Hello!
I am currently doing a regression with panel data and I am quite new to using Stata. I am using the xtreg command with fixed effect and robust after recomendations from my supervisor. I have a couple of questions that I would really appreciate if some of you could answer.
First, I have a dataset with N= 350 and t=4. I understand there can occur a problem with this called "the incidental parameter problem" or the " large N small t"- problem? What does this mean? What are the difficulties with having a large N but small t? How can you justify using this kind of data set event though the problem exists?
Second, I am wondering about the fixed effect. Is it true that you cannot draw inference outside the sample with this kind of model? If so why would you use this model anyway?
I am also wondering about the robust command. I understand that it mitigates violation of the OLS assumptions somehow, but does it mitigate all violations? I am essentially trying to figure out if I have to do some pre/post testing of the assumptions.
I would really appreciate answers to these questions!
All the best
Andreas
I am currently doing a regression with panel data and I am quite new to using Stata. I am using the xtreg command with fixed effect and robust after recomendations from my supervisor. I have a couple of questions that I would really appreciate if some of you could answer.
First, I have a dataset with N= 350 and t=4. I understand there can occur a problem with this called "the incidental parameter problem" or the " large N small t"- problem? What does this mean? What are the difficulties with having a large N but small t? How can you justify using this kind of data set event though the problem exists?
Second, I am wondering about the fixed effect. Is it true that you cannot draw inference outside the sample with this kind of model? If so why would you use this model anyway?
I am also wondering about the robust command. I understand that it mitigates violation of the OLS assumptions somehow, but does it mitigate all violations? I am essentially trying to figure out if I have to do some pre/post testing of the assumptions.
I would really appreciate answers to these questions!
All the best
Andreas
