Date: Nov 12, 2000 5:50 PM
Author: Jerry P. Becker
Subject: RESPONSES: FL Statistical approach
Below are some replies I received from recipients of my last posting
about the website http://madison.hss.cmu.edu . Just for your
(1) Jerry, I've always known that there are lies, damn lies and then
statistics. There is no way to know who specifically double punched
these ballots, nor anyway to re-vote this election. If the voters who
cast these ballots were confused, they should have sought assistance.
It takes resposible people to cast sincere votes, not a herd of cows
told to vote #5 on the ballot. The country should abide by what the
re-counts bear out. I'll abide by it, how about you?
(2) From Japan: One more interesting thing! I heard Bush won
yesterday's afternoon. But the news was wrong. Problem is the 0.5%.
This is an excellent "alive" teaching material, because I think this
is related to significant figure, absolute error and relative error.
(3) Verrrrrrrrry interesting!! Unfortunately 99.999999% of the
country wouldn't understand this. I do tho. It's outrageous isn't
it? (My degree's in Ed. Psych., Statistics, & Measurement.) In my
distant, cloudy past, I have several years experience creating
machine-scored test booklets. The Palm
Coast ballot would never even be considered, much less put out to the
general public without a pilot test.
(4) Actually, this is one of the most positive things about this
whole election. There is plenty of interesting mathematics floating
around. The site below is particularly interesting as it is an
beautiful example of how statistics can be misleading, and I don't
mean this in a pejorative sense as the significant flaw is clearly
(5) I suspect this will be a "textbook example" in years to come.
(6) A thought you may feel free to pass on:
The written Chinese word for the concept "crisis" combines the
characters for "danger" and "opportunity." This election presents
America with a dangerous opportunity. The danger is obvious; we've
seen it all too often in disputed elections overseas. A nation
paralyzed, with no influence in
international affairs, at worst, bloodshed in the streets.
But what of the opportunity? Statistically, the popular vote was as
close to a tie as you'd be likely to see - 200,000 votes out of ten
million - a two percent margin, with only 50% voter turnout. Neither
Gore nore Bush has a clear national mandate. So we as a nation have
an opportunity to demonstrate that we can work around our
differences, agree to disagree, yet work together for the good of the
people. The opportunity is fraught with danger, to be sure -
demagogues and rabble-rousers abound who seek only to provoke a fight
or prove a point - but the America I grew up believing in was capable
of rising to the challenge. Capable of setting aside even serious
differences in political viewpoint for the common good.
Are we still that noble nation? I pray that we are - no matter who
takes the oath of office in January.
(7) Jerry, Thanks for your lovely data!,
Here's something I wrote yesterday and posted on our internal message
board at ____. It goes into a different statistical issue; namely,
the question of whether the errors in the first count could have
fallen so heavily against one candidate (Gore) if errors were
actually distributed randomly. I didn't have precise data on the size
and distribution of "error packets" across Florida. So I simply
assumed that each error was an independent event. Although this
premise needs to be examined when looking at the distribution of
errors, it nonetheless provides a fascinating suggestion that
something fishy (read "non-random") took place in the Florida
(8) There you go again! It is not hard to know how you voted. Why
do you use a Math-List to promote your political views?
(9) The data and Adams' analysis are indeed revealing. Even though,
I would rather academic people stay away from the fray lest some
politicians make an unfair use of academic brooding - exploit it, in
I've just finished watching nightly news. It appears that in Illinois
some 120,000 votes have been disqualified for about the same reasons
as in the Palm Beach county. In Iowa (I may be wrong. Then it's
another state), Gore leads Bush by just 200 votes. It is said that
the beauty of the current system is that irregularities, if present,
tend to cancel out across the nation.
Picking up on a single county may create a wrong impression in a
significant part of the population that big brains in academe somehow
endorse this or that attitude towards the voting results. I think this
would be wrong, because, for one, it's impossible to arrive at a fair
conclusion without getting a more comprehensive study. Other states
and other counties where the race was close must be looked into.
The fact is that the current situation is not academic in its nature.
The ballot was available and was not contested before the vote
started. Adams' research may suggest that the law must be changed.
may only happen and the notion must only be promoted after the
current election is over, not while it is on.
I think the academe may want to closely ponder another question
critically related to what's happening in Florida, the law, and
applicability of technology in education. It's my understanding that
somehow counting by hand is judged by the Gore campaign (and by the
law) as more reliable than counting with the help of computer. May it
happen that the law will eventually have to be adjusted for the lack
of sufficient expertise in counting by hand?
(10) THANKS for the statistical views of the election! Wonder if it
Jerry P. Becker
Dept. of Curriculum & Instruction
Southern Illinois University
Carbondale, IL 62901-4610 USA
Phone: (618) 453-4241 [O]
(618) 457-8903 [H]
Fax: (618) 453-4244