Statisticians Recommend New Measures to Ensure Vote Count Accuracy

[papau]

MODS - this is a PR release with no copyright


Statisticians Recommend New Measures to Ensure Vote Count Accuracy
Release "Ohio’s 2004 Exit Poll Analysis for Novices”

http://www.prweb.com/releases/2006/2/prweb346936.htm
http://electionarchive.org/ucvAnalysis/US/exit-polls/Ohio2004-US-future.pdf

Why should Americans care about possible 2004 vote miscounts? The 2004
election is over. It’s old news. The only reason for rehashing prior
elections is to ensure that our votes are counted the way voters intend
in the future. Should Americans trust that our votes are counted
accurately; or is wholesale electronic election tampering occurring?
How could the evidence of vote tampering be hidden? Are the future of
democracy and U.S. elections at stake? The U.S. press has dismissed
exit polls as surprisingly inaccurate in the 2004 presidential election
when exit polls conflicted with official vote counts. Were exit polls
wrong or were vote counts altered?

On February 14, 2006, the National Election Data Archive, a group of
volunteer mathematicians and statisticians, released a report asking
that new measures be taken immediately in order to assure the integrity
of future U.S. election results. Their new report discusses why current
measures to ensure vote count accuracy, such as testing and
certification, are inadequate; discusses how evidence of vote miscounts
are hidden by current election reporting procedures; and recommends
independent vote count audits, public detailed election data monitoring,
and public exit poll data.

The National Election Data Archive’s report also summarizes a scientific
analysis of Ohio’s precinct-level exit polls in layman’s terms. Why
should Americans care about Ohio’s 2004 vote counts? Ohio was a key
battleground state. Whichever presidential candidate won Ohio became
president. Ohio is also the only state for which pollsters publicly
released sufficient precinct-level exit poll and vote count data to
perform a valid mathematical analysis. Exit polls by the same exit
polling firm, Edison/Mitofsky International, were recently used to judge
when elections in the Ukraine and Azerbaijan were valid. Exit polls in
the 2004 presidential election were not just randomly inaccurate:
Bush's reported vote tally was higher than the exit polls anticipated
and the inaccuracy was highest in precincts with the highest reported
Bush vote. The pollsters said: “Bush voters completed fewer exit
polls.” However, the National Election Archive’s analysis finds that the
exit poll error explanation is inconsistent with the data and claims
that John Kerry might be president today, if votes had been accurately
counted in Ohio.


In June 2005 The Election Science Institute (ESI) and pollster Mitofsky
issued a paper “Ohio 2004 Exit Polls: Explaining the Discrepancy” which
asserts that an exit poll error explanation “is much more likely than
the fraud accusation theory to account for most, if not all, of the
observed discrepancy between the exit polls and the actual results.”
Precinct-level exit poll data released with ESI’s report shows that the
overall average discrepancy between Ohio’s exit poll and certified vote
count margins between Kerry and Bush was 11.7 percentage points.

However, In October, 2005, the National Election Archive released a
paper which gives counterexamples to show that the Election Science
Institute’s analysis is based on an invalid premise. On January 17,
2006 the National Election Archive released its own scientific Ohio exit
poll discrepancy analysis, “The Gun Is Smoking: 2004 Ohio Precinct-level
Exit Poll Data Show Virtually Irrefutable Evidence of Vote Miscount” .
This analysis concludes that Ohio’s exit poll discrepancy pattern is
consistent with outcome-altering errors in vote counts.

Two things are certain in this controversy about U.S. exit poll accuracy:
1. The Election Science Institute and the National Election Archive
cannot both be correct, and
2. Any university mathematics department in America could evaluate
the two conflicting studies and decide which analysis is mathematically
correct.

The National Election Archive challenges every journalist interested in
discovering if outcome-altering vote miscounts or exit poll error is the
more probable cause of Ohio’s exit poll discrepancy; to help resolve
this critical question. The answer may make the difference as to whether
Americans take steps to ensure vote count accuracy in future elections
or not. The National Election Archive urges the National Election Pool
media consortium to accept this “math challenge” by sharing these two
conflicting election studies with mathematics faculty at any university
to determine which analysis is mathematically correct.

http://electionarchive.org/ucvAnalysis/OH/Ohio-Exit-Polls-2004.pdf and
http://electionscience.org/reports/view_reports.org
http://electionscience.org/Members/stevenhertzberg/report.2005-07-19.7420722886/report_contents_file/

The survival of democracy and the future of our civilization may depend
on taking steps to ensure the accuracy of elections. As the Election
Science Institute said, “The public has a right to know exactly how
elections work and to verify for themselves that the voting and the
counting is done right.”

NEDA’s full February 16th report can be found online at
http://electionarchive.org/ucvAnalysis/US/exit-polls/Ohio2004-US-future.pdf

[bullimiami]

1. they are pissing in the wind.

[papau]

2. Or the butterfly effect will bring down the right wing election thieves

Hard to predict.

Only sure thing is that the US media will ignore this.

[DBoon]

3. KnR

nt

[JudyM]

4. This is inspiring. I was a volunteer on the ground in OH in 11/04, and I

(like many DUers)have been gnashing my teeth with frustration and outrage at the travesty I believe happened in Ohio. This is an important step - it is saying to the country: let's get down to the facts and may the best math win. It gives me hope that the truth will win out because when the faux news types put up their best mathematicians to the test, their theories can be taken apart systematically. This effort is centralizing the solution to the problem.

:applause:

[Trillo]

5. Found this conclusion curious:

From one of the OP's links

"Our study indicates that the non-response rate theory is much more likely than the fraud accusation theory to account for most, if not all, of the observed discrepancy between the exit polls and the actual results.
The more detailed information allowed us to see that voting patterns were consistent with past results and consistent with exit poll results across precincts.The fact that there is debate over this at all shows that we need elections to be more transparent, more accountable, and more auditable.
http://electionscience.org/Members/stevenhertzberg/report.2005-07-19.7420722886/report_contents_file/

Edited to add: I wonder how voter disenfranchisement, such as exceedingly long lines and limited machines in certain poor neighborhoods, works into either theory.

[Stevepol]

6. The reluctant Bush voter response theory is ridiculous on the face of it

IMO and hardly requires heavy duty statistics to prove or disprove. Has it ever been true in the past? If it's true now, it's always been true and always will be true and you can forget about polling, despite the fact polls are provably more or less cccurate (w/i a quantifiable margin of error and of course in elections that don't use electronic voting machines or where the machines are closely audited) These are the kinds of explanations people pull out of a hat because there's no other way to explain the facts. It fits with the Repub way of dealilng with any question: just come up with the most convenient theory that matches your ideology no matter how lame-brained or ridiculous.