|From||Jose Maria <firstname.lastname@example.org>|
|Subject||Re: st: -binreg- (was: ORs for non-rare outcomes)|
|Date||Wed, 14 Apr 2004 21:52:09 -0300|
Ronán Conroy escreveu:
Dear Ronan:on 09/04/2004 02:18, email@example.com at firstname.lastname@example.org wrote:the -binreg- command would be really a good solution, if itI agree, but don't be afraid of mixing -binreg- regressions with -logistic-.
would run more often than not run.
I reviewed a good paper a while back that used -binreg- for about two thirds
of the models and, where it couldn't find a plausible solution, -logistic-
for the rest.
Ronan M Conroy (email@example.com)