[Gretl-users] ADF result

Allin Cottrell cottrell at wfu.edu
Thu Dec 4 15:20:06 EST 2008


On Thu, 4 Dec 2008, Hebert Suarez Cahuana wrote:

> I recomend use Eviews, stata or other program for this, or use
> tables of critical values of MacKinnon, see Wooldridge
> "Introductory Econometrics.

Gretl contains the MacKinnon tables and prints out the appropriate
p-value for the ADF test.  You can also access such p-values
directly using the urcpval() function.

The wrong p-value that was being shown prior to the recent report
and fix was not the actual test p-value, it was the p-value for a
lagged difference of y (one of the "augmented" terms), a statistic
which is visible only if you call for verbose output.

Allin Cottrell


More information about the Gretl-users mailing list