Survey USA - Big Source of electoral-vote.com...anyone have their details?

[Roland99]

Look at the polls behind some of the states and you'll find Survey USA behind many of the Strong Bush states. In fact, TN now shows a 19% lead for Bush when just a few days ago it was single digits.

A recent major poll showed Colorado as a tie but Survey USA has a large lead for Bush.

Does anyone have a breakdown of the party affiliations of those surveryed by them?

[Roland99]

1. Check out this comparison!

http://www.electoral-vote.com

vs.

http://online.wsj.com/public/resources/documents/info-battleground04.html (Zogby polling)


See how TN is a virtual tie with Zogby but a huge lead for Bush from Survey USA?

[bullimiami]

2. this is an old discussion. his methodology is screwy.

he uses whatever newest poll he finds. it could be zogby or doofus.com he doesnt seem to make any attempt to keep to more reputable polls or methodologies.

its just for fun i guess.

for a couple days he averaged the 3 newest polls. i think it would have stabilized his mad swings, but he got a lot of e-whining and changed it back right away.

[Roland99]

3. I'm looking for details specifically about Survey USA, though

[Warpy]

7. Not much out there on Google

http://www.surveyusa.com/home.html

What little I've been able to glean from various articles is that they're a subsidiary of Gannet Publishing, who bring us McNews in USA Today as well as a host of papers listed at http://www.gannett.com/web/gan013.htm . Their editorial bias leans right.

Their polls are all pro Bush, so I imagine they poll their subscribers.

[Roland99]

10. Hmm...maybe something here:

http://www.surveyusa.com/currentelectionpolls.html

Notice their polling sizes. About 500-600 people. Much smaller sizes than used by the major polling firms.

And, their "clients" are local TV stations (the one in for the TN poll is owned by Gannett so you may be on to something.)

[bullimiami]

8. why worry about survey usa

even gallop has shown to have serious method flaws. there are so many polls and so little reliability i try to give the most confidence in those with a proven track record.

and electoral-vote.com is a total mess.

this one at least is actually scientific. http://www.geocities.com/electionmodel/


they have a website http://www.surveyusa.com/

[TruthIsAll]

9. Thanks for the plug.

tia

[DemFromMem]

4. Interesting poll comparison

A very Republican friend of mine sent me this link:

http://stolenthunder.blogspot.com/2004/09/good-poll-bad-poll.html

This is a GOPish blogger who has done what looks to be surprisingly honest review of polls (he tells which ones are GOP leaning and Dem leaning as well as which ones hide their methodology and which ones don't).

Might be worth one of our polling experts going over this review and coming up with a similar list.

[Roland99]

6. Thanks for the laugh!!

Wow! Very Solid Methodolgy and he lists Survey USA, Gallup and PEW???

Suspect is Zogby and Rasmussen?


LMAO!!!!!

[TruthIsAll]

5. I avoid using them in the election model if at all possible.

I use the latest Zogby and ARG state polls along with the other polls downoaded from electoral-vote.com.

You want a TRUE picture of Kerry's position?
Go here.
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=132x1021240

[mvd]

11. I would, too

Probably not as bad as Strategic Vision, but personally Survey USA looks suspect.

Rasmussen's really bugging me, too. He now has Bush with a 4 point lead nationally when Zogby's been trending Kerry.

[demzilla]

12. And what all these conflicting polls mean

whatever their methodology, is that it is a very close, unpredictable race.