MUST READ: State-by-State Analysis That Non-Statisticians Can Understand

[Helga Scow Stern]

From MadLinguist:
just posted this paper to truemajority and insert the link here because I don't have enough posts start a thread.

Here is a state-by-state analysis that non-statisticians can understand.
http://www.freepress.org/departments/display/19/2004/997

I think it deserves some attention.


Thanks, MadLinguist!

[Helga Scow Stern]

1. Favorite part so far:

Regarding the uncooked CNN exit poll data used by Simon and Freeman:

Aside from some possible minor errors in calculating overall candidate vote shares from the male and female vote shares shown on the screen, these can be regarded as relatively unadjusted real exit poll data. These are the best exit sample data for the 2004 U.S.
election currently available.

In contrast, the “final” exit poll data analyzed in the CalTech/MIT study, and discussed by Corn and Levin, are useless as an indicator of exit-poll/actual-vote divergence.4 This confusion over the data (or lack of investigatory diligence as this point is spelled out in
great detail in the widely publicized Freeman and Simon pieces) has however sufficed to convince the mainstream media that there was no unexplained exit poll discrepancy.

[gulogulo]

4. The "final" exit polls have been pretty accurate in the past

the "raw" exit polls, on the other hand, have been known to be wildly wrong (something that apparently your article's author is not aware of):

http://www.washingtonmonthly.com/archives/individual/2004_11/005178.php

He looks at the previous four presidential elections' exit polls

"As you can see, the raw exit poll results always overstate the Democratic vote, sometimes by as much as eight percentage points. So the fact that the raw results this year overstated Kerry's actual vote tally is hardly cause for alarm."

So, as you see, 2004 is hardly an exception in how wrong the "raw" exit polls were.

(Edit: URL)

[Helga Scow Stern]

5. Did you read the article, in which the reasons for exit poll error are

thoroughly discussed?

[gulogulo]

7. yes I did -

did you see the link I provided?

As I have shown, the "raw" exit polls are NOT reliable. The national raw exit polls have been off as much as 8% in previous presidential election (1988). The paper that you provided does not address this at all.

If you want to look at foreign exit polls and their reliability, take a look at:

http://www.aceproject.org/main/english/lf/lfd08e.htm

"...their reliability can be questionable. One might think that there is no reason why voters in stable democracies should conceal or lie about how they have voted, especially because nobody is under any obligation to answer in an exit poll. But in practice they often do. The majority of exit polls carried out in European countries over the past years have been failures."

(ACE Project is a joint venture of the International Foundation for Election Systems (IFES), the International Institute for Democracy and Electoral Assistance (IDEA) and the United Nations Department of Economic and Social Affairs (UNDESA))

And if you want to see research someone has done on the inaccuracy of exit polling in general because of various factors, including such things as time of day and age of the pollsters, see

http://www.cpsa-acsp.ca/papers-2004/Docherty%20and%20others.pdf

[Helga Scow Stern]

10. Thank you, gulogulo.

When I return from dinner, I will look into these sources you have provided.

[RaulVB]

21. No studies are necessary to understand the main point

That is, even if the exit polls are not accurate in percentages, they are very accurate in predicting the outcome of the race, that means who will be the winner.

If the exit polls contradict the actual "results" you need to look at tampering like the most likely cause.

I don't have to remind you that based on that rule the US demanded the resignation of Edouard Shevarnadze in the former soviet republic of Georgia, when he claimed victory after the exit polls showed him losing.

[gulogulo]

24. And if you look at the link I provided

http://www.washingtonmonthly.com/archives/individual/2004_11/005178.php

you will see that raw exit polls projected Dukakis winning. Do you remember the term of president Dukakis? I don't.

[RaulVB]

30. Wrong

Try again.

The exit polls made early in the day when the 1988 election took place, never "projected Dukakis as the winner."

Sorry, I'm kind of a "political junkie" and I followed election day back in 1988 with great interest.

Better luck next time.

[gulogulo]

33. If you are claiming that Kevin Drum

is bending the truth in that article of his, please take it up with him. If you have better numbers for those exit polls, please present them.

http://www.washingtonmonthly.com/archives/individual/2004_11/005178.php

[zann725]

26. We also need to remember that the 51/49 Election results....

mirrored the 51/49 program results of Clint Curtis' program designed in 2000. What are the odds of that?

[Helga Scow Stern]

8. Of course the final exit polls are "accurate" -- they've been weighted!

It says right on page one:

The final “official” exit poll numbers, however, cannot be seen as independent predictors as they have been adjusted to match the actual reported outcome. The “final” numbers are not meant to be independent predictors of the outcome but rather a data source on who voted and why.

[gulogulo]

9. But it is a fact that

raw exit polls are NOT accurate and have been even more inaccurate in the past.

Basically here is the question: do you think 1988 election was fraudulent because the difference between raw exit polls and the election results was more than 8%?

[americanwoman]

11. semantics trouble ...

We need to be clear that there were likely NO exit poll results released that were "raw". They had to be weighted to account for over-sampling of various populations. That is fine. The problem is that the final exit poll results were calibrated to MATCH the actual election results and thus are no longer usable as a predictive measure.

[L. Coyote]

36. Hooray, a great point. n/t

[IndyOp]

12. Gulogulo -- I looked at the WaPo article before and I see the same

flaw I see now:

The likelihood that one poll (the national result) would be wrong is lower than the probability that the exit polls of 10 out of 11 swing states would be wrong and wrong all in the same direction!

What is the probability that if you flip a fair coin 10 times you will get heads each time? Very, very, very low. The WaPo article looks only that the national poll -- that is one poll.

What I want an enterprising conservative lad/lassie to do is to go dig up the state-by-state exit polls from the last 4-6 presidential elections and show us how many states are outside the MOE in each direction, how many states are inside the MOE in each direction.

If, in the past 4-6 elections (before the Republicans started wisely investing in voting machine companies), there is an equally biased pattern of state exit poll and vote tabulation discrepancies then I'll be ready to get rid of exit polling too!

:bounce:

[gulogulo]

14. IndyOp - that national raw exit poll is not

independent of the state polls, you know. It is composed of those polls, and is wrong *because* the state polls are wrong.

Your coin example: to get heads 10 times out of 10 the probability would be 1/1024th. What you are presuming, though, is that every one of the 10 events is independent. That is not the correct presumption in the case of state exit polls. The state exit polls are not independent of each other - they are conducted by the same company, and if the methodology is flawed in one state, it will be flawed in all of them.

Basically, the fact that for the last 5 presidential elections EVERY national raw exit poll was skewed to Democrats can mean one of two things:

1. The last five presidential elections were fraudulent.

2. There is a systemic problem with exit polls that skews them for Democrats.

Can you suggest another alternative?

Did you see the other two links I gave above? One discusses similar lack of accuracy of foreign exit polls and the other is a study on what can influence exit polling accuracy (and it seems that such seemingly innocuous things as the age of pollsters can skew the polls).

[IndyOp]

16. Exit polls may be flawed -- equally likely the vote tabulation is flawed

IF the methodology is accurate then the state polls are independent of each other. Given that Mitofsky International won't release the methodology and data obtained we really can't have a useful discussion about whether or how the methods may have been flawed. We can guess and debate, but we can't know.

As of right now "flawed methods in exit polls" is as much of an hypothesis as is "flawed vote tabulation" (unless you choose to cling to the naive claim that fraud this big *just isn't possible*). The best way to solve this problem is to investigate everything -- vote tabulation and exit polls. Because I care more about the vote than the exit polls -- I want to validate the vote.

If you really don't think election fraud this big is possible, please go read about the voting systems: Optech II Eagle Optical Scan readers with modems inside (the tallies of which can be changed using cell-phone technology), Windows-based PC's running Microsoft access to tabulate votes (that can be hacked via modem), touch screen machines that can be set to Bush as the default setting and/or that can be programmed to assign votes correctly unless the candidate you want to have win starts to drop below a chosen value (51% or 54%) at which point votes are assigned so that the vote percentage gets back to your desired value. How about punch card machines that are misprogrammed for precincts in which you think you candidate will loose so that when voters punch the hole for Kerry their vote is counted for Bush and vice versa. Where will you find material about how vote fraud this big could happen? You might start by tracking the links from Edgar Steele's post here <http://www.serendipity.li/jsmill/bushwon.htm>. (Look at the Devvy link and then at the Ronnie Dugger links in Devvy's post). Chuckherrin.com is good, too.

The fact that for the last 5 presidential elections EVERY national raw exit poll was skewed to Democrats could mean that the last five presidential elections were fraudulent. Don't know if you really want to hear this right now, but I was just over at Free Republic website a day or so ago reading a discussion about how election fraud might have impacted the 1996 race. <http://www.freerepublic.com/forum/a370cd2921b85.htm> Of course the Freepers thought that election fraud was how Clinton won the race. If there was fraud (and, yes, even then the voting machine companies were allied with Republicans) maybe it reduced the exit poll/vote tabulation disparity. Maybe the content of the article just means that we ought to get really, really clear in our minds that election fraud on this big a scale could be as likely as flawed exit polls.

Finally, exit poll data is not our best evidence -- the precinct-by-precinct vote analyses is our best source of evidence of election fraud. This sort of analysis is the back-bone of the Arnebeck law suit filed with the Ohio Supreme Court.

:bounce:











 

[gulogulo]

18. IndyOp, I hate to break it to you, but -

neither exit poll analysis nor precint-by-precinct vote analyses nor statistics nor comparisons between judge and presidential races will convince any court about election fraud - ask any unbiased lawer (which rules out Arnebeck). I asked. The response I got was: no, won't fly in court, get a solid proof, not circumstantial stuff that can be explained away.

[RaulVB]

20. "gulogulo," I hate to break it to you but...

Circumstantial evidence is all you need to win a case if is strong, clear and definitive.

No need "to ask lawyers."

Ask Scott Peterson.

[gulogulo]

23. "no need to ask lawyers"

that's where you're wrong.

[RaulVB]

29. And you are wrong here...

If you think that your very weak argumentation is going to "convince" people that the fraud didn't take place, you are making a mistake.

Try again.

[gulogulo]

34. If you'd be so kind as to point out the post in which I stated

that no fraud took place, I'd be obliged.

The fraud may have taken place. The "evidence" that is presented to prove that it has, though, would not stand in the court of law. It may be enough to convince you, Raul, but it needs to convince the judge in order for anything to happen.

[Helga Scow Stern]

32. gulogulo, you forgot to mention the "cover up":

the final damning nail in the coffin of the fraud-free election hypothesis is the energetic resistance to the actual results being verified.

In other words, if there was no fraud, why has there been such an effort to hide and obstruct in Ohio during the "recount"? Why not a transparent recount, according to law? Why refuse inspection of the polling books, violate random selection of precints? Why are officials breaking the law to keep their secrets?

If no fraud, what are they trying to hide?

[gulogulo]

35. I don't know, Ojai, could you explain to me

why Democrat Board of Election directors and members would strive to obstruct the recount together with the Republican ones? The "officials" that you are talking about are just as often Democrats as they are Republican. I know you think that the Republican officials are obstructing the recount in order to hide something, but what are the Democrats among the election officials trying to hide?

[Helga Scow Stern]

38. gulogulo, I am not sure to what extent the obstruction is by Dems....

as I have seen no documentation of party affiliation of obstructionists by the recount volunteers.

However, this proves nothing about fraud. Look at NM--there is quite likely fraud there, and obstruction is colluded with by a Dem governor who demands $1.4M at the last minute for the recount to proceed.

Go figure. I understand in fact, per The Nation, that Blackwell was actually once "a good Democrat". Just the fact that Dems may or may not be involved in no way absolves the charge of fraud.

To further elucidate this point, the tone towards election fraud taken by some liberal left pundits such as Drum and David Corn is quite derisive and almost hysterical in its rush to deny it as a hypothesis. This suggests a hidden perspective of at least some Democrats that is very hostile to the hypothesis of fraud. Of course this does not mean that they committed fraud, but it could certainly explain why they might-- consciously or not--collude in its concealment.

[gulogulo]

39. So, Ojai, would you automatically assume

with a Democratic pundit or official that if they are dismissive or obstructive towards fraud suggestions or recounts that they are "hiding something"? Would you automatically assume that a Republican pundit or offical, if dismissive or obstructive etc. is "hiding something"?

If your answers to the two questions are different, maybe it's time to re-examine your attitude. You're starting to judge people by their political affiliations, and not by their actions.

[Helga Scow Stern]

40. Either is possible in either case, of course.

However, there is quite a difference with Republicans. The burden of proof is upon them in the recount to prove that their results are fair. Breaking the law to keep honest investigation from happening is quite a bit more than being dismissive of the hypothesis, obviously.

And for all I know, some Dems may be directly involved in the fraud.

A crime has been committed, that much is evident. Who exactly perpetrated it where, when, and with whom and how is what is being investigated.

[davidlynch]

25. Gulogulo--Why High Standard Deviation in Battleground States Only?

I can understand that exit polls could be off by as much as 8% in one state, but do you have an explanation of how the standard deviation could be so much higher in *all* battle ground states? What could possibly explain this huge obvious correlation? The more significant the state to the election outcome, the more dramatic the swing. Are we saying that this is sheer coincidence?

[gulogulo]

31. david - I have no idea -

but riddle me this: if you (like TIA and Raul etc) believe that exit polls must be right and the election results must confirm them, please explain how come Nader's partial recount of NH, where the deviation between exit polls and election results was in double digits, and the recount precincts were hand picked by DU's own Ida Briggs in order to find the ones that showed the highest differences between 2000 and 2004, still no significant difference was found between the recount numbers and the election day counts?

Look at this article:

http://www.commondreams.org/headlines04/1118-03.htm

Recount New Hampshire
Just how accurate are optical scanners? A special recount may give an answer.

by Russ Baker


Tomorrow the first recount begins--in New Hampshire, of all places, a state George Bush didn't even win. But in those areas where he did well, sometimes the numbers look decidedly odd. In this case, the person who got the ball rolling was one Ida Briggs, a longtime Michigan software designer and database developer who did a statistical analysis of some election results, and found them perplexing enough to trigger concerns in her mind about the efficacy of the electronic vote tabulation system used.

...

Referring to the recount advocates, a Diebold spokesman told the Associated Press, "I think they're rushing to judgment."

...

If it turns out that anomalies are just that, so be it. Then we need to spend more time understanding why people voted--really, truly voted--the way they did.

-------------

Do you see that last sentence? Today we know that the recounts showed that the exit polls were bunk, the "anomalies" were "just that" - but funny how still no one is trying to "spend more time understanding".

[L. Coyote]

37. Exit Poll Discrepancy Correlation to Kerry Vote and Electoral Votes

Great Point!!

There's a pattern to the discrepancy. Forget the probabilities. The same method was used in each state, so look at the pattern. If they are all off a little, the pattern persists.

Only in "Blue States" is there a positive correlation with the number of electoral votes.
Only in "Blue States" is there a negative correlation with Kerry vote.

The greater the Kerry winning margin, the less correlation.
The higher the electoral vote prize, the higher the correlation.



Duh! Huh?

[davidlynch]

27. Regarding An E-Mail Exchange With Mitosfky

In an e-mail exchange with Mitofsky, he mentioned that he thought that exit polls were of no help in detecting fraud because the so-called raw numbers were already weighted, specifically additional weights forcing the results to election tallies in subareas of each state.

He's indeed planning to create a report, but this report will not focus on fraud, but rather will be an introspective analysis, a post-mortem, an analysis of the exit polls themselves (not of great value to our collective cause).

[mgr]

43. Drum's presentation misleading

This is the old apples and oranges thingee.

Drum cites the spread between Democrats and Republicans as what raw exit polls misrepresent. He does not demonstrate that raw exit polls overstate the support for democrats, only that they may under represent support for republicans.

A relevant question is the manner in which Mitofski (sp.)adjusted the exit polling to address these discrepancies, could potentially make comparisons between both raw and adjusted exits polls in the past and the present election inappropriate.

The one anomaly everyone cites is the Dukakis winning exit poll, what I want to know is whether this was within the MOE?

I am also puzzled as to why this is not extended back further in time to the 1970s, or is it only to the Bush/Dukakis election to make this point.

Mike

[gulogulo]

44. overstating support for democrats or underrepresenting support

for republicans leads to the same result - incorrect exit polls.

Dukakis winning exit poll - the discrepancy was more than 8%. How in the world could it possibly be within the MOE?

70s exit polls - don't know. I could not find that data. Drum probably couldn't either.

[jdog]

28. Thanks Ojai Person - - it does kick ass! n/t

[understandinglife]

2. Thank you for posting this. To motivate everyone to read this article...

....here are three quotes:

"I conclude that, based on the best exit sample data currently available, neither the national popular vote, or many of the certified state election results, are credible and should not be regarded as a true reflection of the intent of national electorate, or of many state voters, until a complete and thorough investigation of the possibilities a) and b) above is completed.

An election that is not “free” because of discriminatory suppression of the vote by not supplying an adequate number of voting machines, or by other means, could lead to an exit sampling discrepancy, as exit sampling is in part based on historic patterns of voter turnout. However the actual outcome in such an election would not be free or fair and would be less reflective of voter intent than exit sampling. Analysis of raw precinct level exit samples (which so far have not been released) should shed some light on where and when changes in the weights for raw precinct numbers, necessary to get state samples, were made.

This raises the more general question of what form of vote counting is more reliable. Vote counting that is overseen by a highly partisan Secretary of State with a clear vested interest in the outcome with election equipment that leaves no audible paper trail and/or central tabulating equipment that has been shown to easily hacked, or vote counting by exit sampling firms whose major vested interest is in getting the prediction right.13 These “unfair” elections in the U.S. mirror the situation in the Ukraine where one party controlled the collection and tabulation of the vote. At a minimum one would have thought that the oldest democracy in the world would implement its elections with unbiased civil servants and a uniform code of regulations. “Neutral” election
implementation was a key demand of the Ukrainian opposition."

Direct link to the .pdf:
http://www.freepress.org/images/departments/997.pdf

Urge all of you to read and disseminate widely.

Peace.

"I'm an American patriot, not a pro-fraud theocrat"

[Helga Scow Stern]

3. This Paper Kicks Ass!!! The title says it all:

The United States of Ukraine?: Exit Polls Leave Little Doubt that in a Free and Fair Election John Kerry Would Have Won both the Electoral College and the Popular Vote

[msongs]

6. would be nice to see this get attention somewhere besides DU eom

[MadLinguist]

41. i did post this to truemajority

and it is obviously available at freepress.org, but where else to send it? Spreading it individually to people that we know who are lukewarm on the topic of election fraud is tactically pretty good. What is best about this paper, to my way of thinking, is that it goes through the explanations for the disparity between exit poll data and official results and dispels them one by one, fat easy targets that they are. I'm glad to see that Baiman's work has gotten at least SOME audience.

[newyawker99]

45. Hi MadLinguist!!

Welcome to DU!! :toast:

[DELUSIONAL]

13. Kick

Good analysis of exit polling -- but still to dense for the average Joe blow American.

Bottom line -- Kerry won.

And the current polls on bush seem to reflect his real election numbers. The SOB lost.

[meganmonkey]

15. This is a thorough and understandable analysis of the exit samples

Thanks for posting, MadLinguist & Ojai Person!!!

[Kellis]

17. When I read stuff like this

It makes me wonder how anyone could call us conspiracy theorists,tinfoil hatters or dissenters.
The facts are right there for everyone to see....It happened in the Ukraine.It happened in the US.Period.
Now lets do something about it.



Thanks for posting-getting this out to everyone.

[seaclyr]

19. kick n/t

[Melissa G]

22. Kickin Truth to the top! n/t

[demo dutch]

42. Kick