Poll: Romney Tops Obama in NH

April 22nd, 2011 at 12:07 pm | 6 Comments |

| Print

CNN reports:

President Obama is getting some ominous news out of New Hampshire Friday in a new poll that suggests he will lose the key presidential state by a sizable margin if former Massachusetts Gov. Mitt Romney is the Republican nominee.

According to the new survey conducted by Dartmouth College, Romney beats the president in a head-to-head matchup by 8 points, 47 percent to 39 percent. Obama’s poor showing against Romney is also the product of his weak approval rating in the Granite State, where only 36 percent of voters there give him positive marks. That compares to a 45 percent approval rating for Obama in New Hampshire in a similar poll one year ago.

But it’s not all bad news for the president: though losing big to Romney, he soundly defeats every other potential GOP candidate the poll tested, including former Arkansas Gov. Mike Huckabee (by 8 points), former Minnesota Gov. Tim Pawlenty (by 16 points), Mississippi Gov. Haley Barbour (by 19 points), businessman Donald Trump (by 22 points) and former GOP vice presidential nominee Sarah Palin (by 27 points).

Recent Posts by FrumForum News



6 Comments so far ↓

  • Carney

    Much as I’d love for it to be true, this poll is junk, like the vast majority of polls.

    Rather than surveying over 1,000 likely voters, it only surveys 426 registered voters.

    • jakester

      Since so few people show up for a primary, just polling registered voters is much preferable to polling airheads who can’t even be bothered to register. I am sure that about 94% of all unregistered voters now will NOT vote in the primary next year

  • talkradiosucks.com

    “Rather than surveying over 1,000 likely voters, it only surveys 426 registered voters.”

    You’d have no problem with a 1,000-voter survey in a national poll, right? That works out to 3 thousandths of a percentage point. This poll works out to 300 thousandths of a percentage point — 100 times greater.

    This isn’t news. If Romney can’t poll consistently over Obama in New Hampshire, he’s finished.

    • Carney

      As I understand it, the 1,000 strong sample size is not about getting to a certain proportion of respondents vs. actual voters, but rather has to do with getting statistical validity by getting a large enough sample size in order to reduce sampling errors below a certain degree.

  • talkradiosucks.com

    Well, I’m not a statistician but it would seem to make sense that the sample size required for a given level of confidence would be related to the population size. If 1,000 people is reasonable out of over 300 million, presumably 426 out of 1.3 million is as well.

    Regardless, there are no surprises (for me at least) in this poll. Romney not only has local ties, he is exactly the sort of candidate that New England conservatives like. If anything, I’d have expected the margin to be a bit higher.

  • joanna

    Actaully, Carney is right, the random sample size of 1000-1500 is statistically significant for any size population. increasing it won’t change the outcome, that’s why you usually see the number is this bracket.(I used to do it, and also it was a necessary part of my degree).