st: re: logistic function

[Kit Baum <Baum@bc.edu>]

<>
Jon said

Are the mu and s mean and standard deviations of the predictions? I'm  
not clear on this.

Yes.  In the case of the normal the predicted values are themselves z- 
scores, so you can feed them to the normal CDF, but a bit more work is  
needed for the logistic (which e.g. does not have unit variance). If  
you compare predictnl for probit and logit for this model, they do not  
differ by much.


Kit Baum   |   Boston College Economics & DIW Berlin   |   http://ideas.repec.org/e/pba1.html
                              An Introduction to Stata Programming   
|   http://www.stata-press.com/books/isp.html
   An Introduction to Modern Econometrics Using Stata  |   http://www.stata-press.com/books/imeus.html



*
*   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/