Prominent Statisticians Refute 'Explanation' of 2004 U.S. Exit Poll Discrepancies

[AlricTheMad]

Actually it doesn't matter where the error lies. If the pols are in error, only a good scientific review can figure out why. If the election results are in error, same thing. It's probably somewhere in the middle.
The largest problem here is, there is no way to verify much of the results. Since there is no paper trail for votes in states like Maryland, the only audit is to re-run the output of a machine.
People didn't like paper because of the designs which created some confusion. Also with paper punches they didn't always punch cleanly causing difficulty in re-counts and the machine readers getting good counts. But at least you could do some counting by hand.

I imagine if the opposite had happened those screaming "Get over it' would be calling for the examination of the data too.

Pat

 

[Train]

Umm, maybe its the exit polls that need to be reviewed, you know, the ones run by solicitors on a sidewalk? Who verifies what that guy does or does not write on his clipboard.? You somehow think those are more accurate than a ballot?

Maybe the smart people just walked right by them. knowing full well the exit poll is not what coints. Democrats tend to be more outspoken about thier politics anyways.

I for one told the pollster to take a hike when I walked out of the poll.

I've heard a few democrats say the conspiracy theories ahve gone so far as to hurt thr democrats credibility with more serious issues. I'm starting to agree with them.

 

[Votingisanillusion]

In Colorado, the exit polls were spot on for the Senate race. Only Bush's results shifted.
http://www.exitpollz.org/pages/colorado.html
It should be noted that it was the same respondents who were interviewed first on how they had voted in the presidential and then in the senatorial election.
So what is the explanation for the much less accurate exit poll for the President?

 

[MadRat]

If a full court press is put on I bet someone cracks.