roger webb escreveu:
I've found that the problems encountered in trying to fit log-binomial
models (using either 'binreg' or 'glm link(log) family(binomial)') can
be overcome by using the 'modified' Poisson regression approach
proposed by Zou (Am J Epidemiol 2004; 159: 702-706). This entails
fitting a Poisson model with a robust variance estimator, i.e.:
poisson y x1 x2..., robust irr
This produces the same Risk Ratio (and SE) as the binomial
model, without failing to converge with covariate adjustment.
Roger Webb
University of Manchester (UK)
Dear Roger:
Thank you. I will try, using the data set I mentioned in the message to Ronan.
Jose Maria
Jose Maria Pacheco de Souza, Professor Titular
Departamento de Epidemiologia
Faculdade de Saude Publica/Universidade de Sao Paulo
Av. Dr. Arnaldo, 715 cep 01246-904
Sao Paulo Brasil
fones (11)3082-3886 (11)3714-2403 (11)3768-8612
fax (11)3082-2920 (11)3714-2403
www.fsp.usp.br/~jmpsouza
[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/
