> Part of what confuses me is that GQ and -hettest- both seem to test, or
> can test, similar hypotheses.
So what's so unusual about it? In many cases there are more than one way
to tackle the testing problem. If you are testing the location parameter,
there are tests for mean and median:
sysuse auto
ttest price, by(foreign)
ranksum price, by(foreign)
with generalization possible, I believe, to the regression context. You
can also test for the mean in a different manner than the standard z-test.
Say, with a single observation -- rejecting the null when the first
observation is beyond (the mean +/- 2 sigma) range. This will give you the
right level of 5%, but not much power compared to the test based on the
mean, since the rejection region is smaller.
Anyway, if all tests give the same answer, then that sounds good. But in
general, as Kit Baum noted, GQ is of less value than others as those tests
are more flexible and more general. Which to me sounds like it has a
greater power, at least in a rather wide range of situations.
--- Stas Kolenikov
-- Ph.D. student in Statistics at UNC-Chapel Hill
- http://www.komkon.org/~tacik/ -- [email protected]
* This e-mail and all attachments to it are not intended to provide any
* reasonable point of view and was transmitted to you in error. It
* should be immediately deleted by all recipients unless they really
* enjoy communicating with the author :). Other restrictions apply.
*
* 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/
