Testing the Presence of Fixed Period Effects
Posted: Fri Dec 03, 2010 1:13 pm
I want to test my panel model for fixed period effects. I want to apply F statistics but I have a few question,
Initially I have some missing observations, so it's an unbalanced panel, does it matter for this test? (I'm not good at panel data, it's an urgent homework)
And I have read about this test in William Greene's book, in the example the formulation is not clear, it just shows the F statistic like "F(14,72)", and i couldn't apply it to my model, i didn't get how to calculate the degree of freedom.
My model is,
299 observations excluding missings
n=27
t=12
with 10 independent variables.
Initially I have some missing observations, so it's an unbalanced panel, does it matter for this test? (I'm not good at panel data, it's an urgent homework)
And I have read about this test in William Greene's book, in the example the formulation is not clear, it just shows the F statistic like "F(14,72)", and i couldn't apply it to my model, i didn't get how to calculate the degree of freedom.
My model is,
299 observations excluding missings
n=27
t=12
with 10 independent variables.