|
|  |
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: NBREG for ordinal scales
At 05:18 PM 10/11/2006, Matthew C. Johnson wrote:
That is one of my primary problems. I see a number of articles using count
models for outcome variables such as those presented earlier, but no
justification was provided for doing so other than the fact that the DV
has a Poisson or negative binomial distribution. I assume that they should
provide justification for using these techniques, but does the
justification exist?
If, in fact, the DV has a Poisson or negative binomial distribution,
then that is your justification for using count models. But how do
you know that it does have such a distribution??? And why would you
even think that it does, given the way the questions are worded? Is
there some "classic" piece these authors cite, or do they just do it
without citation? There must have been somebody who started this -
if you could find out who it was that might help solve the puzzle.
Just to further illustrate the oddity - suppose you added two of
these items together. You could get a score of two because (a) you
committed each act 1 time and hence were coded 1 on both items, or
(b) you committed one act 4 times and the other act zero times, and
hence got a score of 2 for the first act. And even more oddly, you
could get a score of 3 because you committed one act 50 times and the
others not at all (and hence got coded 3 = 5 and above on the one
item) and you could also get a three because you committed one act
once and another 3 times. It would also be possible for somebody to
get a higher score than somebody who has committed more acts overall,
e.g. somebody could have two scores of 2 for 4 total because of a
total of 6 actions.
Again, somebody may have shown that such oddities do not occur; but
it also wouldn't surprise me if somebody just did it and got away
with it and now everybody does it.
-------------------------------------------
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/
