Re: Binomial dta: how to handle don't-cares?

Just my two cents:

The data are meaningless because of the poor response rate, unless you
can prove that the nonrespondents have the same yes/no/indifferent
proportion as the respondents. I doubt that you can do this.

IF you choose to go ahead anyway, it is clear the the indifferent
individuals are a non factor. You can do all sorts of trinomial
analysis, but the no's clearly win out.

Jonathan

.

Loading