Thanks to Kit Baum, a new module, -firthlogit- has been uploaded onto
SSC. -firthlogit- implements a penalized likelihood method for fitting
logistic models. The penalization was originally suggested by David Firth
to reduce bias in generalized linear models. Since then, others, notably
Georg Heinze (medical sciences) and Christopher Zorn (social sciences), have
advocated using the method to counter problems raised by "separation" in
logistic regression.
You can find and download the module via -ssc describe firthlogit-. The
module has a few ancillary files that illustrate its use with worked
examples from the literature.
Bug reports, etc., to
Joseph Coveney
P.S. In the ado-file header, the version is given as 0.99. I might wish to
tweak it based upon any feedback I get in the next two weeks.
*
* 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/
