|
|  |
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: NBREG for ordinal scales
At 12:43 PM 10/10/2006, [email protected] wrote:
My feeling is: There's no particular reason we can only use Poisson or
NBin on count data. Surely the important thing is that the distribution
matches, right? In Poisson or NBin regression, we express results in terms
of Incidence Rate Ratio, which I guess only makes sense if you're thinking
of events happening. But what about calling it 'mean ratios', as
effectively they are just that?
I have no backing from any reference or anything, but just thinking
logically (I feel), that is what I would conclude. Richard, you don't
agree with using count-type regression techniques on non-count data. Why
is that?
For the ordinal variable in question, why would you expect the
distribution to be right? 0 = never, 1 = 1 or 2, 2 = 3 or 4, 3 = 5
or more. In effect, the original count was collapsed, and in a
somewhat arbitrary way; it could have just as easily been 1 = 1, or 1
= 1 or 2 or 3, or whatever. It may well be that the method doesn't
work too badly (just like OLS regression may not give that horrible
of results with a dichotomous outcome) but I'd be surprised if this
was the optimal way to do things. Now, it may be that somebody
somewhere spent a great deal of time validating this approach; but it
may also be that somebody just did it, the reviewers didn't complain,
and now it is semi-accepted practice.
-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
FAX: (574)288-4373
HOME: (574)289-5227
EMAIL: [email protected]
WWW (personal): http://www.nd.edu/~rwilliam
WWW (department): http://www.nd.edu/~soc
*
* 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/
