Danielle Ferry <[email protected]> asks,
> I am comparing 2 models estimated using -arima-. When I run -lrtest-
> afterwards, I get the following:
>
> lrtest v2 v1
> (log likelihoods of null models cannot be compared)
>
> Likelihood-ratio test LR chi2(5) = 9.11
> (Assumption: v1 nested in v2) Prob > chi2 = 0.1048
>
> What does the first message in parentheses mean? Does this mean that
> the results of the LR test shown are invalid?
-lrtest- is trying to be smart and identify null models, to keep you from
potentially making a mistake. Unfortunately, its rule for identifying null
models is NOT smart. It looks only at the main regression equation, which in
ARIMA models is often null.
So long as your models are nested, you can ignore this message.
We will look into improving this message or improving when it is shown.
-- Vince
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/