In 2000 which poll ended up most acurate in New Hampshire?

[Quixote1818]

Anyone have any Data?

[windansea]

1. here you go

"Now I know why they call this the Granite State," said candidate Bob Dole eight years ago after losing a Republican primary in the state for the second time. "It's so hard to crack."

Dole's assessment is as true for pollsters as for the candidates in New Hampshire, where bad methodology, bad timing or simply bad luck have produced some of the most memorable miscues in the annals of polling. Consider these flubs:

In 2000, the headline on an AP day-before-the-primary story was "Nearing the N.H. finish line; Polls declare GOP dead heat. . . . " John McCain then went on to beat George W. Bush by 18 percentage points.
http://www.washingtonpost.com/ac2/wp-dyn/A41186-2004Jan23?language=printer

Krishna Krishna, Hare Hare ... Attention, Zogby Crack addicts: WaPo's Morin and Deane contribute a highly useful explanation of why it's dangerous to bank on New Hampshire tracking polls. I'll let the Mystery Pollster blog it:

They hit two timely and important topics:

1) The poor history of NH tracking surveys in predicting the final result, especially in 2000

2) Why real pollsters distrust Zogby. The money quote from Warren Mitofsky: "Zogby is not a reputable pollster...He is more a salesman and a self-promoter than a pollster. He has made lots of mistakes on election outcomes -- five in 2002." ...
http://slate.msn.com/id/2094126/

I think the most important caution for observers this time around: New Hampshire has a history of very rapid change in the last 24 hours or so that tracking polls miss. My memory of 1984 was that the final Monday night tracking poll had Hart and Mondale running dead even. As you'll recall, Hart did a bit better than that.

[democratreformed]

2. More about Zogby:

Link: http://www.moderateindependent.com/

Also, I have read that most polls are running larger margins of error than usual. Anyone else know anything about this and how it affects their reliability?

[SheilaT]

3. Basically the margin of error

is the point spread. So if someone is polling at 20% with a moe of 2, the real number could as easily be 18% or 22%. The higher the moe, the greater the spread.

If the moe is 5%, and one guy polls 20% and the other guy at 25%, they are within the moe and perhaps tied, although it's likely the guy with 25% really has a little more support than the one with 20%, at least at the time the poll was taken.

I remember reading several years ago, before the 2000 election, that reputable pollsters were getting more and more concerned about the reliability of their polling. One reason is that it is harder and harder to get people to actually agree to participate in a poll. Lots of people with caller ID screen calls and won't pick up for a polling organization. More and more people, especially young ones, only have cell phones these days, which are not callable for polling purposes. The result is that pollsters are more and more uncertain if the people they get are truly representative.

When it comes to November 2, 2004, the only reliable polling will be exit polls, but that may well not happen. We need to be very fearful of what's going on, not simply the horrors the Bush Administration is inflicting on the country and the rest of the world.

[ozone_man]

4. Non status quo candidates do better?

Cain vs. Bush, Hart vs. Mondale. Just a hunch, but maybe New Hampshirites vote away from status quo candidates. Maybe Dean will have a surprise factor, as people react to their gut feelings at the polls.