Hi Simon
I'm examining votes for competing citizens' initiatives on the same
underlying issue. My theory leads me to expect that YY voting will be
predictable by ideology - so for example, I would expect an ordering
like NN-NY-YY-YN along the scale of underlying ideal points. So this
is the model. I don't have a direct measure of ideal points, but I do
have demographics, party ID etc. that (I hope) will predict those
preferences. Does that help?
Cheers
David
On 08/02/2008, Simon Jackman <[log in to unmask]> wrote:
> What is the problem you are trying to solve here? That is, what is the
> question to which having an ordering of the vote outcomes is an
> answer, or a step towards an answer?
>
> I'm also wondering what you mean by "best fitting" ordering. With
> respect to what kind of criterion and/or model?
>
> Pinning down these issues might help list members with your query.
>
> Simon Jackman
> Dept of Political Science
> Stanford University
> http://jackman.stanford.edu
>
> On Feb 8, 2008, at 10:13 AM, David Hugh-Jones <[log in to unmask]>
> wrote:
>
> > I'm running an ALSOS regression to find the best-fitting ordering for
> > some categorical data of pairs of votes. ALSOS turns my dependent
> > variable categories (NN,NY,YN,YY) into numbers, thus imposing an
> > ordering. Unfortunately, it doesn't provide a measure of statistical
> > robustness. So I have been using an ordinal probit, with the DV
> > ordered by whatever ALSOS suggests. It is then possible to test
> > whether the cutpoints are significantly different - this is an option
> > in SPSS.
> >
> > The problem is that however I order the DV - even if I impose a
> > nonsensical order - the tests for difference of the cutpoints always
> > seems to come out as highly significant. This makes me think I am
> > misunderstanding something.
> >
> > Can list members suggest alternative ways to do this? I was
> > considering bootstrapping, and looking at how often I get the same
> > ordering; but if so, how could I calculate a p-value?
> >
> > Cheers
> > David Hugh-Jones
> >
> > **********************************************************
> > Political Methodology E-Mail List
> > Editors: Melanie Goodrich, <[log in to unmask]>
> > Delia Bailey, <[log in to unmask]>
> > **********************************************************
> > Send messages to [log in to unmask]
> > To join the list, cancel your subscription, or modify
> > your subscription settings visit:
> >
> > http://polmeth.wustl.edu/polmeth.php
> >
> > **********************************************************
> >
>
> **********************************************************
> Political Methodology E-Mail List
> Editors: Melanie Goodrich, <[log in to unmask]>
> Delia Bailey, <[log in to unmask]>
> **********************************************************
> Send messages to [log in to unmask]
> To join the list, cancel your subscription, or modify
> your subscription settings visit:
>
> http://polmeth.wustl.edu/polmeth.php
>
> **********************************************************
>
**********************************************************
Political Methodology E-Mail List
Editors: Melanie Goodrich, <[log in to unmask]>
Delia Bailey, <[log in to unmask]>
**********************************************************
Send messages to [log in to unmask]
To join the list, cancel your subscription, or modify
your subscription settings visit:
http://polmeth.wustl.edu/polmeth.php
**********************************************************
|