On Tue, 5 Dec 2000 the anonymous correspondent
"[EMAIL PROTECTED]" wrote:
> Where can I find authoritative data on the undervotes observed in the
> various counties in Florida for the Presidential Election?
>
> My students have suggested that a t-test should be applied determine
> the level of confidence in the assertion that the punch card systems
> produced five times as many undervotes compared to the optical scanning
> system.
And you didn't point out to them that a t-test (or any other statistical
test, for that matter) is not applicable to a census?
If you agree (implicitly) with your students' (implicit) assertion that
the Florida votes represent a random sample drawn from a population in
which one might be interested: (1) What is the population to which you
think the sample refers? (2) On what grounds can the present sample (if
indeed it IS a sample) be described as "random"?
> I already have the machine systems by county, but need the
> undervote by county (to link the two). They have already determined
> that simple comparison of the bulk averages is insufficient to support
> the argument.
>
> This would be a nice demonstration for my Intro class.
A nice opportunity to point out what statistics is, and isn't, about,
perhaps; it is not at all clear what it might be a demonstration OF,
apart from egregious misinterpretation of statistical theory and
practice.
-- DFB.
----------------------------------------------------------------------
Donald F. Burrill [EMAIL PROTECTED]
348 Hyde Hall, Plymouth State College, [EMAIL PROTECTED]
MSC #29, Plymouth, NH 03264 (603) 535-2597
Department of Mathematics, Boston University [EMAIL PROTECTED]
111 Cummington Street, room 261, Boston, MA 02215 (617) 353-5288
184 Nashua Road, Bedford, NH 03110 (603) 471-7128
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================