New USCV Scientific Study Released: Hypotheses of Fraud Remain Credible

[sunshinekathy]

Everyone, Hi. This study shows that vote fraud explains the anomalous 80% Bush vote precincts and E/M exit poll patterns (in pictures). FYI, Print the first 11 pages of the pdf, unless you want to read the mathematical appendices. Please give copies to our elected officials, so we can convince them not to concede in Nov 06 and restore correctly elected candidates to the U.S. Congress in Jan 07.
---------------------------------------------

For Immediate Release - Press Conference Today

2004 Presidential Election: Hypotheses of Fraud Remain Credible; New Scientific Study Released

http://uscountvotes.org/ucvAnalysis/US/exit-polls/USCV_exit_poll_simulations.pdf

Contact: Kathy Dopp, US Count Votes, President

[phone] [e-mail] electionarchive.org

The persistence of credible hypotheses of election fraud, six months after the election, underscores the fragility of the U.S. electoral system. US Count Votes continues its systematic statistical study of the discrepancy between the Edison-Mitofsky exit polls and November's reported presidential election results.

Miami, FL. – Ron Baiman, Ph.D. of US Count Votes and the Institute of Government and Public Affairs of Chicago, will release the new results at the meeting of the American Association of Political Opinion Researchers today, Saturday at a 2:15 p.m. Press Conference in the Hotel Fontainebleau Hilton Resort lobby, 4441 Collins Avenue, Miami Beach, FL 33140.

Peter Pekarsky, the lead attorney being sued by Ohio Secretary of State Kenneth Blackwell, will be available to talk about what has happened in Ohio, and why the election needs to be investigated.

Mitofsky, of the Edison/Mitofsky group who released an analysis on January 19th of their November 2nd exit poll that had predicted a strong win for Kerry, will also be in attendance at the AAPOR conference.

The National Election Data Archive (NEDA) today has released a new report, demonstrating that data from the Edison/Mitofsky analysis is consistent with the hypothesis of a corrupted vote count, and inconsistent with the competing idea that Bush voters were under-sampled in the poll. Using numerical modeling techniques to simulate the effect of polling bias, NEDA scientists are able to reproduce signature patterns in the Edison/Mitofsky data by incorporating a general shift in the official vote tally in the model.

Most telling is the fact that the highest participation rates and the peak disparity between poll and official returns both occurred in precincts where Bush made his strongest showing. This feature of the data is inconsistent with the Edison/Mitofsky assumption that polling
bias was responsible for the gap.

For the complete report, see
http://uscountvotes.org/ucvAnalysis/US/exit-polls/USCV_exit_poll_simulations.pdf

This paper follows an earlier study released on March 31, 2005, by a group of statisticians for the National Election Data Archive Project, Analysis of the 2004 Presidential Election Exit Poll Discrepancies.

[pauldp]

1. Kick! n/t

[Peace Patriot]

38. Profound thanks to you, Kathy, and to other experts who have been...

...working on 2004 election data with clean hearts and clean hands, and despite the considerable danger involved in challenging the Bush regime and its lapdog news monopolies. Frankly, I can't think of anything more difficult or more dangerous.

Be advised: My donation is on its way. It's not big, but I hope it reminds others here that challenging these powers in an organized way needs our support. DUers!!!! Wake up! Contribute whatever you can NOW!

You may mail checks to:
USCountVotes
P.O. Box 682556
Park City, UT 84068

Or make an on-line donation at:
http://www.uscountvotes.org/index.php?option=com_content&task=view&id=13&Itemid=40

-----

The lack of verifiability of the election, and all the confusion that Edison-Mitofsky has created, seem quite planned to me. Why should there be any confusion about what the official tally or the exit polls say?

E/M has created confusion by the type of polling they did, by "adjusting" the polls to fit the official tally (depriving Americans of the real exit poll result), by withholding data, by eking out data and explanations in fits and starts, and by tossing out unsupported theories to explain it all. Why do this?

If they had wanted to verify the election, they could easily have done so. And this election system cried out for proper verification procedures. Many aspects of the election system were controversial, or just plain absurd, including the ownership and control of the secret vote counting software by major donors to the Bush campaign, the lack of a paper trail, and the insecurity and hackability of electronic voting machines and tabulators. So, why didn't Edison-Mitofsky do a poll specifically designed to verify the vote count--as is done in every other democracy in the world?

I have asked this question in other posts at DU, and naysayers have replied that the private business interests who commissioned the poll--that is, all the major TV networks that are using our public airwaves for private gain--had the right to commission whatever kind of poll they wanted, and, if somebody else--the general public, the Democratic Party--wanted verification, they should have paid for it themselves.

This reply is much too simplistic. It fails to acknowledge the reality of the overweening influence that these TV networks have upon our elections, including the billions of dollars they receive in political campaign advertising--which has turned our political life into a corporate pigsty--the untimely prediction of winners (so critical in the year 2000 Florida situation, and also in 2004), and their control over political debate--which has had such a corrosive, trivializing impact on American political life.

Their use of our public airwaves is a privilege, not a right. They are supposed to earn that privilege through public service. They furthermore have an obligation as journalists to monitor, investigate and oppose the government--an obligation that they have repeatedly failed at, throughout the Bush presidency--never so much so as by their commission of exit polls that would be fitted to the official result. Why kind of journalism is that? It is lapdog journalism. There is no other word for it.

The fact is that nobody can prove Bush won this election, and that is the problem.

The official tally was under the control of two corporations, both of them run by Bush partisans, with ample means, motive and opportunity to produce whatever result they wanted. The election system was characterized by severe insecurity and utter lack of transparency. If you try to track it back--from the official results on TV, to the AP computers into which the results were fed, to the state and county computers that were compiling reports from the precincts, back to the voters casting their votes, you are brought to a halt somewhere around the central electronic tabulators in the states and counties. No one knows how they tabulate the results from the precincts. No one--not even state election officials--has the right to review the tabulation software. And if you travel past this rather big hurdle back to the voter, you find NO RECORD OF HOW THEY VOTED that can be independently checked, in one third of the voting machines in the country, and little or no auditing of the machines that do have some kind of record.

Then you find numerous reports (in the Election Incident Reports) of people voting for Kerry and the MACHINES CHANGING their votes to Bush (in 86 out of 88 incidents), or defaulting to Bush, or deleting Kerry votes in a party vote ballot, or counting backwards, deleting thousands of votes, or showing 4,000 votes for Bush in a precinct with 600 registered voters--with virtually all these "anomalies" favoring Bush and none favoring Kerry--an impossibility.

Kerry had near unbeatable advantages going into this election (huge Democratic success in new voter registration in 2004--nearly 60/40; miserable Bush approval ratings; big jump of Nader voters to Kerry; big turnout--always favors Democrats, etc., etc.) And, IN ADDITION, the exit polls showed a Kerry win.

Against all this evidence, what the Bush regime and the news monopolies have is an unprovable election. And so they just made stuff up. Their exit polls showed a Kerry win--so they changed them! Alert citizens and statistics experts got hold of the real exit poll result, so they invented a theory, out of whole cloth, for why their exit polls were wrong (and won't or can't support it with the data). The voter registration statistics showed a huge Democratic success in new voter registrations, so they promulgated Karl Rove and Dick Cheney's "talkings points" about the Republicans' "invisible" voter registration and GOTV campaign.

US Count Votes is the clear-eyed child in the street watching the Emperor's parade go by--and crying out, "The Emperor has no clothes!"

The "adults" in the crowd--those with a vested interest in contributing to the Emperor's delusion that he is clothed--refuse to see the obvious, and can't or wont' say it.

If the Bush regime had wanted a transparent, verifiable election, why didn't we have one?

This is the situation we are dealing with--one of pervasive lying, power-mongering. vast corruption, war profiteering, and worse. The folks at US Count Votes are extraordinarily brave to take it on. I salute you!

Now, take a deep breath! Get some rest! And remember, it's not up to you alone. It's up to all of us.

--------

DUers! Any spare bucks you have in your pockets--or even unspare ones that you would be willing to part with, for this most worthy of causes--do so now!

[helderheid]

39. I have contributed

time, money, and energy to UsCountVotes and UtahCountVotes.

[Teaser]

2. Does this reflect Febble's critique of the fraud hypothesis?

Have to see.

[Febble]

4. According to my contacts at AAPOR

Mitofsky revealed his data today in the form of a pair of scatterplots and correlation coefficients that showed that there was absolutely no difference between either the participation rates, or the disparity between poll and return, between precincts with a high vote count for Bush and precincts with a high vote count for Kerry.

Mitofsky had in fact used my "Fancy Function" (algebraically equivalent to formulae used by USCV) to convert the WPEs into an index of "bias" (which, I will once again make clear, makes no distinction between bias in poll or count) that is mathematically independent of vote-count margin itself. This makes any correlation between bias and vote count valid.

Both correlation coefficients were insignificantly different from zero. Moreover, the plots showed large misestimates of the vote count in BOTH directions throughout the range of precincts, with as many precincts with "implausibly" large pro-Kerry errors as precincts with pro-Bush errors. The mean error was certainly significantly in the direction of an over-estimate of Kerry's vote-count, as we know, but the scale of the mean mis-estimate in Kerry's direction was dwarfed by the sheer scale of the mis-estimates in both directions.

The E-M report contains a very misleading table, whereby mean and median WPEs are given by "partisanship category", and it was on this table that USCV based their model.

Unfortunately, categorizing a continuous variable, like precinct partisanship (by which they meant proportion of votes cast for each candidate) is always a bad idea, as it means that not only are your category boundaries arbitrary, but you have much smaller numbers in the categories at the extreme ends of the range. This tends to distort your mean values, because the smaller the number of data-points you have in a category, the more any extreme values exert "leverage" on the mean value. The data turns out to be extremely "noisy", and extreme values were scattered all over the plot. Surprisingly, many precincts reported large over-estimates of Bush's vote count as well as precincts with large over-estimates of votes for Kerry. Several precincts with large vote counts for Bush showed large over-estimates of Bush's vote.

There was also a huge amount of variance in the participation rates, ranging from near 100% to a very small percentage. Median participation rates were low.

What Mitofsky presented today still doesn't rule out fraud - but the data shown does not support the hypothesis that votes were switched systematically between Kerry and Bush in the manner modelled by Dopp, Baiman et al. The pattern was quite different from their model. If fraud occurred, the data shown by Mitofsky today suggests that it may have occurred in both directions, but that the Bushies were marginally (by a very small degree) better at it.

More likely (IMO) is that the polls were really very inaccurate, with large mis-estimates in both directions. What caused the average misestimate to be slightly more in the direction of Kerry's vote than Bush's may never be known - maybe it was fraud, maybe it was differential non-response. But if it was fraud, it was widespread, not confined to any kind of precinct (as defined by vote-count margin), and tiny in relation to the very large over-estimates of both Kerry's and of Bush's share of the vote in a great many precincts.

My claim that such a pattern was possible, given the rather meagre data we were given in the E-M reported, appears therefore to be supported. The pattern inferred by USCV was not apparent in the full data set as shown by Mitosfky.

[gristy]

6. ..

[Sawyer]

7. ..deleted

since gristy decided to erase his/her comments.

[Febble]

9. OK

Sorry, I assumed people were following Kathy's story.

I was a member of the USCV team until recently, and was even a signatory to the early version of their report.

I found myself unable to sign the later version for a number of reasons, and published a dissenting report here:

http://www.geocities.com/lizzielid/WPEpaper.pdf

diaried by DemfromCT on DailyKos here:

http://www.dailykos.com/story/2005/4/29/161938/921

and by Mystery Pollster here:

http://www.mysterypollster.com/main/2005/04/the_liddle_mode.html.

The new USCV report was originally billed as a rebuttal to my paper, but emerged partially through friendly correspondence between the USCV team and myself.

My "fancy function" was simply an algebraic formula for transforming the WPE ("within-precinct error - a measure of the discrepancy between the poll and the count for each precinct) into a value that would be a pure measure of "bias" - the degree to which either the vote, or the count, was "biased" in favour of one candidate.

Hope this helps.

Sorry not to have provided links in early posts.

Oh - AAPOR - American Association for Public Opinion Research - a big pollster's conference currently underway in Miami, at which USCV were launching their new paper. Warren Mitofsky is the the head of Mitofsky International, one of the two polling companies who conducted the 2005 exit polls. He gave a presentation on the exit poll data at a lunchtime plenary session today.

No cut and paste in my post. Boy, are people quick on the draw around here.

[sunshinekathy]

17. Not quite the way it happened

Let's be truthful Elizabeth. You began publishing dissenting opinions on web sites without once running it by us first for review or discussing it with us.

I do not call that being a team player.

We offered to work with you on our last paper until we could all agree, and you declined. You rushed to publish your derivations of a few new formulas independently, along with your own (in my opinion) unjustified conclusions.

I've shown how simple it is to derive the three formulas that you derived on one page in our work, in Appendix C and D, since you did not reveal how you derived them yourself, and generally obfuscated the math and the issues in your own paper. Everyone else who contributed derivations of new formulas contributed them to the group and wrote up the steps to derive them so we could all easily understand, but not you.

If anyone wants to UNDERSTAND the formulas you are using clearly, they can go to our paper, which shows how easily they were derived from all our previous work.

You also obfuscated by changing all the notation we'd began with and complicating it, by calling precincts defined by percentage of votes for Kerry or Bush, Democratic or Republican precincts, etc.

[Febble]

19. Well I'm not going to get into this Kathy

I am not claiming to be a team player.

I am sorry I misrepresented the wording of your new paper - I had not done a recent download.

If we are now in agreement that the data are consistent with constant mean bias, then there is no problem.

And perhaps the personal stuff would be best left to email.

If someone could tell me how to delete a post, I will delete my posts.

Lizzie

[sunshinekathy]

20. We do NOT agree that the data is consistent with a constant mean bias

Elizabeth,

I just spoke to Ron Baiman at the conference who listened to E/M speak and he said that no evidence was offered at the conference to show that the data is consistent with a constant mean bias as you suggest, just a big blob of dots on a scatter plot which didn't show anything in particular.

Perhaps you are right, but we don't know that yet.

Also, the data IS consistent with a vote shift.

Best,

Kathy

[Febble]

21. OK

I've asked the moderators to delete my posts, and sent you an email.

Perhaps you would also delete your post regarding my "obfuscations". At no point have I attempted to obfuscate, and if there were misunderstandings over the algebra, they went both ways, as you know.

Lizzie

[sunshinekathy]

8. Hi Elizabeth

That seems hard to believe Elizabeth (the claims re. the Bush > 80% precinct data). I just tried to call Ron Baiman, to see what he said, but his reception was bad. It will be interesting to see the scatterplot. Is it posted somewhere?

We still need to know a lot more about the data. i.e. Did they release the vote counting methodology or the size of place data that go along with each precinct?

Did they release their adjustment or weighting information?

You said the variance had a huge range. How do you mean that? What was the overall variance?

You seem to imply that their data is inconsistent with vote embezzlement. How did you determine that?

You said that our model (same as yours) was different from the pattern. Is there any model that would be consistent with their pattern? Are you saying that increasing the standard deviations or varying them in our model, would not be able to approximate the data?

Elizabeth, if you lived in America and were doing the on the ground research that many dedicated voting activists are doing here, you would know that votes are being embezzled ubiquitously here, using many different methods. I got another call this a.m. from someone who is doing laborious investigations in Arizona, where there has been a well organized effort which has been embezzling votes for years.

Your conclusions continue to fly in the face of reality here in America.

If what you say is true, any vote embezzlement may be harder to show it with exit poll analysis, so we need to be able to organize and execute the National Election Data Archive Project to investigate the detailed election results, which will definately be much more revealing than this data which is not being released openly in enough detail to truly know.

You claim that what we "infer" is wrong. Just exactly what did we infer that is wrong specifically? Please quote an exact statement from our report and show us how the data refutes it, rather than making vague allegations that may or may not be true. You seem to contradict yourself in your statement above, first saying the E-M data has refuted the hypothesis of vote shifts, and then saying it hasn't.

Kathy

[Febble]

11. Hi Kathy

I am not saying, and have never said, as you know, that I don't believe that vote embezzlement (great phrase!) occurred.

But I cannot say data says one thing when it says another. If you are convinced fraud occurred then either it took a different form to the one you modelled, or it occurred in precincts that weren't part of the exit poll sample. Both are possible.

I am not concluding. I am noting. I am noting that the pattern of the data presented by Mitofsky today showed a non-existent correlation between either discrepancy or participation rate and vote-count margin. Your model predicted that it would have such a relationship.

The variance was huge. I do not have the data on the variance, but there were large numbers of data points on both sides of the zero line throughout the range, and large numbers of extreme data points on both sides, again throughout the range, as I suspected from the absolute WPE values. And the explanation for the non-existent participation rate correlation (either way) was that again the variance was enormous.

I expect the scatterplots will be made available soon.

Lizzie

[Sancho]

45. Come on Feeble...use your influence and head...

You and most researchers know that a scatterplot looking for a simple a linear relationship on all that data is not the answer. Have YOU asked your buddies at EM to give up the raw data? Let's profile the responders! Correlate the answers to the voter rolls...one person at a time!

Why continue the fake pursuit of an answer when all we need is to see the paper? This is tougher than the Dead Sea Scrolls...sort of an exercise in manipulating mush.

It's time for YOU to call for the data....otherwise, no coefficients or weights or plots are anything but a fun exercise in imagination. Let's see what EM has in the closet.

[Febble]

5. And it is probably worth noting

that several signatories to the original paper are not signatories to the current paper, although I believe there is one additional signatory.

I do not have the sense that this is a majority report.

Modelling is a great tool, and I started the ball rolling with my paper, but in the end, models are models and data is data.

The USCV model predicted that the poll-count discrepancy would be higher in high Bush precincts than in high Kerry precincts.

It turns out that the correlation in the real data was insignificantly different from zero. I did not know this would be the case when I constructed my model - all my model told me was that it could be.

It seems that it is.

The major difference between my model and the one in the new USCV paper lies in the assumptions regarding variance. And it turns out that even my assumptions were too conservative.

The data is noisier than our wildest nightmares, but there is no relationship between either participation rates or discrepancy rates and vote count margin.

No fingerprints.

[sunshinekathy]

10. Elizabeth -you are once again mischaracterizing our paper

I hope people will read our report for themselves and not trust your mischaracterization of it:

http://uscountvotes.org/ucvAnalysis/US/exit-polls/USCV_exit_poll_simulations.pdf

We did not say that WPE would be higher in high Bush precincts than in high Kerry precincts, for example, nor anything close.

We said that vote embezzlement would shift the shape of the curve by shifting precincts from the Kerry to the Bush side of the curve, so that the shape of the WPE curve would show "higher" (actually higher absolute or lower) WPE in high Bush precincts as compared to a WPE pattern caused by response rate. You are really distorting what we said and misrepresenting our statements, which perhaps you did not fully understand.

Please re-read our paper.

I will reserve judgement on the scatterplot until I see it.


We obviously knew that we did not know what the variance looked like without first seeing the data. You do not read our minds Elizabeth, so don't assume that you know what we do and do not know, without first asking us.

I would appreciate it if you would stop misinterpreting our papers to everyone and desist from mischaracterizing us.

Kathy

[Febble]

12. Well itwould also help

If the paper didn't look different every time I download it.

Nonetheless, the latest download states:

New evidence from mathematical simulations conclusively shows that any constant mean exit poll
response bias hypothesis such as the "reluctant Bush responder" (rBr) hypothesis is not consistent
with the pattern shown by the Edison/Mitofsky exit polling data.

And "constant mean exit poll response bias" is what Mitofsky showed today.

I am sorry if I mischaracterized the latest version of your paper.

If it is now consistent with the data Mitofsky showed today, fine.

But in that case it is not consistent with the vote-shift hypothesis.

[sunshinekathy]

13. But it IS still consistent with the vote-shift hypothesis

Elizabeth,

How did you derive the logic that says: If the data is consistent with one theory that it follows that it is inconsistent with another?

Not in any rational logic I've seen.

Can you prove that?

Now you are once again vascillating. One minute you say one thing, the next, the opposite.

Kathy

[Febble]

14. OK Kathy

I have been saying precisely this all along. That the data are consistent with constant mean bias. I have also been saying that they do not rule out fraud.

USCV have been saying that constant mean bias is impossible.

Well, it isn't impossible.

But it requires re-interpretation. The original claim was that the data must show higher WPEs in high Bush precincts. They didn't.

If you now agree with my paper fine. It seems consistent with the data.

But a week or so ago you were very keen on DU to announce that your models had refuted mine.

My models are consistent with the data, and my models were of constant mean bias.

I've taken a lot of flak over here on DU over this. I'm not attacking USCV if USCV now agrees with my paper. But I should at least be able to defend myself.

[tommcintyre]

15. "...mischaracterizing us". In politics, it's called "spin". ;)

http://onlinedictionary.datasegment.com/word/spin
Definition : spin

Spin \Spin\, n.

3. (Politics) an interpretation of an event which is
favorable to the interpreter or to the person s/he
supports. A person whose task is to provide such
interpretations for public relations purposes is called a
spin doctor.

<1913 Webster>

<From politicians I expect this sort of subjective and biased behavior - but NOT from an objective (self-proclaimed) statistician.>

I felt Febble started spinning her position BEFORE you even posted the new paper (a couple of days ago in your thread simply requesting publicizing assistance. I'll be glad to point out the specific incidences of spin there, if requested ;) ).
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x368296

Thanks for reposting here Kathy. You just saved me a lot of "refutation work" on her threads above. ;)

Keep up the good work - the truth IS coming out, in no small part to your efforts.

Edit for grammar

[Febble]

18. No I was not spinning.

Or at least no more than Kathy has been doing.

She promised a "rebuttal" of my paper.

My paper predicted a certain pattern in the data - or at least said a certain pattern was possible. Initially USCV said this was impossible. It appears that new wording of the new paper doesn't quite say this.

Fine. In that case I have no problem.

But in that case their new paper is not a rebuttal.

As we were.

[Sawyer]

22. Any explanation as to why

several authors of the previous paper have not signed under this one?

[sunshinekathy]

24. They have not had time to study it thoroughly

Ron Baiman was asked this same question at the AAPOR conference by Mitofsky and

Ron asked Mitofsky "Why did not more people sign Your paper?"

It is not important who signed it, it is the quality and accuracy of the analysis that is important.

Ron Baiman and I did most of the derivations of the formulas and we wrote this collaborative paper miraculously in about one week (It took us almost two months the last time.)

It was almost an impossible task to get this paper done. I haven't slept more than 3 hours in the last two nights going over and over it.

There was not enough time for others on the list to get comfortable with the material and up to speed and you can't ask people to sign something until they've had a chance to review it.

The wonderful thing about US Count Votes' papers though is that dozens of statisticians or math-types have helped read through them and helped edit them for accuracy of language, including this paper.

While there are only 4 or 5 signees, at least a dozen highly intelligent and talented persons have reviewed it.

It takes a lot of courage to sign the type of work that we're doing and some of the folks who help us have jobs that might be at risk or are afraid for their children or themselves.

The fact that Paul Velleman, PhD statistician from Cornell signed it with us, says it all. Paul Velleman would not have signed something that was mathematically shakey.

Best,

Kathy

[Sawyer]

25. It is indicative if some people sign the paper

but do not sign the followup paper. So let me ask you directly:

Do some of the original signatories dispute the findings in this followup paper?

Are some of the non-signees coming out with rebuttals of the followup paper?

[gristy]

3. Ooooooooh

Reading it now, Kathy. :bounce:

[tommcintyre]

16. NOMINATED! We want ALL of DU to know about this. n/t

[sunshinekathy]

23. Liddle Appears to be working with Mitofsky nowadays

FYI Everyone,

Elizabeth has just admited on another thread that she has been working directly with Mitofsky. Please keep that in mind.

Ron Baiman observed that Liddle seemed to working with Mitosky at the conference and I observed the "inside" information that Elizabeth Liddle claims to have.

Perhaps her obvious efforts to discredit the work of US Count Votes has some other motivations of which we are not aware.

Please judge for yourself the veracity of the work of US Count Votes by reading our work.

Thank you.

http://uscountvotes.org/ucvAnalysis/US/exit-polls/USCV_exit_poll_simulations.pdf

I tried to pull most of the technical materials into the appendices so you can just print and read the first 11 pages and skip the appendices.

Please print and read the 11 pages. They are well worth it IMO.

And Please pass them out to your elected representatives so that we can educate them in time for the Nov 06 election.

Best,

Kathy

[tommcintyre]

33. So it turns out that "Febble" is nothing more than a Mitofsky rBr "shill"

http://www.answers.com/topic/confidence-trick
"The confidence trickster, con man, scam artist or con artist <Mitofsky> often works with an accomplice called the shill <Febble>, who tries to encourage the mark by pretending to believe the trickster."

Febble said the following a couple of days ago, And it got me to thinking, how "dirty" ARE Mitofsky's hands? So, I've done a little research, and what I have found will definitely, solidly, raise some valid questions on Mitofsky's motives, etc.

I'll finish putting it together, and post a thread with the information tomorrow.

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x368296#368417
"Frankly, the thought of a respected academic and social scientist making up data is completely unlikely. Academics will argue vociferously about interpretation of data, and may be biased in their interpretation of data, but the data is absolutely sacrosanct. Data fabrication is a hanging offense in academia, and the chance of someone of Mitofsky's standing indulging in it seems beyond belief."


Hmm... as they say: "if the shoe fits, wear it".

[autorank]

34. Check your PM's. auto. Thanks

[tommcintyre]

37. Kick for "full disclosure" and SCV's hard work n/t

[Filius Nullius]

26. Bottom Line?

O.K.,Kathy, bottom line. I hate to betray my ignorance but is there a 10% discrepancy between exit poll results and final reported votes in favor of Bush in Republican stronghold precincts, or not? Is there a .3% discrepancy in favor of Kerry in Democratic stronghold precincts? I understand that you may not be able to say definitively yet, but can you give us your gut feeling? If you can't, just say so, and we will simply have to wait until you are ready.

I believe you need to find a way to make this stuff more intelligible to the average Joe/Jo Schmo voter. I have to tell you that my eyes don't just glaze over when I'm reading parts of these statistical reports--they triple-glaze over!!! I mean, WTF is a "Gaussian distribution," anyway, and what is all of this natural log stuff? I know what natural logarithms are, but it is about as clear as mud to me what you are doing with them. I took two honors courses in calculus and a course in statistics as an elective in college, but that is the extent of my university math (I was a liberal arts major), and it was a long time ago. I doubt very seriously that many people who do not have heavy math backgrounds are going to take the time to read these reports. If you want the great statistically unwashed masses to have the remotest idea of what you are talking about, you need to do a better job of translating. Can you produce an "Election Fraud for Dummies" version of each of the reports? If not for each of them, could you provide a white paper that explains some of the terminology and methodology used in the reports? Maybe ask a neighbor to read it who never took a course in statistics to see if he or she has trouble understanding it.

Is Febble an ex-patriot American or a different nationality? I got that feeling from one of your posts. Notwithstanding her possible ex-patriot or alien status, I also get the feeling that she may have been a Republican mole in your midst. Have you had other "defections" from the original report?

I also do not understand why you cannot get the information you need from E/M. According to the National Election Pool Web site (http://www.exit-poll.net/pool.html):

"The exit poll data is available for purchase by TV-affiliates, newspapers, internet sites and other news outlets. The polling data will be delivered through a secure web application directly to subscribers on primary and election days. Political candidates may subscribe as well, but the data is available one week past voting day."

Does it cost too much to buy, or is some of the data you need not in the data set?

[sunshinekathy]

28. Please read our paper's conclusions, recommendations, & future sections

Please read our work. I'm very tired and overworked. I need help and we urgently need funding. The work is here:

http://uscountvotes.org/ucvAnalysis/US/exit-polls/USCV_exit_poll_simulations.pdf

Just print and read through page 11, skip the appendices unless you want to check our math.

We're doing our best. Much of this report is written in plain English, with the technical stuff deliberately stuffed into the appendices to make it easier to read then any of our prior papers. I'm sure you can understand the Conclusion, Recommendations, and Future sections.

A Gaussian distribution is a symetrical distribution, sort of like a normal curve.

I'm sorry our papers don't pass muster for you, but we are planning to submit this to a refereed journal, so we have to pass muster in the academic world amonst PhD stats types. If I wrote the paper the way you want it, none of the PhDs would sign it because it has to be very precise. You have no clue of how many agonizing endless hours it takes to get a bunch of PhD mathematicians to sign something like this. As an unpaid volunteer, I'm thinking I do not want to do any more of these. I've already done probably ten collaborative papers with groups of PhDs this year to try to influence America to stop destroying democracy and start reconstructing it, but it has had little effect. No one seems to listen to the experts, especially not the Democrats, who now, I hear are trying to federalize elections by passing a constitutional amendment for a federalize voting rights act - so we can put the Bush administration in charge of elections? As if passing the HAVA act wasn't idiotic enough!? The Democratic leadership must have a death wish or something. They always seem to find the solution for "fixing" elections that makes it tens times worse instead, and very obviously so to the experts, who are doing everything they know how to get people to pay attention and listen.

Liddle is a native of England. I don't think she started out as a mole, but she isn't a team player, and she is being a very destructive influence to our efforts to restore democratic elections, that is for sure. She quit our group several weeks ago at least. I don't recall the exact date. It may've been a month ago. She seems to be getting a lot of the credit for work that we did, in addition to her own. She could not have derived the formulas she did without our deriving the ones we did first. Whereas everyone else in our group, when they derived some new formulas, typed them up to explain their derivations and shared them with the group, Elizabeth decided to publish on her own, and obfuscate the derivation of her formulas IMO.

If anyone wants to completely understand Liddle's work and the results that analysis like hers produce, they have merely to read US Count Votes' papers because it is included there as a part of ours.

I don't know about the raw data. I'm trying to get off this topic and on to the task of restoring democratic elections in America by developing the National Election Data Archive, but we have not received the funding or the help obtaining the funding that would be needed.

Please read our paper - all first 11 pages.

Kathy

[Filius Nullius]

30. Sorry, Kathy. Don't be discouraged.

I'm on your side, and I'm not trying to frazzle you any more than you already are. I just want your work to achieve the greatest currency possible.

[vince3]

32. Thanks for all the tremendous work

[helderheid]

27. let me know if you need help faxing Utah

[sunshinekathy]

29. Why do you want to "fax Utah"? What does that mean?

[helderheid]

31. Kathy - it's me - Clarity

I can fax the media and politicians in Utah. Let me know if you want that help.

[sunshinekathy]

41. We could "fax Utah" regarding Utah's voting equipment selection

Clarity,

That's an excellent idea.

Let's write a press release and fax Utah regarding Utah's selection of voting equipment.

Can you write a draft press release and I'll help you edit it?

We need to make one more try to prevent Utah from getting DRE voting equipment.

Best,

Kathy

[helderheid]

46. Yes

Let's give Utah one more go. Thanks. Email me and let's talk tomorrow.

[tandem5]

35. This is great, thanks! nt

[kster]

36. kick.nt

[Coexist]

40. Wow - amazing

my mouth won't close - help somebody!

[althecat]

42. Fantastic stuff... have been waiting for this...

.. I always felt certain that if the precinct data was analysed closely it would show this sort of clear evidence.

Now to organise a huge show of force in D.C...

[tommcintyre]

43. kick for it's significance! n/t

[Peace Patriot]

44. Kick! Kudos to US Count Votes! n/t