Re: st: Goodness-of-fit test with conditional logit

[Steven Samuels <sjhsamuels@earthlink.net>]

Peter,  The estimated coefficients from -clogit- cannot generate an  
ordinary prediction because there is no intercept.  The estimated  
probabilities will therefore not resemble the true probabilities. If  
all covariates are at zero, for example, the predicted probability  
will be 0.5.


Steven



On Feb 7, 2008, at 10:22 AM, Peter Dijkstra wrote:


> Sorry for bothering with a similar question, but I'm still looking  
> for a goodness-of-fit test after I ran a conditional logit model. I  
> do use the linktest, but that is a specification test and I want to  
> know what the predictive value is of my model. Any ideas?
> Thanks in advance.
> Cheers, Peter
> --
> Psssst! Schon vom neuen GMX MultiMessenger geh�rt?
> Der kann`s mit allen: http://www.gmx.net/de/go/multimessenger
> *
> *   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/

*
*   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/