Panel Unit Root Test (Balanced or Unbalanced)
Posted: Mon Jul 20, 2015 12:29 am
Hi,
I am trying to test panel unit root for my unbalanced panel with time gaps using Eviews. I got confused as STATA user manual said only Fisher-type test possible, whereas, Eviews returns results for all the test:
1. Levin-Lin-Chu : According to STATA, needs strong balanced panel
2. Breitung test: According to STATA, panels be strongly balanced
3. Im-Pesaran-Shin test: According to STATA, does not require strongly balanced data, but there can be no gaps in each individual time series.
4. Fisher-type tests: Possible with unbalanced and time gaps
My question is if STATA manual is true, why Eviews returns output?
Your response will be highly appreciated.
Regards,
Rashel
I am trying to test panel unit root for my unbalanced panel with time gaps using Eviews. I got confused as STATA user manual said only Fisher-type test possible, whereas, Eviews returns results for all the test:
1. Levin-Lin-Chu : According to STATA, needs strong balanced panel
2. Breitung test: According to STATA, panels be strongly balanced
3. Im-Pesaran-Shin test: According to STATA, does not require strongly balanced data, but there can be no gaps in each individual time series.
4. Fisher-type tests: Possible with unbalanced and time gaps
My question is if STATA manual is true, why Eviews returns output?
Your response will be highly appreciated.
Regards,
Rashel