Re: st: Testing the extent of difference between two coefficients in the same model

[David Hoaglin <dchoaglin@gmail.com>]

I don't know the theoretical basis for expecting the two coefficients
to be equal, but if testing equality is the specific objective, it
would be straightforward to use a likelihood-ratio test.  A model in
which the two coefficients are equal is the same as the model in which
the predictor associated with that coefficient is the sum of "old" and
"new."  The two likelihoods in the likelihood-ratio test are the
likelihood for that restricted model and the likelihood for the
unrestricted model.

David Hoaglin


> The test before modification tests the equality of the two
> coefficients. usually you don't move further unless it rejects that.
> So you can decide if the equality in the second (modified) test is
> relevant to your conclusion.
*
*   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/