Page 1 of 1
Choosing fixed/random effects model with Hausman-test
Posted: Tue Apr 28, 2015 8:49 am
by Zerathios
Hello!
I am trying to run a Correlated random effects - Hausman test on this regression in order to determine if a fixed or random effects model is appropriate.
I get the error message "** WARNING: estimated cross-section random effects variance is zero."
Is there a reason to why this is happening?
I've looked it up on the web but I haven't found a good answer.
Kind Regards, Zerathios
Re: Choosing fixed/random effects model with Hausman-test
Posted: Tue Apr 28, 2015 10:46 am
by EViews Glenn
For your data, the estimate of the random effects variance is zero.
Re: Choosing fixed/random effects model with Hausman-test
Posted: Wed Apr 29, 2015 7:09 am
by Zerathios
For your data, the estimate of the random effects variance is zero.
Does this mean that the random effects model is not a well-suited approach to this data or simply that there is something wrong with my data?
Re: Choosing fixed/random effects model with Hausman-test
Posted: Wed Apr 29, 2015 4:46 pm
by EViews Glenn
The estimates suggest that there are no random effects.
Re: Choosing fixed/random effects model with Hausman-test
Posted: Sat May 02, 2015 6:30 am
by Zerathios
The estimates suggest that there are no random effects.
Thank you for the answer Glenn!
I have another question: I want to run a fixed effects model and a random effects model respectively with robust standard errors, adjusting for any heteroscedasticity present. Is this possible in Eviews 8? I have searched the forums but there does not seem to be any clear answer to this question.
Kind Regards, Zerathios
Re: Choosing fixed/random effects model with Hausman-test
Posted: Sun May 03, 2015 3:34 am
by Carlo Lazzaro
Zearthios may find the following thread useful:
http://forums.eviews.com/viewtopic.php?t=362&f=4.
Kind regards,
Carlo