Page 1 of 1
Panel data
Posted: Mon Apr 16, 2012 7:54 pm
by napster
Hi Guys,
I used unbalance panel data for study, so i can't use Huaseman test for choose fixed or randome effect model. Have you any way to fix this problem?
Thanks
Napster
Re: Panel data
Posted: Sat Jun 30, 2012 5:31 am
by eva
Hello,
I have the same problem. I try to reason which is most logical to use (I have countries as crossections so fixed effects) but I estimate both to see if it would lead to a different conclusion and test for robustness.
Good luck,
Re: Panel data
Posted: Sat Jul 28, 2012 2:11 pm
by cecilcarmo
Are you sure you can´t use the Haussman test in unbalanced panels?
I think you can´t use the Haussman test if the panel has gap´s. So you should introduce NA when some year or individual has not information for some variable.
Please anyone correct me if I'm wrong.
Re: Panel data
Posted: Mon Jul 30, 2012 10:26 am
by EViews Glenn
The Hausman test for fixed vs. random effects should be perfectly fine in unbalanced panels. Gaps are no problems as well.