I wonder if somebody in the list or from Stata could explain the following
inconsistency between the original version of "binreg" and the version in
Stata 8.0. Briefly, for the same data set and model the original binreg with
the rr option converges, but the new one does not. Convergence with the or
option does not seem to be a problem.
See example below.
(obinreg is the original binreg)
. obinreg obeabdo obesidad edadr if sex==1, rr eform nolog
Residual df = 1066 No. of obs = 1069
Pearson X2 = 903.1589 Deviance = 1171.045
Dispersion = .847241 Dispersion = 1.098541
Bernoulli distribution, log link
----------------------------------------------------------------------------
obeabdo | Risk Ratio Std. Err. z P>|z| [95% Conf. Interval]
-----------+----------------------------------------------------------------
obesidad | 1.199856 .0611339 3.58 0.000 1.085825 1.325863
edadr | 1.031502 .0022952 13.94 0.000 1.027013 1.03601
----------------------------------------------------------------------------
. binreg obeabdo obesidad edadr if sex==1, rr nolog
convergence not achieved
r(430);
Leonelo E. Bautista
University of Wisconsin Medical School
Population Health Sciences
*
* 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/
