There were two statistical questions considered by
Nicholas Hengartner of Yale. For one, his
conclusions erred because he assumed the wrong
form for the ballot. For the second, he showed
that the undervote ratio between optical scanned
ballots and punched cards was for practical
purposes independent of county and that the
undervotes when punched cards were used was
substantially larger than for scanned ballots. The
failure of his analysis for the first conclusion
had nothing to do with the validity of his second
conclusion.
The second conclusion is of course the important
one for the plaintiffs, and I am sure that the
judge understood it; however, the popular report
may not reflect this since the error was
skillfully produced. It was a masterful cross, and
great theater up to the best of Perry Mason.
For those who may at some time be called upon to
testify on statistical matters, it would be
worthwhile to find a way to view Hengartner's
testimony. It was one of the better such that I
have seen -- for instance, his explanation of the
difference between association and causality was
clear and properly suited to his audience, and his
tactics in resistance to a hypothetical
contingency table analysis was artful.
[EMAIL PROTECTED] wrote:
>
> At today's Leon County trial involving a manual recount, the statistical
> expert testified that the undervote in Florida counties that use punch
> cards is significantly greater than in counties that use scan cards.
> Part of the crossexamination ignored this conclusion and focused on side
> comments written by the statistical expert about the 1998 election in
> Palm Beach County.
>
> Evidently, in 1998 the Palm Beach vote for US senator was significantly
> less than the vote for governor. The statistical expert was told that
> the senate candidates were in the first page (column) and the
> gubernatorial candidates on the second page (column) of the ballot.
> Consequently, the expert concluded that this data may provide evidence
> to support the claim that some voters may have been unable to punch the
> chad in the first column, thus indicating some type of problem with the
> vote-a-matic machine.
>
> The Bush lawyer presented a copy of the 1998 Palm Beach ballot showing
> that both sets of candidates were on the first page--the senate at the
> top, the gubernatorial in the middle. I don't see how this could
> possibly weaken the witness' conclusions about the punch cards and scan
> cards. Moreover, the 1998 discrepancies could indicate a machine problem
> between the top and middle of the first column.
>
> Domenico Rosa
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
--
Bob Wheeler --- (Reply to: [EMAIL PROTECTED])
ECHIP, Inc.
-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
