MADISON, Wis. — Presidential hopeful Barack Obama had a double-digit lead in the polls Monday, but on Tuesday, Hillary Clinton came up with a big victory in the New Hampshire Democratic primary.
In this case, the polls didn’t only project the wrong winner, but were off the mark by a significant margin. Now some political observers are saying that even considering a margin of error, the outcome needs more examination.
The University of Wisconsin Survey Center conducts polls on a number of topics, but the center’s accuracy in statewide Badger Polls is usually much better than the predicted results in New Hampshire.
“If the sample is pretty representative, we’ve got a pretty good estimate of what we think all 5 and a half million people in Wisconsin think about an issue or whether they favor or oppose it. But there’s always an error,” said Chad Kniss, project director at the UW Survey Center.
But University of Wisconsin-Madison professor Charles Franklin said this time was different.
“We ought to see quite a few polls 5 points of the outcome. That is a huge error and it’s certainly not sampling error,” Franklin said. “The big question that’s lingering today is, was that because of a remarkable change in politics that Clinton’s supporters came home to her and turned out heavily or did it reflect something wrong about the polling?”