannot enter measures of such effects in her model.

Consider a (perhaps too) simple political example.  Suppose political culture explains a cross-national outcome and it is correlated with some of the explanatory variables.  Suppose also that the researcher is unable to measure political culture.  If the investigator can reasonably assume that political culture does not change at least during the sample period, than such unmeasured effects are automatically held constant in fixed-effects models.  Since omitted variable bias is one of the major problems in non-experimental research, this advantage explains the popularity of this estimation approach. 

Of course, nothing (particularly in statistics) is free.  Fixed-effects models unfortunately are extremely sensitive to measurement error in the explanatory variables.

If you need more information, check out any decent econometric text.  I prefer the one by Wooldridge, which can be found in the Stata bookstore.

Dave Jacobs



-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Haillie Lee
Sent: Tuesday, September 27, 2011 3:52 PM
To: [email protected]
Subject: st: How is pooled OLS corrected for heteroschedasticity different from the FE model?

Dear Statalist,

I am having trouble differentiating pooled OLS corrected for
heteroschedasticity and the FE model. Why are these two different? It
seems to me that both are same in that they try to control for
unobserved  cross-sectional heteroschedasticity. If so, why is the FE
model considered to be superior to the heteroschedasticity-corrected
pooled OLS model?
Thank you very much and I would deeply appreciate any help.

Sincerely with Many Thanks,
Haillie
*
*   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/

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