On Saturday, Amrita asked for advice on obtaining marginal effects after
-logit-. She is concerned that she may experience some of the problems
she has read about on Statalist, particularly after the recent mails
concerning -xtlogit-.
Amrita doesn't have anything to worry about when using -mfx- after
-logit-.
First of all, she should find that -mfx- runs relatively quickly after
-logit-. (Running -mfx- after -mlogit- can take a long time because
-mlogit- is a multiple equation estimator and the number of coeficients
can be very large.)
Second, no matter what prediction option she chooses after -logit-
(probably p), Amrita will not encounter any of the problems one can run
into after -xtlogit-. That is because -xtlogit- has some prediction
options that predict values that depend on the grouping variable, and
such functions are not suitable for obtaining marginal effects.
--May
[email protected]
*
* 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/
