CNN official exit poll rationale -- Has this been posted yet?

[davidgmills]

http://www.cnn.com/2005/ALLPOLITICS/01/19/exit.polls/index.html

[Jon8503]

1. If Bush voters did not participate at all, how do they know more Kerry

voters participated just because they responded or am I reading this wrong?

[mdb]

5. They don't know.

They don't know if Bush voters chose not to participate just as they don't know if Kerry voters chose not to participate.

[euler]

8. Yes. They can only investigate...

...a finite set of possible explanations for the problem. If all of them are ruled out, they can only speculate. The speculation, however, is informed by over 30 years exaperience, 3000 exit polls conducted, and the fact that differential non-response because of age, time of day and political affiliation is known to occur in other exit polls.

[qwghlmian]

9. Here's a research paper on

non-responders and whether their makeup is similar to responders (conclusion: no, not similar)

http://www.duke.edu/~mms16/non_response2000.pdf

Our results also lead us to conclude that exit polls are likely to over-represent the opinions of younger and non-white voters. Because non-white voters tend to vote for Democrat candidates, over-representation of this social class will skew an exit poll’s results in that direction.

[davidgmills]

13. Just looking at the stats

I doubt they would have skewed that polls very much. A 13% differential between white and others would not amount to that much when whites were 77% of the voting population.

As for younger vs. older the younger category was up to age 40 so I don't think that says much of importance.

The stats all seemed to be Utah based which is heavily Mormon. So I don't know whether these same stats would be universal in the US.

Maybe what this analysis really means is that white Mormons over 40 don't like talking to pollsters.

[sabra]

2. "CNN did not air those inaccurate results"

Then their exit polls posted @ midnight were correct, and Kerry was projected to win?

[mdhunter]

3. Question.

If we are to believe this report must we not also believe that the polling methodology for perhaps half a dozen additional independent polling organizations is also flawed, given that the exit polling was generally in line with the latest preference polls immediately before the election?

If this is true... "The discrepancies stemmed from problems in interviewing voters at the 1,480 randomly chosen precincts where exit pollsters were stationed, not from how those precincts were selected or the way the data were processed, according to the 75-page report" ...then that bias is likely to exist in the methodology of every polling organization whose data aligned with Mitofsky's raw exit polls.

I, for one, think that to be unlikely.

[tritsofme]

15. Of the last 50 public opinion polls conducted, Kerry only led in 6

GW/Battleground (1000 LV) 10/31 - 11/1 50% 46% Bush +4
Rasmussen (3,000 LV) 10/30 - 11/1 50.2% 48.5% Bush +1.7
TIPP (936 LV) 10/30 - 11/1 46.9% 44.3% Bush +2.6
FOX News (1200 LV) 10/30 - 10/31 46% 48% Kerry +2
CNN/USAT/Gallup (1573 LV) 10/29 - 10/31 49% 47% Bush +2
CBS/NY Times (643 LV) 10/28 - 10/30 50% 47% Bush +3
ARG (1258 LV) 10/28 - 10/30 48% 49% Kerry +1
Newsweek (882 LV) 10/27 - 10/29 51% 45% Bush +6
Battleground (1000 LV) 10/25 - 10/28 51% 46% Bush +5
CNN/USAT/Gallup (1195 LV) 10/22 - 10/24 52% 46% Bush +6
Los Angeles Times (881 LV) 10/21 - 10/24 49% 48% Bush +1
Newsweek (880 LV) 10/21 - 10/22 48% 47% Bush +1
Time (803 LV) 10/19 - 10/21 52% 47% Bush +5
GW/Battleground (1000 LV) 10/18 - 10/21 49% 45% Bush +4
Rasmussen (3,000 LV) 10/17 - 10/19 48% 47% Bush +1
FOX News (1000 LV) 10/17 - 10/18 48% 43% Bush +5
CBS News (678 LV) 10/14 - 10/17 47% 46% Bush +1
CNN/USAT/Gallup (788 LV) 10/14 - 10/16 52% 44% Bush +8
Time (865 LV w/leaners) 10/14 - 10/15 48% 48% TIE
Newsweek (LV) 10/14 - 10/15 50% 45% Bush +5
GW/Battleground (1000 LV) 10/11 - 10/14 49% 46% Bush +3
CBS News (760 LV) 10/9 - 10/11 47% 46% Bush +1
ICR (763 LV) 10/9 - 10/11 49% 46% Bush +3
CNN/USAT/Gallup (793 LV) 10/9 - 10/10 48% 50% Kerry +2
Rasmussen (3,000 LV) 10/7 - 10/9 50% 46% Bush +4
Time (886 LV w/leaners) 10/6 - 10/7 47% 46% Bush +1
GW/Battleground (1250 LV) 10/3 - 10/7 49% 46% Bush +3
Fox News (1000 LV) 10/3 - 10/4 48% 45% Bush +3
ICR (762 LV)** 10/1 - 10/5 51% 46% Bush +5
ARG (800 LV) 10/2 - 10/4 46% 47% Kerry +1
CBS/NYT (561 LV) 10/1 - 10/3 48% 47% Bush +1
Zogby (1036 LV) 10/1 - 10/3 46% 45% Bush +1
CNN/USAT/Gallup (772 LV) 10/1 - 10/3 49% 49% TIE
Newsweek (1013 RV) 9/30 - 10/2 46% 49% Kerry +3
Battleground (1000 LV) 9/27 - 9/30 51% 44% Bush +7
LA Times (1100 LV) 9/25 - 9/28 51% 46% Bush +5
CNN/USAT/Gallup (758 LV) 9/24 - 9/26 52% 44% Bush +8
IBD/TIPP (649 LV) 9/22 - 9/27 45% 46% Kerry +1
Time (877 LV) 9/21 - 9/23 49% 43% Bush +6
FOX News (1000 LV) 9/21 - 9/22 45% 43% Bush +2
Battleground (1000 LV) 9/20 - 9/23 50% 45% Bush +5
CBS News (931 LV) 9/20 - 9/22 50% 41% Bush +9
Zogby (1066 LV) 9/17 - 9/19 47% 44% Bush +3
IBD/TIPP (650 LV) 9/14 - 9/18 46% 43% Bush +3
CNN/USAT/Gallup (767 LV) 9/13 - 9/15 55% 42% Bush +13
CBS News (1088 RV) 9/12 - 9/16 50% 42% Bush +8
Battleground (1000 LV) 9/12 - 9/15 49% 45% Bush +4
IBD/TIPP (674 LV) 9/7 - 9/12 47% 47% TIE
Newsweek (1003 RV) 9/9 - 9/10 50% 45% Bush +5
Zogby (1018 LV) 9/8 - 9/9 47% 45% Bush +2
Time (857 LV) 9/7 - 9/9 54% 42% Bush +12

[davidgmills]

4. This is the crock we have all been waiting for

[davidgmills]

6. Does someone want to get a screenshot of this before they

start denying they ever said this?

[understandinglife]

7. I've saved the entire page (n/t)

BE THE BUSH OPPOSITION; 24/7

[davidgmills]

10. Thanks

[davidgmills]

11. This is a hoot

Problems with the numbers first surfaced on Election Day, when exit polls showed Kerry with a 3-point lead nationally and an edge in some key battleground states. Those exit poll results were leaked and became widely known through the Internet.

CNN did not air those inaccurate results or post them on its Web site, and CNN's projections of winners on election night were accurate.

...........

Gee, you mean all those polls I saw on my computer that showed Kerry ahead that had CNN on them were not posted on CNN's website?

Well, who's website was I looking at that had CNN all over it?

[sabra]

14. Exactly! WTF?

[RaulVB]

12. Kick (n/t)