Friday :: Sep 17, 2004

Liars, Damn Liars, and Statisticians


by pessimist

"Anybody who believes these national political polls are giving you facts is a gullible fool. Any editors of newspapers or television news shows who use poll results as a story are beyond gullible. On behalf of the public they profess to serve, they are indolent salesmen of falsehoods." - Jimmy Breslin

Fresh on the heels of two polls which show that George w. Bu$h is tied with John F. Kerry comes this obviously partisan Gallup poll which shows Bu$h ahead by 13 points. The difference? The Gallup poll talked to 'likely voters', defined as

Results based on likely voters are based on the subsample of 767 survey respondents deemed most likely to vote in the November 2004 general election, according to a series of questions measuring current voting intentions and past voting behavior. For results based on the total sample of likely voters, one can say with 95% confidence that the margin of sampling error is 4 percentage points. The likely voter model assumes a turnout of 55% of national adults. The likely voter sample is weighted down to match this assumption.

You can bet your last dinar that this Gallup Poll is a political psyop to counter to lack of a sustained bounce for Bu$h and increased polling numbers for Kerry. Look at this Google search and read the headlines and the teasers. Look at the listing dates and times. This was a rush job, seeking to find something to counter Kerry's advances.

USA Today posts the Gallup Poll Q&A.

As is cited here, polls are volatile things. This article cites a recent Pew poll which came up with widely divergent result in just a couple of days:

The results from the Pew poll are more interesting because they show a massive internal divergence. Taken in two sweeps, one between 8-10 September and the other between 11-14 September, the internal results indicate a massive, overnight shift took place within the electorate -- or point to an undiagnosed problem with the poll itself.

As anyone who has ever conducted a statistical poll knows, one keeps weeding out the responses which don't fall inside the infamous 'bell curve' until one comes up with the results one is after. This has to have been the case for such a wild swing in the poll numbers showing a great increase in favoratism for George W. Bu$h during a week full of news that goes against ALL of the (mis)Administration's talking points - and Kitty Kelley's expose of the Bu$hpranos to boot!

So what kind of results did Pew get in these two sweeps?

In the first sweep, the Bush/Cheney ticket held a 16 point lead (54-38 per cent) over the Kerry/Edwards ticket.

Over the following three days, that lead all but vanished, leaving Mr Bush with a one point lead (47-46 per cent) even though undecideds remained at 6 per cent in both polling sweeps and there was only a one point movement (from 2 per cent in the first sweep to 1 per cent in the second) away from Mr Nader.

Both sweeps were of likely voters -- but, quite obviously, not the same likely voters.

The implication drawn by New Zealand's National Business Review is: "Perhaps they found that pocket of respondents Pew talked to in its first sweep of the country."

Or they were created. As Breslin points out:

... most polls are conducted by landline telephone, he says, and that leaves out a huge swath of the population -- namely, those who can only be reached on their mobile phones. ... there are 169 million mobile phones in America as evidence of just how distorted this could make landline polling ...

James Zogby tends to agree:

Zogby uses direct interviews in his polling, because, when it comes to telephones "there is no easy way to use them." Mr Zogby told Mr Breslin, and Mr Breslin tells the world, that "people who are using telephone surveys are in denial" because the population that can be reached by telephone isn't a valid sample of the actual population in America.

But because people will believe what they hear when it is attached to a well-respected name like Gallup, the desired results can be obtained even though the polling itself is seriously flawed. This makes it easier to create the desired result. Most peole aren't astute enough to know when they are being dazzled by brilliance, or baffled by statistical bullshit.

I repeat: Bu$h has taken some heavy hits this week, and it's time for them to change the course of the discussion away from damaging negative news and toward something 'positive' that Bu$hCo drones can now run around to the nation's media whores and brag about it. You can bet that Rush has taken a double dose today and is ready to rant!

And - it's Friday.

The longer this campaign goes on, the more clear it becomes that Bu$hCo is not above (or beneath) doing anything to win, for they firmly believe that the ends will always justify the means. The sooner we who oppose the BFEE/PNAC Petroelum Pirate Posse recognize this, and act in a manner that will counter these efforts, the better off America - and the entire world - will be.


Copyrighted source material contained in this article is presented under the provisions of Fair Use.

FAIR USE NOTICE

This article contains copyrighted material, the use of which has not always been specifically authorized by the copyright owner. I am making such material available in my efforts to advance understanding of democracy, economic, environmental, human rights, political, scientific, and social justice issues, among others. I believe this constitutes a 'fair use' of any such copyrighted material as provided for in section 107 of the US Copyright Law. In accordance with Title 17 U.S.C. Section 107, the material in this article is distributed without profit for research and educational purposes.

pessimist :: 7:01 AM :: Comments (11) :: Spotlight :: Digg It!