Tell me more ×
Cross Validated is a question and answer site for statisticians, data analysts, data miners and data visualization experts. It's 100% free, no registration required.

I have a question about using panel unit root test and panel co integration test. I used IPS unit root test then I realized my variables are I(1) can I now use kao co-integration test for checking long run relationship? I think IPS is for heterogeneous panels & kao is for homogeneous panels so we can not use both of them? Is my vision true? if anybody can help me I really appreciate it.

share|improve this question
Which reference are you using, Baltagi? – mpiktas Sep 19 '12 at 13:56
so u think it depends on reference! I don't know I saw it in several paper! so what do u suggest? could u introduce me a reference which can confirm using that 2 kind of tests? If it is possible please tell me. I need answer. I am waiting for your reply. – gibran Sep 20 '12 at 6:56
1  
No, I do not think it depends on reference, the reference would give me an idea to look up which tests you use exactly. Panel co-integration and unit-root testing is a bit murky. A lot depends on what kind of problem you are trying to solve. I suggest looking up papers of Pedroni, in my opinion he gives the most coherent picture of how to proceed with unit-root testing and cointegration. – mpiktas Sep 20 '12 at 7:29
well I estimated panel models of 7 period and 20 cross section, one of them with fixed effect and the other one with random effect . – gibran Sep 20 '12 at 17:50
then I realized that R^2 is 0.99 and dw is 0.56 so I thought maybe I have unit root problem. then I try with Im Pesaran & Shin(IPS) test . they were integrated of order1 , I(1) . at the end I checked long run relatioship with Kao cointegration ship . the results show the variables were cointegrated. but the problem is when I saw that kao is for homogeneous panel and IPS is for heterogenous panels?! Is that so? – gibran Sep 20 '12 at 17:55
show 2 more comments

Know someone who can answer? Share a link to this question via email, Google+, Twitter, or Facebook.

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.