Dear Statalisters,
I want to test fixed-effects VS random-effects.
I am using the xtoverid after the random-effects model (I cannot use the Hausman test); my problem is that I get the following message:
"variable _cons not found"
I cannot exclude the constant (like I have done for a few other dummy variables that otherwise would have been omitted) because there is no "noconstant" option.
What can I do?
Thank you in advance
I want to test fixed-effects VS random-effects.
I am using the xtoverid after the random-effects model (I cannot use the Hausman test); my problem is that I get the following message:
"variable _cons not found"
I cannot exclude the constant (like I have done for a few other dummy variables that otherwise would have been omitted) because there is no "noconstant" option.
What can I do?
Thank you in advance
Comment