On Tue, 4 Sep 2007, Mariusz Doszy? wrote:
Do you know formula for Wald test statistic (of joint
significance of 0-1 time variables) in case of models for panel
data with fixed effects?
Well, it's just the standard formula. In gretl you can do
omit dt_*
for the F-form of the test, based on the restricted and
unrestricted sums of squared residuals
or
omit dt_* --wald
for the chi-square form, based on the covariance matrix for the
unrestriced estimates.
--
Allin Cottrell
Department of Economics
Wake Forest University, NC