FEBBLE: WILL YOU COMMENT ON THIS MYSTERY?

[TruthIsAll]

In the Final National Exit Poll (13660 respondents, posted at
1:25pm on 11/3/04), the Bush 2000 voter weighting was 43% of
the total 2004 electorate (122.26 million). Gore voters were
37% of the total.

1. How could this weighting have risen to 43% in the Final
13660, with only 613 additional respondents added, from the
13047 respondents at 12:22am when it stood at 41%?

2. How could 43% (52.57) million voters even be possible,
since only 50.456 million voted for Bush in 2000 - and around
1.75 million of them died? That would lower the weighting to
48.7/122.26 = 39.8%, a difference of almost 4 million votes.

3. How do you explain this? Was the 43% a sampled result, or
was it an arbitrary weighting? 

4. How does this jibe with RBR?

5. Finally, how could Kerry be winner by 4.63 million votes
(13047 respondents) at 12:22am and a loser by 3 million
(13660) at 1:25pm? 


National Exit Poll Timeline:
11/2/04, 3:59pm 8349 respondents: Kerry 51-Bush 47
http://www.exitpollz.org/mitof4zone/US2004G_3737_PRES04_NONE_H_Data.pdf

11/2/04, 7:33pm 11027 respondents: Kerry 50-Bush 47
http://www.exitpollz.org/mitof4zone/US2004G_3798_PRES04_NONE_H_Data.pdf

11/3/04, 12:22am 13047 respondents: Kerry 51-Bush 48
http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=265121

11/3/04, 1:25pm 13660 respondents: Kerry 48-Bush 51
http://www.exitpollz.org/mitof4zone/US2004G_3970_PRES04_NONE_H_Data.pdf



NATIONAL EXIT POLL TIMELINE				
VOTED IN 2000 FOR				

3:59pm 8349 
		KERRY	BUSH	NADER
No	15%	62%	37%	1%
Gore	39%	91%	8%	0%
Bush	42%	9%	90%	0%
Other	4%	61%	12%	16%
Total	100%	51.01%	46.95%	0.79%
Votes 	122.26	62.36	57.40	0.97
				

7:33pm 11027 				
No	17%	59%	39%	1%
Gore	38%	91%	8%	1%
Bush	41%	9%	90%	0%
Other	3%	65%	13%	16%
Total	100%	50.25%	46.96%	1.03%
Votes 	122.26	61.44	57.41	1.26


12:28am 13047				
No	17%	57%	41%	2%
Gore	39%	91%	8%	1%
Bush	41%	10%	90%	0%
Other	3%	71%	21%	8%
Total	100%	51.41%	47.62%	0.97%
Votes 	122.26	62.85	58.22	1.19

				
1:25pm 13660				
No	17%	54%	45%	1%
Gore	37%	90%	10%	0%
Bush	43%	9%	91%	0%
Other	3%	71%	21%	8%
Total	100%	48.48%	51.11%	0.41%
Votes 	122.26	59.27	62.49	0.50
				
				

[Botany]

1. TIA?

So what I am supposed to be looking @?

[TruthIsAll]

2. These are the Kerry/Bush National Exit Poll results from the

first 8349 respondents to the Final 13660.

Kerry led by 51-48% up through the first 13047 exit poll responders.
I show the percentages and equivalent vote totals (in millions).

[kster]

3. kick.n/t

[Sancho]

4. The more I look at this...the more I'm convinced E-M knows..

When you look at the questions on the surveys, it seems that with the stack of original poll papers it would be possible to take all the 43% 2000 Bush voters or any other sub group and test how likely it was that they profiled Bush voters.

I suspect that binomial tests and correlations would quickly precincts where there was fraud. It would likely be clear enough that E-M must be able to rule out RBR's.

[kster]

5. Kster taps desk while waiting for FEBBLE... kick......n/t

[Febble]

6. Responses:

1. How could this weighting have risen to 43% in the Final
13660, with only 613 additional respondents added, from the
13047 respondents at 12:22am when it stood at 41%?

Because new weights were used. You have proved this, and so have I. The initial weights did not match the incoming results, so the respondents were reweighted. This could have been because of fraud. it was why I became interested in finding out.


2. How could 43% (52.57) million voters even be possible,
since only 50.456 million voted for Bush in 2000 - and around
1.75 million of them died? That would lower the weighting to
48.7/122.26 = 39.8%, a difference of almost 4 million votes.

This is a good question with several possible answers, one of which is that the count was fraudulent. Another is that retrospective questions generate notoriously unreliable answers.

3. How do you explain this? Was the 43% a sampled result, or
was it an arbitrary weighting?

I am sure the weighting was not arbitrary. It would reflect incoming information, probably including both turnout data and count data. As you point out, the problem may be that the count data was wrong.


4. How does this jibe with RBR?

rBr might account for the data. So might fictitious votes. The numbers do not tell us which. All I have shown is that whichever it was, it was probably not, as USCV infer, more prevalent in high Bush precincts.


5. Finally, how could Kerry be winner by 4.63 million votes
(13047 respondents) at 12:22am and a loser by 3 million
(13660) at 1:25pm?

I would love to know. I hope that fact that I have cleaned up a key variable will help us find out.

Now, will you please STOP this harassment, because frankly that is what it is.

[TruthIsAll]

7. Good give-and-take is NOT harassment. It's an attempt to focus on FACTS...

To even call it harassment is being grossly unfair, not only to me, but to all DUers.

We welcomed you in good faith.
We heard your story.

We asked you some very pointed questions, direct, factual, numerical.
That is not harassment.

We asked you to review ALL of our analysis.
That is not harassment.

We expressed our views regarding your hypothesis.
That is not harassment.

We expressed our views on the scientific method.
That is not harassment.

We have put the facts before you for comment.
That is not harassment.

Now we have your views on record.
That is what it was all about.


In closing, let's dispel the following myths from those naysayers who since the election continue to claim that exit polls are:

1) unsophisticated and unreliable,
2) not based on true random samples,

**********************************
3) biased by reluctant responders
**********************************

4) do not take historical trends into account.

There should be no Mystery regarding Exit Polls. And believe it or not, the Mystery Pollster and I are in complete agreement:

Late in the day exit polls, like the 12:22am National Exit Poll of 13047 respondents and the various state exit polls (both had Kerry winning by 51-48%) are scientific, accurate and reliable.

http://www.mysterypollster.com/main/2004/11/exit_polls_...

EXIT POLLS: WHAT YOU SHOULD KNOW

snip

SOPHISTICATED AND RELIABLE
I have always been a fan of exit polls. Despite the occasional controversies, exit polls remain among the most sophisticated and reliable political surveys available. They will offer an unparalleled look at today's voters in a way that would be impossible without quality survey data. Having said that, they are still just random sample surveys, possessing the usual limitations plus some that are unique to exit polling (I also remain dubious about weighting telephone surveys to match them, but that is another story for another day).


A RANDOM SAMPLING OF 1495 PRECINCTS
A quick summary of how exit polls work: The exit pollster begins by drawing a random sampling of precincts within a state, selected so that the odds of any precinct being selected are proportionate to the number that typically vote in that precinct. The National Election Pool Exit Poll, which is conducting the exit polling for the six major networks today, will send exit pollsters to 1,495 precincts across the country.

A RANDOM SELECTION OF 100 VOTERS PER PRECINCT
One or sometimes two interviewers will report to each sampled precinct. They will stand outside and attempt to randomly select roughly 100 voters during the day as they exit from voting. The interviewer will accomplish this task by counting voters as they leave the polling place and selecting every voter at a specific interval (every 10th or 20th voter, for example). The interval is chosen so that approximately 100 interviews will be spread evenly over the course of the day.


******************************************************
RELUCTANT RESPONDERS: STATISTICAL CORRECTIONS FOR ANY BIAS
When a voter refuses to participate, the interviewer records their gender, race and approximate age. This data allow the exit pollsters to do statistical corrections for any bias in gender, race and age that might result from refusals.
************************************************************
snip

WEIGHTINGS BASED ON HISTORICAL DATA AND VOTER TURNOUT
One of the unique aspects of the exit poll design is the way it gradually incorporates real turnout and vote data as it becomes available once the polls close. The exit poll designers have developed weighting schemes and algorithms to allow all sorts of comparisons to historical data that supports the networks as they decide whether to "call” a state for a particular candidate. When all of the votes have been counted, the exit poll is weighted by the vote to match the actual result.

snip

[Febble]

8. This isn't harassment?

...telling us how to analyze our fucking election? I'm sick of this Febble fixation. What's the matter with our press and our users looking to AMERICAN analysts and statisticians.

I for one say in the analysis of election fraud, AMERICAN'S FIRST.

Febble, we're happy you have time to do all this. Now go back to what you were doing. As for the dilettantes and obscurants fixated on her work, give it up or admit you're afraid to look at the ugly truth head on. THE ELECTION WA"S STOLEN, WHERE'S THE OUTRAGE?"

Accusing me of a biased premise isn't harassment? Drumming fingers waiting for a response isn't harassment? Bombarding me with questions all over multiple threads, and then writing snarky comments when I don't address them all at once isn't harassment?

On this board, my nationality, my professional credibility, my honesty, my ethics, my motivation, and my political positions, have all been called into question, and my attempts to defend myself have been met with mockery.

I call that harassment.

And nobody, of those who have been indulging in this kind of behaviour, appears to have bothered to check on the USCV website to see the work I have been doing to try to uncover the appalling travesties of democracies that undoubtedly occurred in this election.

I think some people around need to do a little self-reflection.

[mgr]

14. My personal apology

I cannot speak for others on DU, and since I was never welcomed in the first place, I am probably not a member of community, although I am the contrarian statistician (I'm trained as a biologist and geographer) It would be my wish that this were a discussion, not a debate, nor a xenophobic diatribe against the Scots, as you may be our best direct window on USCV.

Please correct me if I have it wrong, my understanding has been that with NEP's claim that bias affected the characterization of the exit poll to show greater support of Kerry than may have been present, there exists potential to rehabilitate the exit polls and remove that bias. The manner in which I saw this issue being addressed was, since WPE represented a 'goodness of fit' test, would be to remove those precincts with high bias rates through an ANOVA, and recalculate the mean and MOE with the residual. What I see your position as being that even precincts with low WPE may be suspect to bias, and that a further correction or adjustment may be necessary before proceeding.

Another concern that I have is with the pattern of bias itself (being a geographer). I noted that there is apparent effort on the part of the Ohio SOS to limit the effectiveness of the exit polling in Ohio (NEP, 1/19/05, p.70?), would this potentially allow for less bias within precincts in other states?

Thank you for your time and attention.

Mike

[Febble]

15. Thanks Mike.

I probably can't address your questions though as I am in no way a polling expert, and do not know much about how the projections are made.

However, I am fairly sure that the WPE is simply a quality control measure - it isn't used to make the projections. Although voters are randomly sampled from precincts, and precincts sampled from states, voters are nonetheless treated as though randomly selected from states, and a factor is applied to the standard error in order to allow for the "clustered sampling" (the notorious "design effect"), otherwise, of course the variance will be too low, and the MoE too small.

The problem I identified is that the WPE is a very poor dependent measure for quality control because of its confound with margin. I have devised a transform (actually simple algebra) to convert the WPE into a measure that is simply the log of the numerator of the ratio between the response rate for one group of voters and the response rate of the other. So it is independent of vote margin, and also independent of overall response rates. This would seem to be a far better variable on which to regress factors that are hypothesised to be associated with variance in bias (bias, to make it clear, that can be equally bias in the polls or bias in the count - if the ratio between the response rates is not equal to 1 it could be because either votes didn't exist or voters were phantoms - or because one group of voters was more ready to respond than the other).

Anyway back to the point - if my transform is applied, there should be no confound, and the factors associated with bias more validly identified using the GLM.

[mgr]

16. I'm beginning to feel that gulf

Between statistical analysis as a gray box when compared to a white box. I've accumulated much rust, and what I know may only be enough to make me dangerous. I had an inkling of what you so clearly expressed, is that WPE cannot function as a goodness of fit because you cannot fix the boundaries between the static variable from the dynamic, but as is often the difference between the dilettante and the artist, I thought that it did not matter.

I will still plod on, at least you haven't cornered the concession on fruitful hypotheses, and I think there is a spatial aspect to the bias that needs looking further into. As to your variable, is the peer review still mid May? Keep us posted.

Mike

[tommcintyre]

18. "Accusing me of a biased premise isn't harassment?" ABSOLUTELY NOT!

That's called healthy questioning. History is replete with examples of "biased premises" that needed/deserved to be questioned.

Example: Was the earth REALLY the center of the universe? Well... it sure looked like it to the "powers-that-be" at the time (and they wanted it to be true). And Copernicus, Galileo, etc. paid dearly for "accusing <THEM> of a biased premise", didn't they? Hmm... I wonder if the Inquisition accused Copernicus or Galileo of "harassment" before they lit the match under the stake, or locked him up, respectively? ;)

YOU, as a student of psychology, should realize this better than others. The challenge of objectivity vs. subjectivity is ALWAYS with us, isn't it? That's why science prefers the double-blind study* whenever possible - to eliminate (in reality only further minimize) the "contamination" of an experiment with bias - even when it is totally unintended.

Since you are working only with inferred, not direct data, your postulations are rife with the possibilities of bias contamination. If you contend that the certainty of numbers protect you from this, you are only deluding yourself. After all, you are choosing the framing of the questions, etc. - so, YES, your premise(s) could be EASILY biased.

Question: Can YOU see any way(s) you could've introduced bias into your paper?

I agree. At times, there has certainly been less than courteous and/or fair treatment of you here. But the fact that you tried to "slip" this "biased premise" claim into the middle of your complaint should be a wake-up call to you.

Yes Febble, you (and your work) are just as subject to bias (subjectivity) as the rest of us.
--------------
* Double-blind study: A testing procedure, designed to eliminate biased results, in which the identity of those receiving a test treatment is concealed from both administrators and subjects until after the study is completed.

Double-blind describes an especially stringent way of conducting an experiment, usually on human subjects, in such as way as to attempt to eliminate subjective bias on the part of either the experimenters or experimental subjects.
http://www.answers.com/topic/double-blind

-------------------------------------------
<In case you missed it, I replied to your request that I read you paper here (some of the main ideas are listed below:>
http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=365014&mesg_id=365835

"Taking EVERYTHING into consideration (the total "environment" in which this paper exists), I am leaning toward the concept of "red herring"...".

"The fact of the matter is you can only infer the validity of this hypothesis - you can NEVER verify it."

"Since you can not EVER verify the rBr (since that time has past, and can never be recovered), the very idea is rendered moot."

[TruthIsAll]

9. Thank you, Febble. I see that we are very close to being in agreement...

I appreciate your response.

Let's summarize (please correct me if I am misstating your views):

1. You believe it was either fraud or RBR or a combination of both.
2. I believe it was fraud.

If RbR is shown to be a non-starter, then we both agree:
It was Fraud.

[Febble]

10. Appreciated.

But we need to sort out some terminology. "rBr" was used by USCV to refer to the hypthesis that uniform "reluctant Bush responders" across all categories of vote margin accounted for the poll-count discrepancy. They claimed that it was refuted on the grounds that if the degree of reluctance was uniform, response rates should have been lower in high Bush precincts and higher in low Bush precincts. In fact the mean response rates in high Bush precincts was slightly higher, than in low Bush precincts, but not significantly so.

I do not agree that this constitutes a refutation of the uniform rBr hypothesis, however, as we do not know the variance of the response rates, and the effect size predicted by the null hypothesis (uniformity) is small. But leave that for the moment. USCV then tested (inasmuch as the meagre information in the E-M report allows us to test) the alternative hypothesis that Bush voters were "shyer" in high Kerry precincts. They claim that the E-M data refutes this, as the mean WPEs are actually more negative in the high Bush precincts.

On both counts I believe that their refutations of the "uniform rBr" are not justified, and are based on a misreading of the data, for the reasons given in my paper.

Which means that "uniform rBr" is unrefuted. However "uniform rBr" is indistinguishable, mathematically from "uniform vote count corruption" if the form of corruption is vote swapping, ballot stuffing, vote spoilage, or electronic versions of any of them.

It is the non-uniformity of the disrepancy I have challenged, or at least the non-uniformity relative to vote-count margin in the precincts, not whether the discrepancy was due to rBr of vote-count corruption, although for quite other reasons, given elsewhere on DU, I think there is evidence that does point to the possibility that rBr was at least a partial contributor to the discrepancy.

What we need is to test are real fraud hypotheses: where would we expect the discrepancy to be greatest if the cause of the discrepancy was fraud? Swing states? Ohio? Those are the hypotheses we need to test. We need to ask: what would fraud look like? What would rBr look like? If my variable is used, the test, either way, will be more convincing, as it is not confounded, as the WPE is confounded, with vote margin itself. It measures bias, and does not distinguish between bias in the poll and bias in the count.

I would also hypothesise that if the discrepancy has anything to do with the rBr, it will be larger where protocol was poor, and more Bush voters were able to avoid being polled. This hypothesis is partially supported by the E-M report, but the report does not say how much variance in bias was accounted for by these factors, and as I said, I also believe that their measure of bias is a poor one. So I think the jury is still out on rBr.

Most of the misunderstanding on this site has been of the fact that what I did was question a particular refutation of rBr. rBr may be refuted in other ways, but I don't think this refutation is valid. I maintain that uniform rBr is consistent with the data, just as uniform vote corruption is consistent with the data, apart from the proviso noted above. And when I say uniform, I mean uniform over all degrees of precinct partisanship. It clearly has a wide variance. It will by analysing this variance that rBr will be either refuted or supported.

But we need specific hypotheses to test. Sadly, though, it is in the nature of statistics that failure to find an effect does not mean it is not there, especially if the data is noisy. This is very noisy data.

[TruthIsAll]

11. Can you comment on this from MP (above)?

My caps:
RELUCTANT RESPONDERS: STATISTICAL CORRECTIONS FOR ANY BIAS

"When a voter refuses to participate, the interviewer records their gender, race and approximate age. This data allow the exit pollsters to do statistical corrections for any bias in gender, race and age that might result from refusals".

Were these adjustments not made in 2004?

[TruthIsAll]

29. Febble, still waiting for a response on the refuseniks....

Did the exit pollsters correct for the refusals?
If they did, why is RbR even an issue?
If they didn't, why not?

[FogerRox]

12. Appreciate you staying with us here @DU

At times your word choice has made me cringe.

Your mention of Touchscreen precincts in Ohio. made me really pissed. I felt like I was being talked down to--

Ohio IIRC has 4 DRE county's.
44 punchcard
40 Opscan

Ohio votes not by precinct but by Municipality IIRC.

So Febble I felt like you were throwing me a bone (and others) by mentioning touchscreens in Ohio-- there may have been some here @ DU, including myself, who would consider that sort of comment as indicating lack of knowledge.

And I understand that the use of the word "precincts" when refering to Ohio, can be a sort of generic use.

So if you felt like you were the silver ball in a pinball machine--

LOL--- welcome to DU

[Febble]

13. DREs

in Franklin County Ohio were what were scandalously rationed by the BoE. You might want to check out this:

http://uscountvotes.org/index.php?option=com_content&task=view&id=65&Itemid=63

But you are right, they were pushbutton not touchscreen, I have just checked. I had been misinformed about this, because I did enquire, the reason being that when I was analysing the data there were votes for someone called "removed". I enquired about this, and was told that it was Nader, who had been "removed" from the ballot, but that the DREs had not been reprogrammed so the touch screen area was still live, and that in some precincts a sticker had been placed over it. A a small number of votes were nonetheless cast for "removed". It appears that my informant was not correct. I assume the stickers he mentioned must have been placed over buttons, not a part of the screen.

I wasn't talking down to anyone, and it was the DREs in Franklin County Ohio that I was referring to. But yes, it appears they were button machines, not touchscreens. Sorry.

I realise that word choice is a sensitive issue on DU - to the extent that even a completely technical statistical term like "error" has everyone reaching for their tinfoil hats.

And the unit of analysis in the E-M report is precincts. Precinct level data is what we are dealing with.

But, thanks for the olive branch.

[emcguffie]

17. I'm no statistician or scientist

And I do not understand all of this. But can someone try patiently to explain to me why it would be likely for a human trait (is that the word?), such as willingness to respond to an exit poll, to be consistently more/less common in a conservative versus a liberal voter? (I didn't express that very well, sorry.) In very different environments? Such as, say, the south versus the north, where attitudes and cultural differences can be generalized in a very loose way. I mean, people hurry more in the northeast, and are more "apparently" friendly in the south. Gee, what I am trying to say is that, with so many other human factors that make people so different, like millions of genetic combinations, why would one factor such as who you are voting for correlate to your willingness to participate in an exit poll? Especially when you factor in that many of these voters may have voted a different way in another election? Would becoming a * voter make you suddenly more shy? Why not more in-your-face? In my experience, those guys on the other side are much more in-your-face than thoughtful lib types.

I just don't get it.

What I do get is that lower income people are much more likely to be in too much of a hurry to stop and fill out a form. People who have to get back to work or pick up children at school or daycare, who can't afford to pay a dollar a minute late penalty. People with bigger incomes, such as folks who are more likely to be executives and managers and bosses, can afford babysitters more easily and can take time to vote as they wish. Or they can afford to pay a big penalty for being late to pick their child up.

What is the quality that would make people who voted for Bush unwilling to respond to exit polls, but people who voted for Kerry more willing?

This is not a facetious question. I am absolutely sincere.

Thank you to anyone who will take the time -- and please forgive my inability to articulate very clearly.
E

[TruthIsAll]

19. You are using common sense. RBR is uncommon nonsense.

/

[tommcintyre]

20. "uncommon nonsense"? That's a pretty good way of summing it up ;)

My only qualifier is that that the nonsense has gotten all to common since Bushco took over.

What did Orwell call it? "New speak"? During the Viet Nam "nonsense", this now infamous quote epitomizes the absurdity that is alive and not-too-well today:

"We had to destroy the village to save it."
"Destroying Iraq to Save It"
http://www.commondreams.org/views04/1121-21.htm

So, are the neocons trying to destroy our Democracy, to save it (from the American people)?

"They" that would (inadvertently or not), help them in their goal - come and go. But, WE are still here.

And we're NOT going away until this wrong is made right.

Drip, drip, drip

[FogerRox]

21. Additionally

My understanding of exit polling (limited -by far) Says its accurate.

But--if you have a white male running against a black female.
there should be some bias. A white male may not admit he voted for the black female.

When you have 2 white guys running--Bush vs Kerry-my understanding is that these sorts of bias -- or reluctance--- are basically eliminated.

Febble -- I wasnt aware of any pushbutton systems in Ohio--news to me. Though I have not looked at it hard and have relied on others for research---- These machines are they like --25 - 30 years old?

Since ohio spent a lot of money in the last 6 years on new equipment.
I am surprised.

[Febble]

22. Info in Ohio voting machines here

Apparently it was "touchpad". I wasn't far out.

http://www.buzzflash.com/contributors/04/11/con04517.html

And incidentally, exit polls have have overstated the Democratic vote in every year since 1988, always significantly. In 1992 the discrepancy was nearly as large as in 2004, and actually more "significant".

Interpret that as you wish!

[TruthIsAll]

23. Febble, you can throw that canard out of your talking-point toolbox.

When Clinton ran there were three candidates. Are you aware of that?Perot got 19%.

And there were no touchscreens, so a lot of democratic spoilage was from punched cards. Now the central tabulators, toucshcreens and optiscans do the work.

In any case the deviation trend has been down - until it reversed in 2004. I have the graphs to prove it.

Question for you:
How come the Republican deviations always exceed those of the Dems?
Skimming third party votes?


http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=203&topic_id=341940

[FogerRox]

25. buzzflash article

at this link--

http://www.buzzflash.com/contributors/04/11/con04517.html

5. Cuyahoga and Summit Counties use punch cards, as do most of the counties in Ohio. Franklin and Mahoning use paperless touch screen machines (made by two different manufacturers). Mahoning's is a true computer touch screen system, while Franklin's machines are an older touch pad type that is described in detail on the Franklin County Board of Elections web site. Mahoning was switched over to a touch screen voting system only recently. Both Franklin's and Mahoning's machines are considered to be DRE (Direct Recording Electronic) machines. They are "paperless" because they do not produce any paper trail that can be used for recounts. There is no TANGIBLE way to verify that votes were recorded and counted properly.

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

IIRc the difference is touchscreens are galvanic response--while pads are pressure sensitive

[emcguffie]

26. can you explain the theory to me?

Of what would cause some consistent -- over-sampling of kerry voters? or the opposite?

There must be some kind of theoretical explanation, if people are going to keep saying it as a possibility. Some reason. Something.

[Chi]

27. I'll take a stab at this....

Since both Kerry and Bush supporters were not really inclined to respond, we have to assume whatever has affected the response rates is suppressive.
Bush Response rate 50%
Kerry Response rate 56%
Since Bush responders were less inclined, we must conclude they are more effected by the negative stimuli.
The likely factor would be network/cable news.
I believe that subliminal messages broadcast by network/cable news affected Exit poll response.
When someone would decide they wanted Bush to be President again (easy to believe, between the record deficits, special attention to the rich, and the body counts), that an internal trigger would occur that made them inherently less likely to answer questions from an exit pollster, but somehow makes them much more vocal at parties, and always reasonable to talk to intelligently.
This affect seems to cross all U.S. political borders and cultures throughout the entire country.

Sorry, I've been asking myself that for the last week, and I couldn't resist 8)

:freak: :freak: :freak: :freak: :freak: :freak:

[emcguffie]

28. Thanks for your answer, and excuse me for not quitting yet --

But is it reasonable to work backwards like that? To say, well, obviously these guys were less likely than those guys (which isn't necessarily obvious and isn't necessarily true), and then make up a reason for it?

Excuse me for being dense there, because even what you suggest cannot be taken seriously. How could that could apply all across the country and across all types of Bush voters? That to me is utterly entirely absurd. Like expecting all women to look good in the same lipstick, or the same clothes. Which does seem to be expected by some companies, but which turns out to be not true. We are to accept that a certain percentage of voters are voting for Bush but feel badly about their choice, and thus unwilling to 'fess up? Like a five-year-old that broke something and says he "can't remember" what happened?

Can we say * supporters like coke more than pepsi or the other way around? Rednecks or liberal elites are more likely to do this or that?

That's all just as silly as the red state blue state business. Character is much too complex to break down into two columns, kerry voter versus bush voter. And to character you have to add circumstances, which brings in the poor voter versus rich voter, and who has time and who doesn't.

Excuse me for going on and on, but I don't see a plausible argument for this reluctant responder business anywhere. From a human behavior point of view. But then, what do I know? Obviously not the same things that * voters do.

On top of which, isn't the whole basis for polling that one error one way will be cancelled out by an opposite error elsewhere? I admit I don't know what I'm talking about, but isn't that why polling supposedly works?

[Chi]

30. Not a problem....

But I was joking.
I was playing Bill O'Reilly, and making stuff up.

I too wonder how deciding on a particular candidate makes you more prone, or less prone, to answer poll questions...everywhere.

I think your question is valid, but I don't think you'll find anyone to answer it.

Good Luck

[emcguffie]

37. glad to know you were fooling

Since I know nothing about the field, hey, you could tell me anything. But I might not accept it.

I think people who statisticians or mathematicians get so caught up in whether this is evidence or proof or is valid or not that they cannot see the forest for the trees.

Thanks for the answer.

[Chi]

39. Your best bet for an answer....

would probably be Mike (MGR), that's not a blind endorsement of his conclusions though. ;-)

Febble would be best equipped, IMO, but she doesn't seem to want to touch on anything outside the scope of her paper.

[mgr]

42. Your subliminal argument might have merit

If the intense partisanship between the parties suggest that individuals of one party's persuasion avoid interaction with individuals of the other party, then anything that might bring an interaction would be avoided. The thing that shocked me (and woke me up) was the intense and hateful rhetoric of day time radio about three years ago. I don't know if we have the same rhetoric on our side, so there may not be as strong a sense of animus to wards republicans by democrats, as there is by republicans to democrats. So if republicans view a poll taker as possibly liberal, this animus that has been ingrained into them, may be the factor.

Right now its speculation as to why, but that a pattern exists; and that apparent pattern serves to obscure any means of addressing fraud through the exit poll. What Febble argues hurts my agenda far more than TIA's.

Mike

[Chi]

43. Actually...it was humorous, or so I thought

Being emcguffie didn't catch on to it, I guess I was too subtle.

I'll take it you actually are not giving credence to the subliminal, but to an unspecified source of conscience influence.
In my mind the repeat problem for that scenario will always be, what such influence could so consistently affect the populous.
If I pick one, I can always find an area where that influence should show a peak, or valley when matched to the filtered responder results.
And I don't think that would happen.

If you think talk (hate) radio was bad 3 years ago, you should of heard it under Clinton. I used to listen to Rush, just to know what my father-in-law would say next time I saw him.
It was a deliberate pattern of deception, almost evil.

To me (& emcguffie by the looks of it) this is an hypothesis that makes no sense, and which no one tries to actually answer the question emcguffie asked about what could cause this 'trait' among the (Bush voting) populous.

(shrug) But WTF do I know.

[davidgmills]

32. You think was subliminal broadcast messages

I thought it was Jupiter aligning with Mars.

What did I know?

I thought it was the continuing of the age of Nefarious, Nefarious.

[mgr]

41. Actually, that is a very good question

What has been done in this argument has been reductionism to behaviors based on party identification. Your point is that there may be cultural or class differences that may overwhelm the party identification component of refusals, and you may well be right.

There is often a bit of truth to regional stereotypes, like southern hospitality, and such; but with the recent migrations from the east and mid west to elsewhere in the country these may no longer be valid regional characterizations. With the homogenization a lot of these cultural or class differences break down, after all we now have white guy who is thought to be the best rapper (remember Vanilla Ice, how far we have come).

With the exit poll, we have a preliminary statement after the election
by Mitofski that he thought the poll was strongly biased to wards Kerry due to non-response, with possibly little analysis of the data at the time.

We can take this as hypothesis or self fulfilling prophecy when we review the January 19 report. But what it is, is that somehow, due to some characteristic of the poll takers, it resulted in a pattern that suggests more republicans may have been non responsive than democrats. It may not be a big difference, but when the ability to resolve a dynamic pattern (movement to Kerry)with the measure of a static pattern (party loyalty), the dynamic (and the reason we are polling in the first place)gets lost in it, it can hurt.
But it is a pattern that probably should not be reified into a characteristic.

The thing is to realize that at the precinct level this is a very small error--all you have to do is get two too many Kerry supporters to respond to the pollster to exceed the expected response based on party registration--this can give you a WPE of 5 to 8% depending on the ratio of D:R.

Mike

Mike

[FogerRox]

24. I did a search for 3 words on that PDF--

touchscreen
DRE
touchpad

I got zero results

this pdf--
http://uscountvotes.org/ucvAnalysis/OH/FranklinCountyReport_v2.pdf

[Febble]

31. Well I know they were DREs

I had thought they were touchscreen. Turns out they were touchpad.

But the point of the paper was that they were scandalously rationed, resulting in a major loss to Kerry.

[FogerRox]

33. understand that you have waded into a forum where some very

knowledgeable Peeps reside.
Accuracy is respected, as is being vague--if you arent sure-- say so

that SHOULD be respected.

I remember some one going around DU and correcting any post on voter fraud---its ELECTION fraud--was the mantra.

The allocation of Equipment you refer to was based on the 2000 & 2002
Voter turnout in Ohio--thats a violation of Federal statute. YOu must use the Current voter reg roles.

"scandalously rationed" no, a violation of federal statutes. Which is a criminal act. Relating "scandal" or "ration" to criminal act in this arena may be considered diluted language--see "weak knee'd".
So by using "scandalously rationed" you set a tone.

By mentioning "Touchscreen precincts" you set a tone

---ohio votes by Municipality IIRC and 4 out of 88 county's used DRE's. So the tone you set "might be" one of---------

"this gal dont know shit about ohio". And that will only serve to deride your efforts.

I grew up with my family in the academic world--My father served as Advisor to many a doctorial candidate--- these kinds of mis-steps may have brought about a dressing down for many a PHD candidate-
(in some programs)

[Febble]

34. Please read the paper

It contains evidence that may interest you.

If you are right that allocating voting machines on the basis of voter turnout is a violation of Federal statute, then my paper contains vital statistical evidence of the violation.

But my field is statistics, not Federal law. For this reason I forwarded it to the Kerry Edwards counsel in Ohio, and also to Cliff Arnebeck.

Thankyou for pointing out that I used the word "touchscreen" instead of "touchpad" above. I stand corrected. However, whether the machines were touchscreen or touchpad is quite immaterial to the evidence, and I do not refer to machine type in the paper.

If this is an area that interests you, please read the paper. I am concerned that no legal action appears to have been taken over the evidence it presents. This is one reason I posted the link here.

Perhaps someone could pursue this.

[FogerRox]

35. if you went to college -got a dgree then you should to be more carefull

Dont BS about something you are not sure about.

And touchpads may be 20 to 30 years old soo yes that is of import----

DOH

Sorry Feb but if your area of academic expertise is stats then use that and go with it---I think you need to study up a bit on the general area of VOting equipment, election law --oe dont make the suppositions you have--

I have no college degree but I have learned a lot since nov 3rd.

Which suggests that you are capable of the same. Without the general understanding you limit your self-your work and how peeps may interpret your work.

Maybe i'll read your paper--I have 3 lawyers working with my election reform group--NJVoting Issues Project [email protected]
And i'm a founding member of Voters for Open and Transparnet Election
vote.org
as well as the NJ coordinator of 51capitalmarch.com
As NJVIP coordinates a state wide series of County petition drives to have a referendum banning DRE's 1 county at a time.

things I look forward to:
Suing Seqouia
Filing for injunctions to void-delay existing contracts to deliver more DRE"s into my state
Educational forums
RAllies- protests

When I read your paper I hope it atriculates your position well. Otherwise you will have wasted my time. I fear from the missteps here at DU that may be so

[autorank]

44. Amen!

[FogerRox]

36. yeah yeah yeah--that link is to a pdf doc

that PDF doesn't contain those 3 words so why did you post it?

press CTRL & "F" to use the search feater on a PDF--there is no friggin mention of what you said was there. SO I'm callin tour bluff.

If you cant post an article that pertains to your post then WTF are you doing?

[Febble]

40. The link is to a page on the USCV website

It has the first few paragraphs of my paper. If you click on the continue link you will get a link to a pdf of the whole paper.

The paper is reports an investigation into the way voting machines were allocated in Franklin county Ohio.

I actually don't know what you are talking about. You seemed at one point to object to a post in which I had used the word touchscreen to describe voting machines in Ohio. I presumed you were talking about this paper. You were correct, apparently they were touchpad.

As the paper is not about the type of machine, then the document probably does not contain either word.

If you are not interested in the paper don't read it. I understood you were interested, so I posted the link.

But it seems all you are interested in is abusing me.

[Peace Patriot]

38. I want to reintroduce my rrrK theory: Reluctant Republican Respondents...

...who voted for Kerry. This election, and the political and social climate in this country, are quite unique in my memory (and I go way back). I cannot remember anything similar except when I was a child and the Army-McCarthy hearings were on our grainy little black and white TV screens. Peoples' lives being ruined for being "pinkos" or "fellow travelers." I grew up in an area where the Christian Anti-Communist Crusade held big tent gatherings, and where the John Birch Society was founded. These were truly scary people--not much better than the KKK.

These people are back--only they are now up to far worse things than McCarthyite witch-hunts. And they are now running our national and many of our state governments, and the libs and Dems in government are scared of them. I'm scared of them. If I were living in a Republican county now, I would most certainly be reluctant to admit a Kerry vote to a strange pollster, who approached me after I voted. For one thing, I wouldn't be at all sure who this person really was. I now live in a very Dem area, and guess what? I never did put a Kerry sticker on my car--for fear of tire slashing and other vandalism by Bushite thugs. In a Dem area!

We are being ruled by thugs. Real scary people. Now imagine yourself, say, as a lifelong Republican. You've decided Bush is nuts. (I know some elderly Republicans who said just that--"Bush is nuts.") You've decided to vote for Kerry. You live in a white, Christian Anti-Communist Crusade type of county. You've never voted for a Democrat before in your life. But you can't take this war and this deficit any more. These are not conservative policies. Being a Republican, you don't talk much about politics with your friends and family, so you don't know much about how others are feeling. And the only people who speak up are loud Bushites. So you go ahead and cast your vote for Kerry, and you may not even have told your husband that you were going to. And somebody comes toward you, out of the blue, and asks you to state publicly how you voted. And either you just walk away--because it's none of their business--or you say you voted for Bush, because you don't want any repercussions.

I think this is a much more likely picture of a Reluctant Responder than one who voted for Bush--especially one who voted for Bush in a Republican precinct. What would they have to fear? Why would they be shy in this Bush Cartel controlled country? On the other hand, what would they have to fear if they defected from the Bush paradigm (which, quite frankly, I suspect a lot of them did)? I think they would have much to fear, and much reason to be silent.

I have never seen such fear-mongering, such out of control fascism, such naked bullying of the poor by the rich, such arrogance and irrationality, and such...what is the word? ...insanity in my country, and I have never, ever been so afraid for my country, as I am now.

This election can't really be compared to any other. We have not had these political and social conditions ever before. In the 50s, we had sane people in government, like Dwight Eisenhower, who helped put a stop to McCarthyism. And even in the 1960s, at the height of the Vietnam war protests, even when protesters were being shot dead, I never felt that the government had gone bonkers. I thought it would eventually respond to the clear will of the majority--and it eventually did. Even Nixon wasn't this crazy.

So I'm just saying that maybe the rules, and precedents, and what happened in previous elections are not all that relevant. Maybe some Republicans were shy of pollsters in previous elections. But everything has changed now. This is a far, far different country than existed in 2000. And, as I said above, to me, the opposite of the rBr is what makes sense. People being afraid of Bush and his bully boys--and voting against them, and then being afraid to say so--because Republican social culture has become so poisoned with repression and coercion.

I noticed above some comment and response about Febble not being an American. Maybe that IS a factor, Febble, in your view of these statistics. Because I don't know if people who are living in free countries can fully appreciate what it has been like living in this one, over the last several years. Maybe an older east German would understand.

I recall reading commentary somewhere, last year--I think it was from an American who had been ensconced in France for a while, and traveled a lot between Great Britain, France and the U.S.--about Americans being misinformed and asleep. And I just felt that this person had no idea what it was like being trapped in a supposed democracy with a constant barrage of propaganda and brainwashing from every news monopoly, and being too poor to travel and visit other countries for a new perspective, and trying to sort all this out, and figure out what is real, and what is true.

The startling thing, to me, is that, given these conditions, the majority of Americans did figure it out. They saw through all the B.S., and came out in droves to vote the Bush Cartel out of office. That's what the total of the evidence shows--from the Dem blowout success in new voter registration (almost 60/40) to the long lines in Ohio, to the overwhelming disapproval of Bush policies and dismal Bush approval ratings in all opinion polls.

To many of us it is patently obvious that Bush does not represent the majority of people here, and was not elected. What we are looking at is a fascist coup. And since the evidence of the fraud has been largely hidden from most Americans--I mean, the TV networks CHANGED these exit polls, just went and CHANGED them, denying Americans this important information--we can easily get impatient and even nuts with just normal academic questioning. We have been lied to so much. We have seen so much corruption and ass-kissing from people who are supposed to be our intellectuals--not to mention outright disinformation campaigns, and smear campaigns, and black ops. And what we are facing--what this fascist regime may well lead to--is we, ourselves, and our loved ones, and our friends and political compadres, being ruined, imprisoned, or disappeared in the dead of night.

That's what this supposed election MEANS TO US. That's what its consequence could be.

So please, please, forgive us, if we get impolite sometimes--or even unreasonable and wild. It is a very difficult time to be an American, and to see our democracy destroyed, and to see the truth about this regime smothered over.

We have seen so much crap like this "rBr" theory (not incidentally from people--E/M--who are trying to cover up their cover up)--we get crap like this from Karl Rove and Dick Cheney every day, and from all their ditto-heads--you can understand our touchiness. It reminds me of a discussion I had with someone close to me about Colin Powell's 100% lies to the UN about Iraq WMDs. At the time, I only knew for sure that it was about 50% lies, but I could FEEL it was all B.S. I KNEW it, as sure as I've ever known anything. She thought there was something to be concerned about--a reason to invade Iraq (even if only half of it were true). And I couldn't seem to get across that these Bush toadies lie for breakfast, lunch and dinner, and then go home and tell lies in their dreams. That's the real trouble with "rBr"--it smells like a Karl Rove lie, like this mythical "invisible" Republican voter registration campaign they were talking about the other day. And whenever Rove or Cheney tell a lie like this, you can ask yourself what the opposite of that is, and you will probably find the truth.

rrrk. And that's probably where they stole the most votes, too--from Republicans who voted for Kerry.

[emcguffie]

45. You are right.

I know you are right.

Bush voters are obnoxious and they are not shy. In fact, they scream at you.

Many registered R's voted for Kerry.

People in this country are not crazy.

I also imagine there are quite a few who did vote for Bush that are sorry they did so.

[emcguffie]

46. You know, TIA --

I may not know statistics, but I saw that the day after the election. I mean, the basic shift in percentages without adding enough voters to shift the percentages that much.

But I thought that was just evidence that they changed the polls to match the vote. Which we know they did. It doesn't prove the exit polls were right, does it? Although I believe just up and change for one day on November 2 because it likes *. It had to have some help.

[TruthIsAll]

47. Prove the recorded votes were "right". You can't. eom

/

[emcguffie]

48. of course not.

i should have edited that post, sorry I didn't read it.

People seem to want absolute proof, like a fingerprint, that the election was stolen. It certainly seems that there is plenty of statistical evidence that says it was stolen. And if all the statistical evidence points in the same direction, that in itself could not be random, could it? Like all the mistakes coming out in favor of the same side?

The problem is, many people want some other kind of evidence, because, for whatever reason, statistical evidence is not persuasive to them. They need agreement from elsewhere -- such atheir neighbors or their friends. And in this instance the media is not going to build any kind of consenus. So public agreement might help to persuade them. But absent their own confidence (in what is in front of their face) with no reinforcing agreement, they can't accept that there was fraud. Too afraid of appearing to be a nut or a sore loser.

Now, regarding your stastistical stuff, I have to sit down with it and concentrate to figure it out, and often I am not up to it.

could you pick what you think is the most blatant, obvious, anyone-can-understand-it evidence?

Then I think we should try to do little mass mailings. Since we have no media access whatsover, all i can think that we can do is on a grassroots level do community mailings. Because we can get our community mailing addresses.

"Dear neighbor,

Has it occurred to you that our last election was not entirely honest?
Perhaps the one before wasn't either.

No matter who you voted for, we must make sure that our electoral system is ..." gee, I don't know, i'm just tossing this off right now.

My point is, many small mass mailings to our neighbors is the only way I can come up with to reach them, absent any legitimate media.

Then I think we have to start some legitimate media. we have to start a newspaper and a news channel. Period.

I know that sounds ridiculous, but it is what we must do.

But we need a thorough strategy before we waste the postage on an incomplete message.

Maybe even phone banks would be useful. "Hello, I'm your neighbor. Has it occurred to you that the 0204 election was stolen? Are you interested in evidence that the corporate media will not put it in front of you?"

This is percolating in my brain and won't go away.