Re: st: "cmp" command and rho LR test statistic for SUR bivariate probit model

[Maarten buis <maartenbuis@yahoo.co.uk>]

--- On Fri, 4/3/11, David Quinn wrote:
> I know to do a LR test, you have to compare constrained and
> unconstrained models.  I'm used to doing this with typical
> OLS and logit regressions, though, where the constraints
> are something different. So, for clarification purposes:
> you will always set rho to 0 for the constrained model when
> doing a LR test using heckprob, biprobit, etc.?

No, what I use for the constrained model depends on the 
hypothesis I want to test. This has nothing to do with whether
I estimated a linear regression, logit, heckprob, or whatever
other model one can think of. So the first step is to determine
your null hypothesis, than you figure out what the constraint
is that is implied in that hypothesis, and that will be your
constrained model. I thought that you wanted to test whether
the error terms of both equations are independent, which means
that the correlation between these error terms is 0, so I 
used that for my restricted model. If you had a hypothesis that
stated that the correlation should be .2, then you should set
atanhrho_12 to `= atanh(.2)'. I cannot tell you what your 
null-hypothesis should be, that is your job.

-- Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany

http://www.maartenbuis.nl
--------------------------


      

*
*   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/