1. What really happened in the New Hampshire polls
2. Amazon has the best book reviews
3. Thinking with your body, not just your brain
4. Anthony Bourdain’s new blog
5. Martin Feldstein: the economics of the FBI
Assorted links
Previous post: Markets in everything
Next post: All of Beethoven for $125
Second, we’re talking about Democratic primary voters, NOT the general electorate. These people are the least likely to be racist.
The inference here for Replubicans is not flattering.
That first article is utter nonsense. Some of the stuff is just false (HC did not have anything close to a 20-pt lead before Iowa), and the rest is just varyingly implausible speculation. The correspondent in the link appears to be some random person with a statistical background who made some cursory guesses.
I don’t understand why everyone seems to claim that the polls must have been “wrong.” I assume they would have been quite different if taken on election day. Indeed, Zogby said that there was a dramatic shift on Monday night that wasn’t reflected very much in his results, since he uses a several-day tracking poll. Shouldn’t Occam’s Razor suggest that the crying incident/media inanity about said incident led to Hillary’s extremely strong showing with women and very % of late deciders, which means there was an actual shift rather than every single poll (with their different methodologies) failing?
The race thing does not strike me as very plausible on its face (the effect applies to white women in NH and not anyone else?) and is taken care of decently here:
http://www.tnr.com/politics/story.html?id=17dabbce-95ad-40c1-805c-5c12d0158ba8
One thing that always struck me about physical expression aand thought and language is that adjectives expressing scale tend to use expanding mouth movements for ‘large’ words and contracting mouth movements for ‘little’ words. This seems to be true across languages.
In regards to the accuracy of polls, I’ve heard the phrase “It is simply unprecedented for so many polls to have been so wrong” and its ilk quite a bit. Perhaps it’s just cynacism and a deep disbelief in the competence of cable news pollsters, but I always had the impression that polls were wrong a good deal of the time.
I wonder if there are any statistics on the accuracy of polls out there…
It was not just the polls that were done before the election that were wrong, but the exit polls also predicted an Obama win by several percentage points. Someone suggested that working class women are reluctant to talk to pollsters so both the pre election polls and the exit polls under sampled them. Because the Obama support was primarily from up scale voters this produced a much larger error in NH than it usually does.
The argument that the Democrats are less likely to lie about voting for the minority seems wrong to me. To lie about your voting preference implies that you feel guilty about supporting the white candidate over the minority candidate. Now, which do you think would feel guilty about their preference… the person who has a conscious race bias, or one that has a race bias, but is embarrassed by it or tries to explain it away.
thank you for this excelent article
Comments on this entry are closed.