Why Gallup is crap...

[MAlibdem]

Courtesy of Electoral-Vote.com ...

"Exactly what to normalize for is a controversial issue. Should the pollster make sure his poll has the statistically correct number of Catholics, gun owners, retirees, veterans, immigrants, union members, millionaires, welfare mothers, fat people, lesbians, and ? Where do you draw the line? More specifically, should the pollster normalize to make sure the effective number of Democrats and Republicans is correct? Some pollsters do and some do not. And how many is correct? Gallup is currently normalizing to 40% Republicans and 33% Democrats which some pollsters think is highly unrealistic, which explains why Gallup's polls show Bush doing so well.

As we reported earlier, in an extraordinary step, both John Zogby and Scott Rasmussen criticized Time and Newsweek for also having too many Republicans in their samples. Zogby said: "If we look at the three last Presidential elections, the spread was 34% Democrats, 34% Republicans and 33% Independents (in 1992 with Ross Perot in the race); 39% Democrats, 34% Republicans, and 27% Independents in 1996; and 39% Democrats, 35% Republicans and 26% Independents in 2000." Thus a score of Bush 52%, Kerry 40% doesn't necessarily mean of the 1000 pollees, 520 will vote for Bush, 400 will vote for Kerry and 80 will vote for someone else or are undecided. It might mean this is the result of weighting the Republican votes to force them to represent 40% of the sample."

stupid gallup

[sendero]

1. They aren't stupid per se...

... they are simply skewing the results the way they want them to be, and hoping nobody notices their flawed methodology.

[rockchalk]

2. Do you have proof of that?

Or are you just hoping and willing to defame professionals if it suits your bias? How many statistics classes have you taken? A lot of people around here cavalierly dismiss one poll or another as biased without the slightest evidence.

[rogerashton]

3. I have taught statistics and published statistics-based research.

The normalization issue is a real one. I would not impute motivations for choosing those numbers, but the key point is that they seem to be chosen, not estimated, and the Zogby critique looks sound to me.

By the way, Rockchalk, welcome to DU.

I think.

[still_one]

4. You want proof

Gallup Poll, October 26, 2000

Gore 39
bush 52

[Joanne98]

7. Gallup needs to be called on the carpet.

They're openly helping Bush steal the election by screwing the polls to adjust to the future e-voting magic. Maybe we should start and e-mail campaign to the foreign press. If they do steal it there will be alot of letters from Americans saying I Told you so. It would start a global wave of anger and hate for Bush and anybody who voted for him. As Pat Buchcanan would say, "Let the bloodbath begin".

[Jim__]

9. Given the numbers in the opening of the thread, ...

... namely:

Gallup is currently normalizing to 40% Republicans and 33% Democrats which some pollsters think is highly unrealistic, which explains why Gallup's polls show Bush doing so well.

and:

Zogby said: "If we look at the three last Presidential elections, the spread was 34% Democrats, 34% Republicans and 33% Independents (in 1992 with Ross Perot in the race); 39% Democrats, 34% Republicans, and 27% Independents in 1996; and 39% Democrats, 35% Republicans and 26% Independents in 2000." Thus a score of Bush 52%, Kerry 40% doesn't necessarily mean of the 1000 pollees, 520 will vote for Bush, 400 will vote for Kerry and 80 will vote for someone else or are undecided. It might mean this is the result of weighting the Republican votes to force them to represent 40% of the sample."

If that is not proof, what would you consider proof?

And, given the numbers from post 4:

Gallup Poll, October 26, 2000

Gore 39
bush 52


Numbers that are pathetically wrong; and wrong at the time that Zogby was calling the race close, what kind of professional doesn't adjust a procedure that has been demonstrated to be both wrong, and not even close to accurate procedures being used by other professionals?

[sendero]

10. The proof...

... is how badly their results compare to real elections.

And if you need MORE proof, simply compare polls that should have the same numbers within a margin of error that mismatch each other by twice as many points.

Want even MORE PROOF? Then notice how certain pollsters ALWAYS give the Republicans several points lead over every other ostensibly equivalent poll.

You have anything else to say?

[Fla Dem]

6. Unfortunately, self-fulfilling prophecy

My concern is that as undecideds see and hear about these polls (whether or not they are skewed) they may feel that the majority knows best and will vote with the majority. I do feel people are influenced by polls. People unable or unwilling to become familiar with the issues will believe there is something failing in the Kerry campaign therefore, not worthy of their vote. So they will either not vote,or vote for the status quo.

The other concern is that if the polls continue to be lopsided in shrub's favor, Dems will not bother to vote because they will think it's a lost cause.

[AmerDem]

5. Why aren't Dems talking about

These loaded polls. It would make sense to everyone that if you ask more repukes then dems the results will be nonsense. I have personally seen endless times when Dem talking heads were asked "what do you make of these polls". They go on to talk about Kerry's message and campaign but never address this corrupt way of polling. Once again another thing that boggles the mind!

[AnnInLa]

8. The PURPOSE of these outlying polls seems to be

to dissuade voters from actually voting. IF the Gallop, et al are influenced by rove ops (VERY possible), they do not want Dems to vote. They are afraid of the OUTSTANDING job the dem groups are doing to register new voters and to encourage existing voters not to desert Kerry. If newly registered dem voters and discouraged dem voters believe these poll results, they will give up and stay home. I think rove/bu$h is worried about the Dem effort to get out the vote---thus, the repug-influenced outlying polls.