Posted on 10/25/2008 5:02:33 AM PDT by HalfFull
WASHINGTON, DC: On presidential election day, November 4, all eyes will be on what exit polls say voters did, and analysts fear they could give a rosier view of support for Democrat Barack Obama than reality.
The combination of the exclusion of early voting, and the greater tendency of young voters--more often Obama supporters--to take part in the surveys, could skew the results of the exit polls, which, taken just as voters leave polling booths, are the closely-watched first indicators of how the election is going.
In a National Journal article in March, Mark Blumenthal of pollster.com stressed that overestimation of Obama support was already seen in Democratic primaries earlier in the year.
In 18 of 20 states he reviewed, votes in favour of Obama, who at that time was duelling Hillary Clinton for the party's White House nomination, were overestimated by an average of seven points.
The gap was likely due to the fact that many of Obama's most fervent supporters, often younger, more politically active and in many cases better educated than Clinton's, were more inclined to respond to exit polls.
The same thing could happen on November 4 if Obama voters turn out to be markedly younger than those of his Republican rival John McCain.
According to a recent Gallup poll, 65 percent of voters aged 18-29 plan to vote for Obama, against 31 percent for McCain.
Meanwhile, pollsters themselves tend to be on the younger side.
"One problem (with polls) has been that interviewers tend to be young people, and older voters are less likely to respond to young pollsters--if most of the interviewers this year are again young people, that might lead to some overstatement of the Obama vote," said Herb Weisberg, a professor of political science at Ohio State University.
Weisberg also stressed the unknown factor: ballots cast before November 4--which researchers say could total as much as one-third of the national total.
"Another problem will be that many votes are being cast absentee (early) this year. Exit polls can't pick up these votes," Weisberg said.
"Some pollsters will undoubtedly do phone surveys the weekend before the election to ask people if they've already voted and how they've voted, as a means of trying to capture this effect."
Exit polls have been conducted since 2003 by the National Election Pool (NEP), a consortium of ABC News, Associated Press, CBS News, CNN, Fox News and NBC News, which uses the Edison media research and Mitofsky International to carry out the post-voting voter questioning across the country.
On November 4, the Edison/Mitofsky team plans to carry out more than 100,000 interviews of voters at more than 1,000 US polling stations.
Concerns about the accuracy of this particular type of polling are nothing new.
In the 2004 presidential election, early exit polls gave a slight advantage to Democrat John Kerry, prompting some premature celebrating by Democrats and hand-wringing by Republicans, only to have George W. Bush eventually win the White House.
"People ought to be very skeptical of leaked (exit poll) results they might see between 5:00 pm Eastern Time (2200 GMT) and when the polls close on November 4," Blumenthal said. (AFP)
bmflr
Maybe some of the MSM restrained themselves in 2004. But there is no law prohibiting calling elections early. No way the MSM will follow that restraint this year...as we have clearly seen with MSM shenanigans of the last couple of months.
No doubt they already have the numbers they will be using and Biden is chewing at the bit to leak them ;-)
Would love to see his extravagant stage turned into a flaming funeral pyre of his campaign!
I bet if any of you do, you lie through your teeth just to drive them nuts.”
Last month I was poled about Oregon politics. I was hesitant to participate as most questions are biased and written to ascertain their desired conclusions. Of course they wanted to know what race I was and I told them I was American. He was not satisfied and eventually I lied to him and told him I was black. I pretty much lied from that point on and messed with him.
Nader and McKinney could siphon enough votes from Obama in some states to make it difficult for him to win them!
That first firing can be a good motivator for some to get their acts together...I hope it helped some of the ones your firm fired!
I’m dying to tell them I’m Cherokee and I don’t like the Dem’s because of what Andrew Jackson did to my people in the 1830’s...
How would selection bias not dwarf everything else the polls are trying to measure?
Obama won't lose on November 4. If there's even one state that goes for him, I wouldn't count on him not to seek to have all of McCain's Electors disqualified or eliminated. I don't think he'd succeed, but I'd still be wary of the possibility.
If they weren't that far off, they how come they called it for Gore rather than holding off and saying too close to call?
People lie all the time especially in surveys and polls.
Rxit poll data is worthless and useless. It’s never accurate.
Furthermore, the thing I hate most though are those network news stations who always call states prematurely for a candidate instead of waiting for all the votes to come in [100%]. That really curdles my guts. And that’s what really happened in 2000. It wasn’t Bush’s fault. It was the damn network news assholes that called Florida way too early.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.