2 new polls sharply divided

[QCassidy352]

http://www.cnn.com/2004/ALLPOLITICS/09/17/presidential.polls/index.html

Kinda weird... one poll has bush with a 13 pt. lead, and the other has them exactly even. And yet both claim to have a margin of error of 3.5-4%... apparently not.

Polls often differ, but this is by so much! I wonder what the explanation is...

[yg17]

hopefully dubya doesnt have that 13 pt lead

[wordmunger]

http://www.cnn.com/2004/ALLPOLITICS/09/17/presidential.polls/index.html

Kinda weird... one poll has bush with a 13 pt. lead, and the other has them exactly even. And yet both claim to have a margin of error of 3.5-4%... apparently not.

Polls often differ, but this is by so much! I wonder what the explanation is...
I think the difference usually comes down to how you define "likely voters." Very often Democrats lead in "overall" polls, then their lead narrows when you ask if responders are registered voeters, and narrows even further, or comes out in favor of Republicans when you try to determine who's actually likely to vote. So the differences most often reflect how the "likely voter" determination is made.

[IJ Reilly]

Which only goes to show once again that Kerry should be spending less time trying to sway the undecided voters and more time trying to get people to understand what is at stake in this election. In the end, it's going to be about turnout.