PDA

View Full Version : debunking the validity of Gallup polls...




fourameuphoria
09-27-2007, 03:28 PM
I was reading the comments section atthis site (http://zealfortruth.org/2007/09/internet-buzz-is-irrelevant/), and was thinking, just to muster our talking points, we should use this thread to compile all the incidents we've had of how unreliable the phone polls were. Since we've already nailed fundraising, straw polls, meetup numbers, rallies, etc. all that the neocon bloggers have left are a) oh, he's low in the gallup polls b) just plain making up stories (ferrygate). We already know that the landline polls don't account for ex-Dems, independents, people without cellphones, etc. but I think it'd be definately good to have a compilation of what happens during the *actual* phone calls.

Some details from the link I posted:

"As a registered Republican since 1978 when I first became eligible to vote, I have participated in many phone polls in the past. I have been called 3 times since June 2008, and in 2 of those polls, Dr Paul was not a choice I was given to pick from.

In the 3rd call that was identified as being from the RNC I was asked who I supported from the following list of Candidates, I did not wait for her to list them and stated that Ron Paul was my choice. I was then subjected to a brief pause of dead air, a terse thank you and a dial tone as they hung up."

"Kevin, “adjusting for demographics” doesn’t make a scientific poll. A randomly selected sample and an equal presentation of the choices is what makes a scientific poll. Calling only land lines is a form of sample bias. Another is the way the selections are presented. Burying the Paul selection under layers of menus, or not presenting him as a selection at all, will result in the data reinforcing whatever foregone conclusion was behind the choice to bury his selection in the first place."

And I believe in another phone poll, the candidates were listed in "tier" order, rather than alphabetical.