Talk:2004 United States presidential election controversy, exit polls

For a December 2004 deletion debate over this page see Wikipedia:Votes for deletion/2004 U.S. presidential election controversy, exit polls

---

Statistics for exit poll/actual result discrepencies

The article tries to point out that the discrepancies between exit polls and actual results were greater at e-voting places, but the cited reference [1] does not give a mathematical calculation of this (1 in 50,000 refers to the chance that Bush did better than exit-polls preedicted over-all). Does anyone have a reference for the e-voting discrepancies? If not, it could be fairly easy to calculate given the raw data (by performing a chi-squared test or similar - I assume that someone, somewhere, has done this). The article says that the discrepancies at e-voting booths are significantly greater, but this is a precise statistical term and should be backed up with numbers. — Asbestos | Talk 11:36, 30 Nov 2004 (UTC)

Moved from main article talk

New Discrepancies Map

This page has requires using the history to read it, I've never encountered that before. Anyways, I have a new map, of discrepancies as of nov.3 12:22am, from this source (and the updated excel file is uploaded too, if anyone wants it)

Kevin Baas | talk 22:40, 2004 Nov 12 (UTC)

What is the maximum gain (among the blue states according to this map) of exit poll discrepancy in kerry's favor? Are there pre election day polls that agree or disagree with blue state (according to this map) discrepancies? What are the odds of all the irregularities in kerry's favor? Is this based off of final exit poll data or non "weighted"? Zen Master 23:06, 12 Nov 2004 (UTC)

Kizzle, this isn't how talk pages are normally used and this kind of break from the norm should be discussed first. I feel people should be free to create threads here as they wish, and you shouldn't archive discussions that were started literally minutes ago. I moved all of today's discussions back to this main talk page, where they belong. Rhobite 22:55, Nov 12, 2004 (UTC)

Kevin, please stop using sources which anyone can post an article to? Wanna take a look at my article [2]. -- Netoholic @ 22:56, 2004 Nov 12 (UTC)

If you don't think like wikis or forums, don't write in them. Kevin Baas | talk 23:01, 2004 Nov 12 (UTC)

Wikipedia is not a discussion forum, nor does it use source which can not be independantly verified by our readers. It seems I am not the one who is writing in the wrong place. -- Netoholic @ 23:34, 2004 Nov 12 (UTC)

I agree with you on all but the last point. Kevin Baas | talk 23:44, 2004 Nov 12 (UTC)

Actually Netoholic, that is the essence of your misunderstanding, wikipedia IS a discussion formum for resolving article dispute, please learn. Zen Master 00:29, 13 Nov 2004 (UTC)

new map:Hybrid us map

Kevin Baas | talk 21:47, 2004 Nov 15 (UTC)

Opps, I compensated the wrong way. Sorry. Here's the real map. According to the a variance of the discrepancy between the compensated polls and the vote counts for the non-BBV states, there is a 22.5% chance the poll is off by enough for Bush to have won Ohio (and a 50% chance that it was off in that direction), and a 47.5% chance that it was off enough for Kerry to have won Florida (and a 50% chance that it was off in that direction), giving Kerry a 91.4% chance of victory. Kevin Baas | talk 06:55, 2004 Nov 16 (UTC)

By the same logic, if we use instead the average of the vote count and the exit polls for the BBV states (meet halfway), Kerry is given a 55.5% chance of victory.

And if we give the vote-count a 50% chance of being perfectly accurate, and a 50% chance of being meaningless, that gives Kerry a 91.4%/2=45.7% chance of victory. Kevin Baas | talk 07:18, 2004 Nov 16 (UTC)

Opps, those figures are with New Hampshire in the BBW states. (I am not doing well with clerical perception lately!) Those figures are interesting as well, and with New Hampshire included in the bias compensation, the figures are pretty close. If people think any of this should be in the article, or if the data is really important to someone, I can recalculate.

Commentary about exit poll accuracy

A good part of the suspicions of fraud in the election seems to hinge on the accuracy of the exit poll data. Doing a google search I came across this commentary by Howard Fienberg and Iain Murray. Quote:

Historically, exit polls have been more reliable
than regular polls and the news media treat them as gospel.
But if the poll result is close, anyone who tells you
that they know who has won is lying - November 7, 2000

(veryfied with wayback machine: this page was up august 2002) Besides being definitely unbiased concerning the current election, it mentions several cases of exit polls that were quite of. Those might be worth investigating to get an idea of the circumstances under which exit polls can yield inaccurate results. --Icekiss 14:04, 12 Nov 2004 (UTC)

When I said "evidence", thats exactly what I meant. Nice one!

Wariness of exit poll reliability is also why I asked for a similar comparison of exit poll / popular vote for other recent or parallel elections too, to indicate how well they track in other elections for comparison. FT2 18:30, Nov 12, 2004 (UTC)

Please see dnamining.com/exit for another analysis. I disagree with your characterization of the Caltech/MIT paper, although I agree it is not a good paper. -wjb(newbie)

Here is a relatively thorough explanation of how exit polls work and an evaluation of their reliability: http://www.mysterypollster.com/main/2004/11/exit_polls_what.html Quote:

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 

(anon, 18:00 Nov 15, 2004 PST)

The issue with the exit polls is not with their general methodology, which is certainly very reliable. There is always selection error, and it's important to list that as a source of error and discuss it. This article has been written so thoroughly from the POV of democrats/liberals that it's not even possible to insert gracefully the real issue with exit polls: that democrats are more likely to live in cities, therefore, they are more likely to vote multiple times, therefore they are more likely to 'vote' in exit polls multiple times. This is why Democrats say, 'voter disenfranchisement' but Republicans say 'voter fraud'. And in our day and age, I'm afraid that the staunchly pro-Democrat African-Americans are more likely to be disenfranchised by a criminal record. I'm not racist, I've looked at the demographics, and unfortunately, young african american males are more likely to go to jail than to college. It's terrible. But it's also the case that a lot of votes should be thrown out that aren't, and that it benefits the Republicans to be more picky than the Democrats. I think it's a mistake to spend so much time on sour grapes, but if you're going to have this article, you should offer more than one perspective.

Like this quote from a link above: d about voting for Wilder.

"* In the Republican primaries in New Hampshire in 1992 and in Arizona in 1996, exit polls overestimated the vote for Pat Buchanan. The most likely reason for the mistakes, which resulted in misleading news coverage? Zealous Buchanan voters were more willing to participate in exit polls. In 1992, exit polls predicted a George Bush win by only a small margin over Buchanan -- Bush actually won by 16 points. In 1996, exit polls put Bob Dole a distant third after Steve Forbes and Buchanan - Dole actually came in a close second to Forbes. " And I think everyone will agree that the Democrats were more zealous in this election. I'm not asking for a 'Fox News' style article, but I'd like some thought put into these things by people who care about balance rather than proving a point. It's not an encyclopedia for the world if alternative POVs are deleted- as my previous attempts were (though I didn't have an account and I don't know if that hurt me).

I think you do offer a variety of sources, but you haven't structured this as an article, but as an argument. A balanced argument would include the actual evidence rather than citations- instead of quoting biased professors verbatim, and name-dropping, simply include the statistics and reasoning itself. I don't care about anyone's paper, I care about the evidence. - solomonrex (noob)

the real issue with exit polls: that democrats are more likely to live in cities, therefore, they are more likely to vote multiple times, therefore they are more likely to 'vote' in exit polls multiple times.

How does geographical clustering of party affect one's probability of multiple voting? Do you have any articles or prominent discussions you can reference to in hypothesizing that democrats were more likely to vote multiple times? And yes, while its not official wikipedia policy, if you register with an account, your edits are more likely to be interpreted as contributions and not vandalism. --kizzle 16:27, May 6, 2005 (UTC)

Kizzle's right, this is an unsubstantiated allegation, and, as such, violates the Wikipedia:Civility policy. And on a side note: let me correct your logic: that would not effect the accuracy of exit polls: the sample remains in proportion to the vote. Kevin Baas^{talk: new} 17:48, 2005 May 6 (UTC)

Thom Hartmann states that in Germany, [3]

"... people fill in hand-marked ballots, which are hand-counted by civil servants, watched over by volunteer representatives of the political parties. ... It's totally clean, and easily audited. And even though it takes a week or more to count the vote ... the German people know the election results the night the polls close because the news media’s exit polls, for two generations, have never been more than a tenth of a percent off."

Unfortunately he's wrong. Generally all votes are counted (twice) by midnight. Then the preliminary final results are announced and the candidates who won are contacted so they can state whether they accept their seats or not. They have a week to reply and then the final results are announced; it's got nothing to do with counting votes for a week (btw. the final result of the election as a whole can take even more time because somtimes some precincts vote later than the other. This year a candidate died in Dresden after the ballots had been finalised and therefore the precinct voted two weeks(?) later) I therefore removed that part from the article. If Thom Hartmann's got something useful to say about voting in the US and Germany perhaps someone could find a correct quote to replace the one above.

CNN screenshot redux

Look, I know this was discussed on the main page already, but that stuff about someone's screenshot showing dodgy info: that's really NOT reliable. C'mon guys! I've been very supportive of these articles, and voted to keep on nearly all of them. I've also gotten into a massive fight with Netoholic about the whole deal because I was trying to save the page, but I totally agree with him about this issue. Can we please consider removing this information? It's honestly eating into the credibility of the article. - Ta bu shi da yu 16:57, 17 Dec 2004 (UTC)

Is your concern/thought process centering around the information in the screenshots, the possibility that the screenshots were forged, or the fact that an allegation of irregularity surrounds the 'fixed' exit polls? Cause I can see a lot of ways editorially to approach the issue in 3. without the screenshots themselves. -- RyanFreisling @ 23:24, 17 Dec 2004 (UTC)

Changing of the exit poll results to reflect the actual precinct counts is Mitofsky standard procedure. What would "eat into the credibility of the article" is if the article showed the results not changing to reflect the actual vote counts. Kevin Baas | talk 00:09, 2004 Dec 18 (UTC)

The screen shots have actually been entered in as evidence in the lawsuit in ohio that challenges election results. But more importantly no one officially connected with the exit polling companies, as far as I know, including Mitofsky, has disputed the accuracy of those screen shots as reflecting the actual "unweighted" exit poll data, that should tell you something (it tells me a few things). So, because the screen shot data has effectively gone unchallenged by anyone officially connected with exit polling companies themselves, and because it has been submitted as evidence in an Ohio court of law, it's certainly acceptable to include in a wikipedia article I believe. If you have citations of someone officially challenging the data please provide them. zen master 00:22, 18 Dec 2004 (UTC)

exit poll data link

Where'd you get that data from? Is it corrected to match the vote count? --kizzle 06:09, Mar 4, 2005 (UTC)

NEP. Yes, the weighting was applied 1:34am, Nov 3. 2004.

From the Wordfile:

"Respondent weights are included with each data file. They can be found in columns 18 through 23 of each file. These weights should be applied in all analyses. Unweighted tabulations will be seriously misleading and should not be used. In the SPSS data file, these values can be found under the variable label WGT. In the combined data file, this value can be found under the variable label XSWGT and is located in columns 55 through 60 of the ascii file."

"For additional information on weighting, see the methodology statement on page 12."

I believe weight and "correction" is different... as in the 12:23am info was weighted in several aspects (as in towards gender breakdown) but it wasn't adjusted against the official vote count... I could be wrong though. MysteryPollster has an article on it: --kizzle 01:00, Mar 5, 2005 (UTC)

I have not had time to do more than skim the some of the documentation, but on first glance, the files appear to be consistent with the previous releases of respondent level exit poll data. The files include only one "weight" variable -- the one that includes a "correction" to match results with the actual count. I also see no precinct level data nor any other means of replicating the "within precinct error (WPE)" analysis from the Edison-Mitofsky report. If I that turns out to be true, those who have been demanding the release of "raw data" are going to disappointed, to put it mildly. - http://www.mysterypollster.com/ (January 31, 2005)

Indeed. -- RyanFreisling @ 01:04, 5 Mar 2005 (UTC)

Freeman and Mystery Pollster

Hi there. THe article now states that Freeman "listened to several posts on Mystery Pollster" (a blog). Is there corroboration of Mystery Pollster's role in Freeman's revisions, - and if so, can we correct it to 'considered' or 'read', as appropriate? -- RyanFreisling @ 00:41, 6 Mar 2005 (UTC)

Yeah, if you visit the site you'll find the post by Freeman himself:

Hello, Mark:

I’d like to thank you for taking the time for offering this detailed critique of my paper, and more generally for your knowledgeable commentary on polling processes. Since writing the draft you read, I’ve learned a great deal about polling – in large part from reading through your site. I’ll have a revision of my paper out in a few days, which will be much stronger for having read your commentary.

Regarding your post, I’m going to respond to several points and then give a general response to what I see as the big question:

1. Data. I’m happy to make my CNN data available. I have 49 states & DC (only Virginia missing if anyone has that), although for a few I don’t have sample size. Just tell me where you’d like me to send it to or post. (My own personal and University websites have been going down from too much traffic.)

2. High degree of certainty (Your point 1). I agree that I overstated the case, should never have cited Hartmann, and did not understand the logistical challenges you explain. NEVERTHELESS, logic and evidence still indicate that exit polls should be a good basis for prediction, and although I can understand why the logistical challenges would increase the margin of error, it’s not at all clear why they should skew the results.

3. 250-million-to-1 (Your point 2). I see that I did put too much faith in stratification counterbalancing the effects of clustering, and will redo the calculations with the 30% increase. That’s a very good citation. NEVERTHELESS, as you point out, it doesn’t change the finding that **random error can be ruled out as an explanation.** This is really the main point of the first draft, because once chance is ruled out, some other explanation needs to be found.

4. Official “explanations. (Your point 3). My key point about explanations is that all we have -- at best -- are hypotheses. Perhaps Bush-voter refusal is a better hypothesis than I gave credit for, but it still is only a hypothesis. (Too many women would be irrelevant to the CNN data. Male and female preferences are reported separately and thus automatically weighted appropriately.) On the other hand, there are also creditable hypotheses, some with substantial evidence, which could have effected the tally.

I object most to belittling dismissals of these second set of hypotheses and allegations (e.g, Manuel Roig-Franzia and Dan Keating, “Latest Conspiracy Theory -- Kerry Won -- Hits the Ether” Washington Post, November 11, 2004; Tom Zeller, Jr. "Vote Fraud Theories, Spread by Blogs, Are Quickly Buried" New York Times November 12, 2004-Page 1), along with unquestioning acceptance as “explanations” the hypotheses and allegations about poll error.

In summary, I think that perhaps I biased my paper somewhat unfairly towards suggesting count errors as explanations, but that was probably in response to what I still see as an extreme bias at the press in dismissing them.

When you say that suggesting the possibility of count errors is delusional, perhaps you have done the same? (It seems as though you spend a lot of time on the tin foil hat circuit.)

Thinking coolly and scientifically: Is it delusional to question the Bush-voter-refusal hypothesis as conclusive without independent evidence? On the other hand, considering the scores of allegations, the history (especially in Florida), the lack of safeguards with electronic voting, the conflict-of-interest in election oversight, etc…, etc… (and now the Berkeley study) is it delusional to consider that, just possibly, even part of the discrepancy might be due to the possibility of miscount?

Yours truly, Steve Freeman

Posted by: Steve Freeman | November 18, 2004 10:32 PM

--kizzle 04:18, Mar 6, 2005 (UTC)

Visited the site, didn't see it at first. Thanks. -- RyanFreisling @ 04:26, 6 Mar 2005 (UTC)

Venezuela comparison

The article about the recall referendum referenced a poll by a "firm that was hired by businessmen who were almost surely opponents of Chavez".

But in the Venezuelan_recall_referendum,_2004#Claims_of_media_bias paragraph says other polls, some conducted by the opposition, were close to the results.--P8 07:48, 28 Apr 2005 (UTC)

Stale information

I am about to delete our statement that "Unofficial results currently indicate a victory by George W. Bush over John Kerry." Merecat already did this, but it was reverted by RyanFreisling (along with other edits), with a request for discussion here. I am considering replacing the deleted statement with "Leaked exit poll results indicated that John Kerry led George W. Bush in the popular vote, but official counts recorded more votes for Bush." Any comments? -- Avenue 00:56, 18 April 2006 (UTC)

The deletion looks fine to me. -- User:RyanFreisling @ 01:01, 18 April 2006 (UTC)