CBS, CNN, NBC Florida Exit Polls 2004

[RaulVB]

Take a look at this:

http://election.cbsnews.com/election2004/poll/poll_president_fl0.shtml

http://www.cnn.com/ELECTION/2004/pages/results/states/FL/P/00/epolls.0.html

http://www.msnbc.msn.com/id/5297152/

There are 2 WAYS OF READING EXIT POLLS DATA:

- Vertical (CBS model)

- Horizontal (CNN, NBC models)

[Goldeneye]

1. What about it?

44% of Kerry voters were women, and 56% of Kerry voters were men. What am I missing?

[L. Coyote]

6. Florida exit polls and the vote count. OJ and OZ !!

This is why exit polls are "adjusted." Since more than 44% of the voters are women, the answers by the women polled must be used to determine how about 50% voted.

When analyzing the election canvasses, the problems of inference do not prevail. The entire population has been polled, and there is no need to toy with the data to "adjust" it to hopefully accurately predict, within a margin of error determined by sample size, how the entire population voted.

Of course, in a fraudulent election, the final results are also "adjusted" by deception. By working on the election canvasses, we can determine far more, and without inference, adjustments, or margin of error. The exit polls do point to where to look.

Check out this relationship:



Ordinarily, the percentage of Democrats predicts the number of votes for Kerry, on average for counties in IA, CA, and SD about a 0.83 correlation. Not in Florida, the land of OJ and OZ, where the opposite proves true for the county percentages. This data is based on actual election results, not inferences from samples.

[RaulVB]

7. Great points

Undisputable!

[L. Coyote]

26. Move to this thread for election results analysis, the real numbers

not the sample we know very little about. http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=201&topic_id=7087#7156Charts"> Charts of Bush 2000-2004 shift and Pres.-Senate split.

[tritsofme]

21. This is what Gallup said about exit polls before the election

When asked why they do not weight their polls for party idenification based on exit polls.

Some consumers of the polls (including the Gallup Poll) have questioned poll results because party identification and other characteristics do not match the 2000 exit poll data. There are very good reasons why they may not match the exit poll data.

First, some treat the exit poll as a census. It is not a census, it is a survey based on sampling of voting precincts. There is a reason it is called an "exit poll" and not an "exit census." That's because it is a poll, and as such is subject to sampling variation and other polling errors just as any other poll would be. In fact, because exit polling relies on quota sampling (hand selection of survey spots according to population size and other factors), it has a higher degree of potential error than do the random samples on which telephone surveys are based. Thus, there is no basis on which to believe the exit poll numbers are in any way more accurate than any other number you get from a poll. They are all estimates. The one advantage of the exit poll is that they know everyone they interview is a voter, while pre-election polls rely on models to determine who is likely to vote and who is not. However, that does not mean their estimates are necessarily better, and they are definitely not error-free estimates of the electorate as many treat them. In fact, when multiple exit polls existed in the past, they very routinely differed in their estimates of the vote as well as their estimates of the demographic characteristics of the electorate. Even today the Los Angeles Times exit poll differs from the larger exit poll used by the networks.

Second, the exit poll measure of political party ID is fundamentally different from ours. We know that survey results can differ depending on how the data are collected. Our questions are read and responses obtained verbally over the phone. Their responses are obtained in self-administered questionnaires that present the questions in a visual format. Most survey research experts would be extremely cautious in comparing data obtained by a telephone interview versus that obtained in a self-administered paper-and-pencil questionnaire. That is in addition to question wording differences in the party ID question that can also have an effect on the results.

Third, a lot has changed since 2000. In the post-9/11 environment, terrorism has become one of the chief problems for government to deal with. The Republican Party has a large perceptual advantage on the terrorism issue. To assume that everything is as it was four years ago is a very risky assumption. While it is possible that in the end things could change once again so that partisanship looks much like it did in 2000, that is by no means certain or even likely.
http://www.gallup.com/poll/blog/default.aspx?a=10012004

[TruthIsAll]

18. Wrong. 56% of Kerry voters were women, 44% men.

The columns add to 100% for all categories.

[L. Coyote]

2. Too bad the percentages are rounded off.

[RaulVB]

3. The other way around!

Are you reading it vertical?

[RevCheesehead]

4. I honestly don't understand what you are looking at.

What is it that we're supposed to see here?

[RaulVB]

5. I edited the original post

We have to compare what is posted by the 3 networks.

[RaulVB]

8. CBS Exit Poll posted focus on voters universe

Kerry Data:

- 56% of Kerry's voters ID as WOMEN

- 44% of Kerry's voters ID as MEN

Bush Data:

- 52% of Bush's voters ID as WOMEN

- 48% Of Bush's voters ID as MEN

Let's take NBC now:

- Bush "beats Kerry" 50% to 49% among WOMEN

- Bush "beats Kerry" 53% to 46% among MEN

[RaulVB]

9. Guys, are you really not getting it?

Are you familiar with the concept of "Gender Gap"?

[RevCheesehead]

10. Please don't be offended, but your posts are a bit like talking to my mom.

She sometimes starts in the middle of a thought, and I have to get her to backtrack to tell me what she's talking about.

I can only guess at this point, but you have obviously found something interesting related to the "gender gap" in the exit polls.
So, could you please relate it to the larger context?

i.e., what has the gender gap been in past elections? Is this anomaly related solely to FL, or are there other states that should be looked at carefully? How is gender gap relevant to this election?

I don't come here with a lot of experience in polling, etc...I'm kind of learning as we go. But it would be helpful if your posts included some kind of summary at the beginning, telling us what you are looking at, and why. Otherwise, a good find might be ignored, due to lack of information.

PS - LOVE your avatar!

[RaulVB]

11. Sorry, is a very "peculiar" thought process I know, and I don't mean it

I posted something more direct in a different thread but the point would be:

"12 points of gender gap for Kerry."

"4 points of gender gap for Bush."

* That is what is CBS telling us. Curiously enough, they do not care about repeating by how much "Bush beats Kerry", no, they want to explain to us how the voters of Kerry and Bush ID themselves.

CNN and NBC are telling us:

OK, what you "need" to know is only this:

- Bush "beats" Kerry by 7 points among men
- Bush "beats" Kerry by 1 point among women

That explanation does not quite close the gender gap between Kerry and Bush.

Thanks for your comments.

[RevCheesehead]

12. And thank you for replying!

So, would it be reasonable to ask if the gender gap for Bush was legitimate? Could there have been a padding of votes for Bush that closed that gap?

(am I close?)

[RaulVB]

15. More than reasonable!

It shows that votes were padded to open up the gap among males for Bush with respect to Kerry.

I'll explain with more detail in a new post below in few minutes.

[Bill Bored]

13. Or are you simply saying that the massaging of the exit polls

(as reported by the various news organizations) just doesn't add up? (Sorry I haven't had a chance to crunch too many of these numbers myself yet.)

I do note that although income levels were requested, employment status was not, or at least it's not in these reports. I had a theory that was met with little or no interest in another thread, that unemployed voters (at least 5.4% of the population) voted earlier and tilted those results toward Kerry, but I'd be happy to have it debunked!

[RaulVB]

17. Massaging males backs

Indeed.

Let's look at the total of Florida votes and let's assume that the exit polls are an "accurate representation" of the voters numbers.

CNN reports this:

The total combined of votes says that Bush + Kerry =

* 7.530.165 millions

- Bush "gains" :

2.056.941.12 millions (females) + 1.898.714.88 millions (males) =

* 3.955.655 millions for Bush.

- Kerry "totals":

2.001.775.1 millions (females) + 1.572.783.9 millions (males) =

* 3.574.558 millions for Kerry.

Now two major considerations:

1. The portrayed difference between both candidates is 381.000 votes.

- Kerry "lost" 2 percentage points respect Gore and also "did not gain" the 2 points that Nader got in 2000.

Compared totals adding both figures, 2004 value:

* Kerry "loses" 313.000 votes. I sustain that he really got those votes, based on electoral voting patterns.

2. He has an advantage of 323.000 votes among women respect to his own male voters, or a 12 points gap.

Bush has "advantage" of just 158.000 votes among women respect his own male voters, or a minor 4 points gap.

IS JUST ONE THIRD OF KERRY'S GAP.

I sustain that the voting patterns reflected IN KERRY'S CBS EXIT POLL gender results represent the trends of his REAL VOTING TOTALS.

HE DID WIN THE FEMALE VOTE IN FLORIDA IN THE SAME PROPORTION, perhaps an extra 1.5% over the totals or 112.000 votes (roughly)

The final tally would be:

Kerry : + 425.000 votes

Bush : - 425.000 votes

Real Numbers:

John Kerry : 3.999.558 millions (53% roughly)

George W. Bush : 3.530.655 millions (47% roughly)

I assign a MOE to my numbers of +,- 1.5 %, founding all of the above in what I believe is a clear model of Gore's victory in Florida and the Nader factor plus the addition of elements from Florida 2004 voting patterns, both internal and external, like the voter registration, voter turnout, AARP new plans, the war in Iraq, resentment for Gore's loss, etc., etc..

* Sorry for the length of the post.

[Dolphyn]

14. Here's a breakdown. The gender data looks consistent to me.

CNN data gives:

2862 * 46% * 46% = 606 Male Voters for Kerry
2862 * 46% * 53% = 698 Male Voters for Bush
2862 * 54% * 49% = 757 Female Voters for Kerry
2862 * 54% * 50% = 773 Female Voters for Bush

According to the above numbers,
Kerry voters % men is: 606/(606+757) = 44.4%
Bush voters % men is: 698/(698+773)= 47.5%

This is consistent with the CBS data (within rounding errors).

So ... Kerry is more popular among women than among men, no surprise there.

[ClintCooper2003]

16. Excuse me, but the msnbc and cnn polls are the "adjusted" data...

from after 1 a.m.

[althecat]

19. Precisely...

And given that these news organisations have posted this data. Isn't it logical that they should want the actual data in place.

Perhaps the correct approach would be to direct requests on this to each of the NEP's ombudsmans..... this datas is misleading. We could embarass them into releasing the actual data.

al

[Krocksice]

20. also

that cnn site probably has their 'adjusted' exit polls up. I'm sure everyone is familiar with the big switcheroo that CNN pulled at about 1 AM on election eve. They pulled all their early exit polls with no explanation and put exit polls up that more closely matched the outcome.

[newyawker99]

22. Hi Krocksice!!

Welcome to DU!! :toast:

[Bill Bored]

23. OK, it was the wee hours of the morning. I was tired but

from what I could see, Dolphyn is right. There is no discrepancy between the CBS and CNN polls although they have different reporting times.

So what are you saying, Raul?

That Kerry actually won the female vote in FL while the exit polls say he lost it?

That Bush's gain of 5% (to 50%) shown on CNN is bogus?

And what is the proof of this again? I don't see it in the poll results anywhere. Was it in the earlier leaked ones before the infamous RED SHIFT?

Thanks for clarifying. :-)

[troubleinwinter]

24. See if this thread helps

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x143475

[Bill Bored]

25. I understand the apples/oranges stuff perfectly, but

the numbers in both polls do in fact agree, if you look at Dolphyn's post above in this thread.

I'd like to know how we can determine that they are wrong? It has something to do with either the 2000 results, or the earlier polls of Nov. 2, 2004.

To my untrained eye, it seems that Raul is questioning the movement of 5% of the female votes to Bush this year. So what am I missing?