Shrub's Margin Shrank in the Gallup Poll Even though they OVERSAMPLED MORE

[liberalpragmatist]

From the Left Coaster: http://www.theleftcoaster.com/

1. Look at the previous survey:

Bush 52, Kerry 44

Reps (38%), Dems (35%), Ind. (27%).



2. NEW poll: Bush 52, Kerry 46

Reps (39%), Dems (34%), Ind. (27%)


So, basically, the sample is EVEN MORE Republican (+5 v. +3) and Bush's margin still WENT DOWN.

[The Velveteen Ocelot]

1. Could they possibly be bigger whores?

Why don't they just drop their bullshit pretense of objectivity, sample all Republicans and say everybody in the whole damn country is voting for Little George?

[patrice]

2. How do they get away with that sampling error?

That's research 101. What's the reasoning behind the difference?

[gulfcoastliberal]

4. I heard somewhere they justify it by saying it's accounting for

voter fraud. Not sure if that's true or not.

[patrice]

6. Oh I get it . . . Like you have to include extra Republicans to

make up for those who really are Democrats but say they are Republicans because that's more socially acceptable and economically beneficial?

[ThePhilosopher04]

8. No, what they CLAIM to do...

is take a random sample of the public, poll their preference for President, and after the fact, ask what their party affiliation is. Gallup claims that more people are identifying themselves as Republicans among their random samples.

Personally, I think they're full of shit but that's the answer they give.

[George W. Dunce]

10. Yep when we get out the vote it's

Voter fraud, when Chimpy and his brother disenfranchise thousands of people their basic right to vote it's an "irregularity".

[whistle]

3. Maybe Gallup needs to oversample more often, it seems to give....

...more accurate results. They really are what off in right field with their stuff. Why don't they just take a huge sample from each segment to yeild a 99% confidence level with a +/- 1% error instead of wasting everyones time with their bogus results.

[high density]

5. When are these pollsters going to be held accountable?

Most of them were all wrong regarding the popular vote in 2000 and once again this year they're all over the map.

[DeepModem Mom]

7. kick

[radfringe]

9. kid polls mixed

http://www.post-gazette.com/pg/04300/401766.stm

Nationwide presidential election polls of school children are by no means scientific, but their organizers claim they usually are right.

This time, somebody has to be wrong.

Nickelodeon's underage voters made Sen. John F. Kerry the clear winner, but the mock elections by Weekly Reader and Channel One News gave President Bush a decisive victory.

Kerry received 57 percent of the vote in the Nickelodeon poll, but Bush received 65 percent from Weekly Reader and 55 percent from Channel One News.

Combined, the organizations recorded more than 2 million votes, most of them online.

Channel One News says it predicted the last presidential election right. Nickelodeon is claiming total success since it started the poll in 1988.