Kelzie said:
Your link didn't work. And the exit polls were off by like,
2%. And that's within the margin of error. I've already said that.
Here is the whole article.......It disagrees with your assesment:
Posted 11/3/2004 11:26 PM
Predictions burn pollsters, pundits — again
By Mark Memmott, USA TODAY
The polling industry is on the defensive for its surveys of voters on Election Day, which some TV pundits relied on Tuesday to overstate support for Sen. John Kerry.
And the USA TODAY/CNN/Gallup Poll is catching flak for pre-election surveys in battleground states that turned out to be off the mark.
Polling errors are important, critics say, because they create false impressions about who's ahead or who's behind in a campaign, and that may affect voters' thinking.
Pollsters defend their work, say that their findings are often misunderstood, and question whether many voters were affected by any mistakes.
Both sides agree that, despite heavy investments in money and technology, there's more criticism than ever being aimed at pollsters.
"Let's be honest: They spent four years and a lot of money and were just as bad with their 'exit polls' (of voters Tuesday) as they were in 2000," says Larry Sabato, director of the University of Virginia Center for Politics. "It's disgraceful how bad the information was."
The controversies involve:
• Exit polls. Since 1990, television networks and the Associated Press — known as "the consortium" — have worked together to count votes on election days and to survey voters. Those surveys, known as "exit polls" because they're conducted as voters leave polling places, are designed to give early clues about how candidates are faring and about which issues are driving voters' decisions.
Only the five TV networks (ABC, CBS, CNN, Fox News and NBC), the AP and other news organizations that pay for the information are supposed to have access to the data. (USA TODAY is among the media organizations that pay for some of the information.)
In 2000, problems with exit polls in Florida contributed to a humiliating Election Night for networks. They awarded the state to Al Gore, then took it away, then called the state and election for George W. Bush, then took that back.
In 2002, the exit-poll computer system crashed during the mid-term elections, leading the consortium to disband the operation and start over. This year, a beefed-up computer system and the expertise of two new polling firms, Edison Media Research and Mitofsky International, were supposed to produce a smoothly run operation.
Tuesday's problems were partly out of the control of the National Election Pool, as the consortium is formally known. As NEP began releasing "early waves" of its data, for instance, much of the information was leaked to Internet commentators known as "bloggers." By early afternoon, the material was posted on Web sites such as Slate and the Drudge Report.
Soon after that, commentators on Fox News, CNN and MSNBC were dropping hints that Kerry was doing well. It was clear they were relying on the exit polls for support.
While the information colored TV reports, newspaper editors say it didn't influence their coverage of the election.
"We had prepared for most scenarios," says Sandra Mims Rowe, editor at The Oregonian in Portland. "The early exit polls may have colored our expectations a bit, but they didn't change our plans."
"We looked at everything through the prism of what happened here in Florida in 2000," says Tom Fiedler, executive editor at The Miami Herald. "We just couldn't let exit polls push us after something like that."
Bill Wheatley, a vice president at NBC as well as a spokesman for the consortium, says the NEP has always said early waves of information "are, at best, incomplete pictures" of what might be happening on any Election Day.
The data are not meant to be reported and are meant only to help media users plan their coverage, Wheatley says. The results are not fully adjusted, for instance, to reflect the overall makeup of the day's voting population. That's only done much later as voting results start to be reported.
Jack Shafer, the Slate columnist who posted the early exit poll information, agrees with Wheatley that the data is rough. "I try to put lots of 'don't drive drunk' warnings with it," he says. But he posts the material, he says, because his readers should have access to the same information as TV pundits.
Sabato cuts the consortium no slack. He says, for example, that the early waves included far too many female voters to be reliable indicators of anything. "Anyone hearing about it was getting false information about what was happening," he says. The early exit polls also may have included too many rabid Kerry supporters, who might have been more motivated to get to polling places early than Bush supporters were.
Wheatley says the consortium stands by its work. "In the end, we got it about right," he says. What the NEP may do, he says, is further restrict the number of people who get access to the early data in an effort to keep it from leaking.
POLLPOURRI
President Bush defeated Sen. John Kerry 51%-48% in the national popular vote. Of 10 major polls of "likely voters" taken in the week before the election, seven showed Bush leading:
Poll Bush Kerry
Marist 49% 50%
Harris 49% 48%
FOX News 46% 48%
Reuters/Zogby 48% 47%
USA TODAY/CNN/Gallup 49% 49%
NBC/Wall Street Journal 48% 47%
ABC/Washington Post 49% 48%
CBS/New York Times 49% 46%
Pew Research 51% 48%
Newsweek 50% 44%
Source: RealClearPolitics.com
• Pre-election polls. For much of the year, Kerry supporters criticized the USA TODAY/CNN/Gallup Poll for allegedly including too many Republicans in its sample and for excluding too many young people, who might support Kerry, from its pool of likely voters. Critics said those factors were why the poll showed Bush leading for most of the time since August.
The final USA TODAY/CNN/Gallup Poll before the election, though, slightly understated Bush's strength. The poll, conducted last weekend, showed the candidates dead even. As of Wednesday, it appeared Bush had won the popular vote by 3 percentage points.
Two polls were on the mark. Both the CBS/New York Times and Pew Research surveys last weekend showed a 3-point Bush lead.
More at odds with the results were the USA TODAY/CNN/Gallup Poll's final surveys of key states:
• Kerry was said to be leading in Florida by 3 percentage points. He lost there by 5 points.
• Bush was said to be leading in Pennsylvania by 4 percentage points. He lost by 2.
• Kerry was said to be leading Ohio by 4 points. He lost by 2.
Frank Newport, editor in chief of the Gallup Poll, says "battleground states are the most difficult challenge that pollsters face" because voters there are bombarded by campaign ads and other efforts that push and pull their allegiances. But, he says, "we would like to be just as accurate with the battleground polls as we are with the national poll. We will carefully analyze and look at what happened."
Newport says the polling firm will also examine its controversial decision to allocate most of the undecided vote to Kerry.
Jim Norman, USA TODAY's polling editor, says the newspaper "is disappointed and wants to learn" why the polls were so far off. He expects to join Gallup in meetings to review the polls' methodology.
