I would tweak Maarten's example, slightly.
He had
contract rep78 foreign
I think there is good reason to go
contract rep78 foreign, zero
so that the cells with zeros are explicit.
Nick
[email protected]
Maarten Buis replied to Lord Yupa
> > > An alternative is a loglinear model,
> > > that is, the log cell frequency is a function
> > > of the covariates mapped to dummy variables.
> > > That would be most easily done with -glm-
> > > or -poisson- after a -contract-. That
> > > might be a sledgehammer, but it's a systematic
> > > approach.
>
> --- Lord Yupa wrote:
> > I'm interested in such a problem too; i used to aggregate the data
> > creating 2x2 tables, but this is a workaround (with many
> cons and few
> > pros) and not a systematic approach.
> >
> > Please, could you make an example using glm for such analysis?
>
> example is my middle name (actually it is Leendert)
>
> *----------- begin example -------------
> sysuse auto, clear
> recode rep78 1/2 = 3
> keep if !missing(rep78, foreign)
> contract rep78 foreign
> xi: glm _fre i.rep78*i.foreign, family(poisson)
> *---------- end example -----------
> (For more on how to use examples I sent to the Statalist, see
> http://home.fsw.vu.nl/m.buis/stata/exampleFAQ.html )
*
* 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/
