Ramsey Reset test for Panel Regression

[Alex Mai]

Hello,

Estat ovtest is the command for Ramsey Reset. In many instructions, people use it after the command of "reg". But if I want to do a panel regression, shall I still use the command combination of reg y x + estat ovtest before doing, for instance, my fixed-effects regression? I know that estat ovtest does not work after xtreg y x, fe.

Thank you very much!

[Carlo Lazzaro]

Alex:
you may have a rough idea of what's going on by applying -estat ovtest- after running a pooled OLS with the same dataset.

[Alex Mai]

Alex:
you may have a rough idea of what's going on by applying -estat ovtest- after running a pooled OLS with the same dataset.

Thank you very much! just another question about the suitable threshold value for the reset result. some people say 0.05, but is this too loose? Or if 0.1 would be better.

[Emad Shehata]

http://econpapers.repec.org/software...de/s458101.htm

RESETXT: Stata Module to Compute Panel Data REgression Specification Error Tests (RESET)

[Carlo Lazzaro]

Alex:
my possible fault, but I've never heard about a threshold p-value for RESET.
Usually, the substantive issue is whether a quadratic or cubic specification fits your data better.

[Philipp Ottenstein]

Gentlemen, this contribution here describes Carlo's first comment with an example: http://econometricstutorial.com/2015...n-stata-fe-re/
Thanks, Carlo! The pooled OLS approach with 'estat ovtest' worked for me.

[Carlo Lazzaro]

Philipp:
welcome to this forum.