The "Turkey" problem (Re: [EM] 2-rank and N-rank Condorcet)

[Forest Simmons]

On Thu, 8 May 2003 josh at narins.net wrote:

> On Thu, May 08, 2003 at 01:07:22PM -0700, Forest Simmons wrote:
> > On Sun, 4 May 2003 josh at narins.net wrote:
> >
> > >
> > > The Social Sciences have repeatedly held that people most prefer 7.
> > >
> > > Agree strongly
> > > Agree
> > > Agree weakly
> > > Neutral
> > > Disagree weakly
> > > Disagree
> > > Disagree strongly
> > >
> > >
> > > I read it again just recently, somewhere.
> >
> > Which election method did the social scientists use to determine that this
> > was the scale preferred by the majority, and what were the other
> > candidates (choices) in the election?
>
> Sorry, I don't remember.
>
> The class was "Research Methods in the Social Sciences"
>
> You _can_ bet the result was statistically significant :-D
>

But statistical significance is always relative to some null and
alternative hypothesis.  What the alternative to seven choices?