Pierre,
You have three possible outcomes: (1,1), (1,0) and (0,1). This means
that you can fit a multinomial logit model.
Friedrich Huebler
--- Pierre Lafaye de Micheaux <[email protected]> wrote:
> Dear Stata gurus,
>
> Here is the experiment we are conducting:
>
> Laboratory rats are smelling some kind of odour and then we measure
> their neuronal activity. We then observe the presence/absence of
> gamma or beta waves.
> But you should always have some kind of wave (beta or gamma). The
> absence of the two waves means a failure in the experiment.
> So you only have (1,1), (1,0) and (0,1) data.
> And never (0,0) data. (But when the rat is dead, or the electrod
> badly placed, etc ...).
>
> We are interested in modeling the presence/absence of gamma or beta
> waves as a function of odour's characteristics.
>
> We then wanted to use a bivariate probit analysis (through biprobit
> for example) but the fact that there is no (0,0) observations
> (structural zero case) seems to be a problem.
>
> Do you know how we could handle this?
>
> Best regards,
>
> Pierre
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
*
* 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/
