st: AIC test level or first differences?

[Merijn Groenenboom <m.groenenboom@students.uu.nl>]

Dear statalisters,

I am writing a research with tests like multivariate cointegration
test and granger causality test.
To determine the number of lags, I use Akaike Information Criteria (AIC) test.
I use first differences for the granger causality, so I expect I
should use first differences for AIC test as well?
And similarly I do not use first differences for the multivariate
cointegration test, so I expect not to use first differences
for the AIC test here?

Can anybody tell me whether I am right or not?

Thanks in advance!
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/