Re: st: Binomial regression

[Marcello Pagano <pagano@hsph.harvard.edu>]

Convenience should not be the determining factor, rather which model 
best fits the data should be what governs our choices.

m.p.

Tim Wade wrote:

> Thanks Constantine for sharing your results.
>
> There are certainly cases where the risk difference is a more
> appropriate  or desirable measure. In these cases,  the identity link
> does have the significant advantage of being linear on the probability
> scale instead of the log odds scale. And while we can convert log-odds
>  to probability, it is often convenient, especially when we have a
> multivariate model to be able to interpret the regression coefficients
> as the expected change in the probability (i.e, the risk difference)
> holding other factors constant. Since probability is not linear in the
> logistic model,  similar inferences about probability or risk
> differences cannot be made with the logistic model. If one wants to
> make a statement about risk differences from multivariate logistic
> model, it needs to be with regard to holding the other covariates at
> some specific constant value, such as zero or at their mean values.
>
> Tim
>
> On 8/2/07, Marcello Pagano <pagano@hsph.harvard.edu> wrote:
>   
> > Sorry to disagree with your first sentence, Constantine.
> > Logistic regression stipulates a linear relationship of covariates with
> > the log of the odds of an event (not odds ratios).
> >  From this it is straightforward to recover the probability (or risk, if
> > you prefer that label) of the event.
> > Don't understand your aversion to logistic regression to achieve what
> > you want to achieve.
> > If you don't like the shape of the logistic, then any other cdf will
> > provide you with a transformation to obey the constraints inherent in
> > modeling a probability.  The uniform distribution that you wish to use
> > has to be curtailed, as others have pointed out.
> > m.p.
> >
> >
> > Constantine Daskalakis wrote:
> >     
> > > No argument about logistic regression. But that gives you odds ratios.
> > > What if you want risk differences instead?
> > >
> > >       
> > *
> > *   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/
> >
> >     
> *
> *   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/
>   
*
*   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/