Thanks Roger.

> What *is* possible is to fit a surrogate Poisson mixed-effects
> regression and to re-interpret the fitted model as a multinomial model.

I will perhaps try this route, although I might also start by fitting 
an old-style multi-nomial regression instead...

I'm also thinking of the possibility of a series of binomial 
(logistic) regressions...

> Also, it is supposedly possible to fit Bayesian multinomial
> mixed-effects regression using WinBUGS (but this is only for Windows :().

Ah, this is also bad news -- I thought there was an almost equivalent 
version of BUGS for unix...

> Hope this helps.  There is clearly a need for this type of model in the
> analysis of linguistic data...

Yep.

Thanks again.


Regards,


Marco









-- 
Marco Baroni
CIMeC, University of Trento
http://www.form.unitn.it/~baroni

_______________________________________________
R-lang mailing list
[email protected]
http://pidgin.ucsd.edu/mailman/listinfo/r-lang

Reply via email to