So I'm watching C-Span

[Goldeneye]

and the guy is talking about exit polling in the 2004 election. He had a piece of paper which he wrote out on election day at CNN headquarters that had Kerry winning 51% to Bush's 48%, however he notes that he was supposed to be very careful with the data, because it was based on 58% women. So I have a few questions

1. Why would the exit poll companies send out info that hadn't been
reweighted, when it is seemingly very easy to do?

2. The guy is named Stuart Rothenberg. Is he reliable?

3. Did CNN put data that wasn't reweighted up on their site? If so, why would they do that, when they had been warned that the data was unreliable.

4. What are exit polls really used for? If they aren't meant to call races what are they really for?

[saracat]

1. Think he is part of the Arnebeck suit.

[Goldeneye]

3. I don't think that would be a good thing.

He pretty much pinned this on bloggers.

[Al-CIAda]

2. Release the RAW data

..like Conyers has asked.


What are they hiding?

[ClassWarrior]

4. Good question.

What are they hiding??

NGU.

[Goldeneye]

5. You guys make me feel better,

but does anyone know how to answer these questions, esp. the first and fourth ones.

[roenyc]

7. how is it possible that everyone

gets away with all of this? lets assume this was a robbery and i am the thief. the store i robbed wants evidence and my lawyer says um, nope cant have it. sorry. cant.

thats it. done! America, wonderful country! i can just say no.

[redacted]

6. Release the RAW data

Agree. This could be the biggest cover up in our nation's history.

[sandnsea]

8. Shouldn't have mattered

Because they should have noticed the uptick in both male and female voters for Bush, which would have offset the difference in more women anyway. Do we have early exit polls that show the breakdown for women voters? There was a higher percentage of women voters this time, they just voted for Bush, supposedly. Where'd these women come from?

[anamandujano]

9. What about all the democrats waiting in line 5 hrs to vote?

They didn't get exit polled. The repukes voted easily and got to talk to the pollers easily while the dems stood in line in the rain.

Shouldn't the exit polls consider them?

[Goldeneye]

10. *sigh*

does anyone know how to answer these questions? They seem pretty central to the exit polls were off scenario.

[sandnsea]

11. That's a good point too

I'm referring to the women vote though, it would be interesting to see how that vote changed throughout the day as well. We really ought to have the raw data.

[Kindigger]

12. I do know I posted

...about that guy yesterday, but couldn't remember his last name (sorry). He said someone "accidently" got hold of the data, and it was never supposed to be made public. The bloggers picked it up, and ran with it.

He also said "there are a few not so sane individuals out there, who think the election was stolen".

[suziedemocrat]

25. I guess that makes me insane!!!! And proud of it!!

[fooj]

13. wasn't 51%-48%

the final faux total? Looks like an interesting "switched" set of numbers, doesn't it?

[SmokingJacket]

14. Good questions.

I can't really answer them either.

But #1 & #3 -- Yeah, why WOULD they send out unweighted numbers? My answer: they wouldn't. They came up with that excuse later.

And #4 -- Exit polls are used in other countries to validate the election. If an unbiased, disinterested party projects the actual winner, then the election is considered credible.

We're given to believe the only reason we have them here is to get the numbers early, as part of the media's game. But what's the point, then?

Fishy fishy fishy.

We're in such a fine mess.

[tritsofme]

26. Its my understanding that the networks

got the raw data as soon as it came in from Mitofsky.

They paid for it, and they got it as soon as any data became available.

[witchhazl]

15. #4 -- The exit polls are used to "analyze" the results after the election

That's why they collect statistics on gender, income, race/nationality, religion, etc. That's why they asked about why people voted as they did -- giving rise to the "moral values" debate.

They're also used, of course, to "help" call winners early, before the total vote is in.

[PBS Poll-435]

16. DEBUNKED!!!!!!!

NO WAY!!!!!!

:puke: :mad: :puke:

CNN/MSNBC/FoxNews own websites showed that there was not 58% women for the sample that showed Kerry winning 51/49. NOT THE ORIGINAL POLL or the "FIXED" ONE

But ok.........lets figure the results with this CRAZY 58% number.

If 58% Women and 42% Men voted for Kerry 51% then using the ratios from the "unfixed" exit poll then......

51.724% of Women voted for Kerry
50.000% of Men voted for Kerry

If 53% of Women and 47% of Men voted for Kerry then

51.887% of Women voted for Kerry
50.00 of Men voted for Kerry

Not a huge change. With like 97% of respondents in for the Ohio exit poll, Kerry had a 51/49 lead.

Am I right?

[Goldeneye]

17. I don't know what percent of women were voting for Kerry

but the 58% is refering to the percent of women used in the sample...thus the "explanation" that women were over sampled.

[PBS Poll-435]

22. My Point Was...........

It didn't matter what ratio of women voted for Kerry.......

Slate/Druge told us that Ohio was 51/49 for Kerry with "58% women"

CNN/MSNBC/FUX told us that Ohio was 51/49 for Kerry with "53% women"

The percentage of women that needed to vote for Kerry for him to be at 51% was very, very small. It was less than a .2 difference based on the fact that the uncorrected exit poll showed 50/50 split for men.


I think my math is correct......

[Goldeneye]

27. Those were the results for the national poll...

but it is significant that Kerry was leading both men and women in the exit polls in Ohio. Something happened with those exit polls.

[sandnsea]

21. Well we don't know

Because that 58% was supposedly early in the day and changed over time, according to the final exit polls.

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

Also, comparing national and Ohio, more men voted in Ohio than nationally. And women went 50/50 in Ohio, as compared to 51/48 nationally.
http://www.cnn.com/ELECTION/2004/pages/results/states/OH/P/00/epolls.0.html

That's why I'd like to see those exit polls early in the day, how were those women voting?? For all the post-mortems done, not much attention has really been given to the woman's vote, except to say it's the security mom vote and I just highly doubt that. My sister is 100% Republican and she isn't even thrilled with this war, so I don't see how Democratic women would have shifted their votes to Bush.

[GetTheRightVote]

18. Let's not forget ABC could not put out the video until after Nov 2, why is

this always the case when it should all be based on facts and the truth which is out there for us to discover. Like * having heart problems but did not get his physical until after the election. We really need to start picking up on these clues and acting on them whether we can get them first hand or not. Why else would they hold them from us, the reports.

:kick:

[Goldeneye]

19. What video?

I don't think I've heard about that.

[shiina]

20. The Freeman paper talks about #3

Under the section heading "Is the Data Valid?" he states:

"NEP’s predecessor, Voter News Service, warns in bold letters in its 2000 Methodology statement never to use unweighted data (see Appendix A). It makes no sense ever to distribute unweighted data to anyone. Pollsters want to get it right. Their customers are depending on it. Broadcasters want to be alerted to probable outcomes, so as to plan their coverage accordingly (e.g., pre-writing stories so they could be completed shortly after poll-closings, assigning top reporters to winners’ headquarters, being prepared for when concession or victory speeches might be forthcoming, etc…). In this case, subscribers were very surprised. Editors and network managers had to scramble, and journalists complained that they had to rewrite their lead stories."

In other words, "no way!"

http://www.buzzflash.com/alerts/04/11/ale04090.html

[madison2000]

23. Exit polls are used to call races and they are considered very reliable

I remember on election night that AP called the race for Senator Feingold based on exit polls, when less than 2% of the vote had been counted. He was expected to win by a wide margin, but it really surprised me that they called it so early. Exit polls are considered very reliable, unless you are screwing with the voting equipment or ballots....

[Carolab]

24. Don't know who Rothenberg is but read this

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=201601&mesg_id=201601

I posted it yesterday. Both articles linked are very interesting and pertain to the discussion at hand.