"Out of kilter" Gallup hints at adjustments to likely voter methodology ~ Chicago Tribune

[mzmolly]

Snip ...

As of Friday afternoon, Gallup's daily tracking poll of likely voters had Romney leading Obama by six percentage points, 51 percent to 45 percent.

The Reuters/Ipsos tracking poll, taken from a sample online, had Obama leading by three points on Friday and for much of this week. A Public Policy Polling daily survey had Obama leading by one point, Rand put him ahead by three points and Rasmussen showed the two candidates to be tied.

"Firms don't like being outliers - it causes a lot of self-doubt," said Harvard University political science professor Stephen Ansolabehere. "It's OK if you're wrong if everybody else is wrong with you. It's not OK to be wrong if you're all alone."

... Questions about the gap between Gallup's findings and those of other pollsters is the latest fuss this election season over polling methodology as partisan passions come to a boil in the heated final weeks before the November 6 presidential contest.

With a record of correctly predicting all but three of the 19 presidential races stretching back to 1936, Gallup is one of the most prestigious names in the business and its outlier status has other polling experts scratching their heads.

"They're just so out of kilter at the moment," said Simon Jackman, a Stanford University political science professor and author of a book on polling. "Either they're doing something really wacky or the other 18 pollsters out there are colluding, or something."

Gallup's editor in chief, Frank Newport, said he didn't know why his results didn't line up with others. Nor did he seem unnerved by the disparity.

"We try to keep our eyes on the boat and do the best job possible," he said. "We're going over some additional tweaks with our methodologists to make sure we're on top of it."

... / Snip

More from: The Chicago Tribune

Also of interest, Gallup's LV model was off in the 2010 congressional race, but their RV data was close to the actual result. See what (Alan Abramowitz - Alben W. Barkley Professor of Political Science, Emory University) wrote at Huffpo today on this subject. > Is Gallup Heading for Another Big Miss?

[Democratopia]

1. Maybe gallup should stop asking 80% white folk and ask an accurate cross-section of society.

Ditto LGBT. Gallup puts the LGBT figure at 3.4%. That is way off. If Gallup's own people say they don't have a clue why their polls are so far out, then maybe they need to get somebody who has.

[Baitball Blogger]

2. Did the Chicago Tribune get the scoop on this one?

Small wonder.

[shraby]

3. He said they "try to keep our eyes on the boat"...problem is they are in the fog...

and boaters can tell you when in a fog, you tend to steer in a circle. Ptuii on his polls.

[Lex]

5. I've always heard "keep your eyes on the ball" not the boat . . .

maybe the guy is off in the fog.

Okay there is a saying "keep your eyes IN the boat" not ON, according to urban dictionary.

[progressivebydesign]

4. Once again The polling firms are fucking up the "likely voter" thing.

The people you've been seeing in the early voting pics are NOT "likely voters" according to the major polling firms. They oversample white, middle-aged, moderately affluent, conservative-leaning, women as "likely voters." And the demographics are changing, especially with more motivated young and minority voters. This is the issue here.

Everyone is freaking out, but the polling firms switched to reporting "likely voters" a few weeks ago, after having "registered voters" or other less curated samples. They start a poll with 1,200 people and dump around 400+ to get the "liklies."

And Yahoo News, the Fox Junior outlet this year, has a HUGE article saying that according to Gallup, Romney would win today. As the other polls show a "dead heat."

Well first of all, bullshit. We do NOT elect presidents based on popular votes. and Second... likely voters, again the polls are skewed, and not just Gallup. Do people KNOW that we don't elect by popular vote??? For fuck's sake the WRITER at YAHOO should know that!!! Romney would NOT win today, even if Gallup had him at 6% up in the popular vote, because he's NOT going to win the swing states.

[Blue Idaho]

6. Nothing would make me happier than a lopsided win for the President.

Not only for the obvious reason - but for the huge embarrassment it would cause our so called news media and crooked pollsters.

[ProSense]

7. "Firms don't like being outliers - it causes a lot of self-doubt"

"Gallup's editor in chief, Frank Newport, said he didn't know why his results didn't line up with others."

Yet some people are trucking along and pretending that they don't see the clownish results as unusual.


IBD/TIPP Obama +3
PPP Obama +2
RAND Obama +3
Reuters/Ipsos Obama +3

Gallup Romney +6
Rasmussen Romney +1

[NCLefty]

8. Polls are so unreliable

[boingboinh]

9. What's Humorous is that MOST MSM report the Gallup outlier and exclude the others

If u want proof they want a horserace that is it. Sure u can report Gallup but how about report other, more new releases that show Obama is ahead?

[Rozlee]

10. I read that Gallup Organization has been under investigation by the

Justice Department for several months for suspicion of millions of dollars in fraud in their contracts with the government. I wonder if they've colluded with Republicans in their polls to discouraged turnout by Dems in hopes that a Romney Administration would drop charges against them. Meh, now I'm sounding as conspiracist as they do. But, they did steal the vote in Florida in 2000...in Ohio in 2004...and they've been doing their damnedest to steal and suppress it this year.

[Midwestern Democrat]

11. The fact that Gallup's final poll has been correct 16 out of 19 times doesn't at all convince me

that they're not the outlier. In most elections, it's not that difficult to predict straight up who the winner is going to be on election eve. There have really only been four elections in that whole time period where there was real suspense about who the victor was going to be on the eve of the election and only one election where the perceived favorite lost. Plus, even in the four squeaker elections there was a general consensus of who was the slight favorite and it turned out to be correct each time.

Elections where the outcome on election eve was not in doubt (13)- 1936/1940/1944/1952/1956/1964/1972/1980/1984/1988/1992/1996/2008

Elections where the outcome was a true coinflip (4): 1960/1968/1976/2000

Elections where the general consensus on the outcome was wrong (1): 1948

Elections with a slight but definite favorite on election eve (1): 2004

[mzmolly]

12. Gallup adjusts toward the end of the race it seems?

Excellent points!