Re: [Gretl-users] MacKinnon critical values
Am 14.08.2012 10:52, schrieb Alessia Via:
Hi Sven,
It is all very clear (p-value too!Thank you) but, when I tabulated
the critical values using "beta0 + beta1/T + beta2/T^2 + beta3/T^3"
there were cases in which there were, in my opinion, conflicting
results (even if the differences were very very small).That is why I
was wondering if there was the need to further verify....
So you're trying to replicate MacKinnon's results, is that it? If you
think you've found differences with respect to gretl's implementation,
then you would need to show us those differences. But be aware that this
is a pretty well-tested area in gretl I think, so maybe you should check
your results first.
-s