What If Gallup Flipped Their Poll Results Artificially?

[berni_mccoy]

10 point, 1-day swings.

20 point 2-day demographic swings (white-women 'flocking' to McCain).

These numbers are highly suspect given that little changed in this race. The polls were extremely consistent all the way through and even past the Convention. Then, suddenly, in one day on the weekend, the race changes? Unlikely.

I suspect that there has been some major number switching, if not downright fabrication. All other national polls showed tightening, or even a slight, statistically insignificant bump for McCain, but not a complete flip of the race.

Discuss.

[Alter Ego]

1. Well, Obama's campaign did call the Gallup Daily the worst thing

to happen to journalism.

[EstimatedProphet]

2. They probably don't even have to do that

If we're going to assume that they intentionally flipped results, then we assume they are at the least biased. If that's the case, then they can bias a poll simply by manipulating the questions or the selection of participants, and it would have the same effect but would be much harder to prove.

But this of course assumes they are biased.

[maseman]

3. No way

Gallup would never do that. They have a longtime credibility and would never through that away. I know we don't like what we are seeing in the polls but we cannot blame the polling companies. I work in research and know that Gallup wouldn't "flip" numbers.

[berni_mccoy]

4. If they were concerned about their credibility, they would release their party-id weighting numbers.

Gallup does indeed modify their numbers. They just don't tell us how they do it.

At least Rasmussen publishes his party weighting numbers.

[tritsofme]

11. Gallup doesn't weight for party ID.

And they wrote extensively about in their blog back in 2004.

I'll see if I can find the posts later today.

[berni_mccoy]

12. They are certainly not presenting the raw data.

They could present the daily numbers as well instead of only presenting the 3-day average.

[tritsofme]

13. This information is available to those who pay for it.

At least it was 2002-2004 when I carried a subscription.

[berni_mccoy]

14. HA! GALLUP BUSTED! THEY WERE!!!!

http://www.huffingtonpost.com/2008/09/09/poll-madness-mccain-takes_n_125158.html

[tritsofme]

16. Same old crap I have been hearing and disagreeing with for five years.

Nothing new...they're about about four years behind the Leftcoaster in "busting" Gallup for something they do not hide.

[berni_mccoy]

17. You just said above that Gallup DOESN'T weight by Party ID. WTF?

[tritsofme]

18. Gallup doesn't weight by party ID, and doesn't claim to.

This is a response from Gallup I found in my notes from 2004:

Some consumers of the polls (including the Gallup Poll) have questioned poll results because party identification and other characteristics do not match the 2000 exit poll data. There are very good reasons why they may not match the exit poll data.

First, some treat the exit poll as a census. It is not a census, it is a survey based on sampling of voting precincts. There is a reason it is called an "exit poll" and not an "exit census." That's because it is a poll, and as such is subject to sampling variation and other polling errors just as any other poll would be. In fact, because exit polling relies on quota sampling (hand selection of survey spots according to population size and other factors), it has a higher degree of potential error than do the random samples on which telephone surveys are based. Thus, there is no basis on which to believe the exit poll numbers are in any way more accurate than any other number you get from a poll. They are all estimates. The one advantage of the exit poll is that they know everyone they interview is a voter, while pre-election polls rely on models to determine who is likely to vote and who is not. However, that does not mean their estimates are necessarily better, and they are definitely not error-free estimates of the electorate as many treat them. In fact, when multiple exit polls existed in the past, they very routinely differed in their estimates of the vote as well as their estimates of the demographic characteristics of the electorate. Even today the Los Angeles Times exit poll differs from the larger exit poll used by the networks.

Second, the exit poll measure of political party ID is fundamentally different from ours. We know that survey results can differ depending on how the data are collected. Our questions are read and responses obtained verbally over the phone. Their responses are obtained in self-administered questionnaires that present the questions in a visual format. Most survey research experts would be extremely cautious in comparing data obtained by a telephone interview versus that obtained in a self-administered paper-and-pencil questionnaire. That is in addition to question wording differences in the party ID question that can also have an effect on the results.

Third, a lot has changed since 2000. In the post-9/11 environment, terrorism has become one of the chief problems for government to deal with. The Republican Party has a large perceptual advantage on the terrorism issue. To assume that everything is as it was four years ago is a very risky assumption. While it is possible that in the end things could change once again so that partisanship looks much like it did in 2000, that is by no means certain or even likely.
http://www.gallup.com/election2004/BLOG /

[Teaser]

6. No, but here's what Gallup does:

they use a likely voter model that is very sensitive to changes in enthusiasm. So a small change in enthusiasm can radically change the likely voter screen and cause the results to swing wildly.

This has been noted in Gallup's polling since 1996. Why do they do it? One suspects because it creates the illusion of a tight race, with wild gyrations and opinions constantly shifting. Which encourages polling units from private news organizations to contract with Gallup for more frequent polls.

The wild shifts in Gallup's numbers AREN'T indicative of bias. They are a business model.

[lurky]

7. Pollsters are known to bias commissioned polls

to suit the agenda of their client (right or left).

Here is a link to a great interview that discusses all these issues.
http://onthemedia.org/transcripts/2008/09/05/02

[fasttense]

9. You forgot your sarcasm thingy

Cause that is just what they did in the 2004 election.

"As voters, we need to ask ourselves “why Gallup would use a methodology that would almost definitely mean that their election predictions would be wrong?” These writers are somewhat baffled in answering that question. Why would Gallup want to give the false impression of a Bush lead?

It is interesting to note that James Clifton who bought the Gallup organization is a big Republican donor. He gave thousands to Right Wing Republican Georgia Senate candidate Herman Cain. (See http://www.opensecrets.org) Cain ran as a huge backer of cutting taxes for the wealthiest Americans. This is essentially the same tax position supported with vigor by the Bush-Cheney ticket."

http://www.leftbusinessobserver.com/Gallup.html

If it was good enough for the bushes in 2004 it's good enough for the creepy McInsane in 2008.

[Uben]

5. Polling companies always cheat

...because it is to their advantage. They don't necessarily report false information, they just manipulate the questions to get the desired result. Ever wonder why elections always seems so close? They make money from subscribers, and the closer the race, the more subscriptions they get! It's called capitalism.....the best, and worst part of American democracy!

[DemocratSinceBirth]

8. I Have Been Polled

It's hard to manipulate the following question.

"If the election were held tomorrow who would you vote for; Barack Obama or John McCain?"

I guess they could influence the answer by changing the inflection when pronouncing the names but that seems far fetched...

[tritsofme]

10. These numbers are coming from Gallup's likely voter model.

Gallup's LV is very good, however the numbers really aren't valid until right before the election, as that is the only way to determine who is very likely to vote on election day.

Gallup's RV numbers seem to be in line with the other pollsters. The big change in the LV numbers is just noise from the convention.

[berni_mccoy]

15. GALLUP BUSTED. Check This Out!

http://www.huffingtonpost.com/2008/09/09/poll-madness-mccain-takes_n_125158.html

[BzaDem]

19. Oh not this again. n/t