Page 1 of 1

Unit root testing

Posted: Wed Feb 20, 2013 6:32 am
by Jagielski
Hi everyone,

I am testing for unit roots with a balanced panel data set. I will be using the IPS and Fisher ADF tests.
My question is what form to use: Level or first difference? I am worried about fixed effects, so maybe first difference is a better choice? Eviews7 btw.

Re: Unit root testing

Posted: Sat Feb 23, 2013 7:03 pm
by Quyen Nguyen
If you want to test I(0) is stationary, you choose Level and I(1) is stationay you choose 1st difference .