hausman test

[vbree]

I've done the hausman test on eviews. But i don't understand the result. How can i now what method that i should use from the output?

i also have this message from the output:
* Cross-section test variance is invalid. Hausman statistic set to zero.
** WARNING: estimated cross-section random effects variance is zero.
But i don't now what is the meaning.

my Chi-Sq. Statistic is 0.00000
my Chi-Sq. d.f. is 3
the prob is 1.0000

Please explain me about that.

[EViews Glenn]

It means that the random effects estimate of the cross-section variance term is zero, so that there is no evidence of individual effects in your data.

[obicna89]

It means that the random effects estimate of the cross-section variance term is zero, so that there is no evidence of individual effects in your data.

does this mean one should use random effects?

[EViews Glenn]

It means that the "best" estimate of the random effect variance is zero, so that there do not appear to be random effects.

[obicna89]

if there aren't any (or appear to be) individual effects, it means one shouldn't use random effects or FE cross-section?
should one use FE time fixed?

edit:
oh and i have a question.. eveiws gave me the same result: hausman test statistic set to zero. but when i compute the test stat manualy, i got 148 as a test statistic. ( i have computed it from the estimated coef and covariance matrices from fe and re models), so what to decide? fe, re? :(

[BMM]

We have some dummy variables for time in our RE, and we get this answer from the hausman test:
"Cross-section test variance is invalid. Hausman statistic set to zero."
Can someone give an intution about this. If we remove the dummy for time in our model we get a test statistic.

[EViews Glenn]

http://forums.eviews.com/viewtopic.php? ... test#p9055