Manipulating Polling Data(figures lie and liars figure)

Accuracy in Media ^ | August 23, 2004 | Cliff Kincaid

[AmericanMade1776]

In a story about how George Bush, not John Kerry, got a "bounce" from coverage of the Democratic National convention, Susan Page disclosed something significant about a poll done on the matter. She revealed, "Because the results were a surprise, USA TODAY extended the survey an additional night, to Sunday, to create a larger and more reliable sample." In other words, because the media feared the results would undermine Kerry, they did their best to find more people for their "survey" who would validate the media's assumption that Kerry would get the bounce. But it was not to be. They couldn't find those Kerry supporters.

Despite the blatant manipulation of the polling data, Bush came out the winner. But the admission by Page demonstrates how these polls could possibly be rigged to produce the desired result. It shows that when there is a "surprise" in the polling data, the pollsters always have the option of extending the survey by polling more people over a longer period of time to produce different results.

[wideawake]

the pollsters always have the option of extending the survey by polling more people over a longer period of time to produce different results

Or changing the questions to massage the results. Or asking 40 questions and only reporting the data from the three whose answers best match their agenda.

Etc.

[Jeff Gordon]

If they are attempting to bias the poll by extending it, they are wasting their time. Once they have a statistically accurate sample, extending the poll only will change the results in the decimal point range.

[SPRINK]

This is a non story.. Everybody thought that Kerry would get a significant bounce and when the data showed other wise they wanted to make sure. It was the logical thing to do.

[Juan Medén]

The publishing of public polls should be banned six months prior to an election. They are manipulable and serve only to sway public opinion.

[Chewbacco]

Yes, but they can also change the demographic, by polling more in D.C. and California than in Alabama or Texas and then claim it as a national poll.

Remember Mondale, Dukakas, Carter, etc...
The Alphabet polls always said up to the day of the election that the result was 50/50 or with the Dem leading.

The fact that the Alphabets are reporting 50/50 most likey means that President Bush is way out in the lead, too far for the pollers to fix. (imo)

Also, I have always suspected that there are 2 results, the one that they supply to Kerry's campaign (i.e., the "real" result) and the results that they publish in the paper for propaganda. It almost has to be like this because they are so consistently wrong, Why would anyone pay them?

[IndyMac]

"Once they have a statistically accurate sample, extending the poll only will change the results in the decimal point range."

Not necessarily, Jeff. I've heard and read several times that polling results can be significantly influenced by the time and day (weekend vs. weekday) that the data is gathered. I don't know that this is true, but it does make some sense that the likelihood of a particular person answering that their phone at a given time could be affected by demographics.

[NationSoConceived]

One should note in relation to the claim that 760-odd respondents is insufficient to form an accurate sample is not necessarily true. If a sampling is truly random, as you approach 1000 respondents, the margin of error is significantly reduced.

In short, IF the polling organizations ask a sufficiently large number of truly RANDOMLY selected people, the sample is failrly accurate.

For those of you who want to understand polls a lot more throughly, an excellent quick introcuction to sampling statistics can be found here: http://www.isixsigma.com/library/content/c040607a.asp

[DoctorZIn]

This article has been posted to DoctorZin’s New News Blog!

[Publius6961]

POLLS DON'T DECIDE ELECTIONS. VOTES DO!

[LS]

I remind EVERYONE that in 1996, according to a study by a Purdue prof. EVERY SINGLE POLL was off on the final prediction, and every one was off in Clinton's favor! In 2000, it was closer, but more were off in Gore's favor than Bush's. In 2002, all the major pollsters badly missed the 3-4 competitive races (again, in favor of the Dems) and in CA's special election---you guessed it---they BADLY underestimated Schwartzenegger's popularity (by as much as 12-13%, depending on the poll).

The HISTORY of the EVIDENCE is that the polls are notiously UNRELIABLE and usually off in favor of the Dem.