Statalist


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

st: Re: Hausman Test when the RE OLS model includes time-constant variables


From   "Pavlos C. Symeou" <p.symeou@jbs.cam.ac.uk>
To   "Rodrigo Alfaro A." <ralfaro@bcentral.cl>
Subject   st: Re: Hausman Test when the RE OLS model includes time-constant variables
Date   Thu, 17 Apr 2008 19:47:33 +0100

Rodrigo,

Thanks for your reply. I understand that there might be correlation between the random effect and x(i,t) using the RE. However, the RE model on the one hand allows me to include the industry dummies (crucial for the model's specification) and on the other hand most coefficients are statistically significant and support my theoretical arguments. To the contrary, in order to include the industry dummies in the FE model, I need to create interactions and even if I do that, the model hardly explains any relationship as all of my coefficients are statistically insignificant. In addition, the RE reports much higher R-squares (around 0.3) than the FE (around 0.02). Is there a way that I can account for this correlation whilst maintaining my existing (and only) variables in the model?

Thanks,

Pavlos


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