How do you negotiate with crazy people?

[andrewducker]

• 67 percent of Republicans (and 40 percent of Americans overall) believe that Obama is a socialist.
• 57 percent of Republicans (32 percent overall) believe that Obama is a Muslim
• 45 percent of Republicans (25 percent overall) agree with the Birthers in their belief that Obama was "not born in the United States and so is not eligible to be president"
• 38 percent of Republicans (20 percent overall) say that Obama is "doing many of the things that Hitler did"
• Scariest of all, 24 percent of Republicans (14 percent overall) say that Obama "may be the Antichrist."

From

I mean, I know a lot of, say, Conservatives in the UK have beliefs I don't agree with.  But the vast majority of them, so far as I can tell, just have different experiences to me, and different opinions about how things should be organised.  They don't believe that the leader of the oppositon is the fucking antichrist, or other things that can be disproved by 30 seconds with Google.

Comments

[matgb]

I just went to look; 2K+ people surveyed by Harris, who I respect enough.

I'll look at the proper methodology when I've time, but odds are it's wrong because it's taking the numbers from registered Repubs, which isn't the same thing as those who voted repub.

[wildeabandon.livejournal.com]

Ah - here's some criticism of it.

[matgb]

Criticism, yes, and some of it is valid, however he says

I’ll lay off the sampling, though this survey was done among people who sign up to click through questionnaires via the Internet in exchange for points redeemable for cash and gifts – not a probability sample. Been there before.

Which, according to HArris themselves, isn't a fair assesment

Respondents for this survey were selected from among those who have agreed to participate in Harris Interactive surveys. The data have been weighted to reflect the composition of the adult population.

I studied polling methodology, and while I have a problem with weighted samples, that's what virtually all UK poll companies do these days, and Harris switched from face to face, to telephone and now to internet panel surveys.

Their methodology is very close to YouGov's over here from what I can see, and my only problem with YouGov currently is that they're under-weighting for Lib Dem identified voters compared to other weightings (and I'm talking about that with their stats guy currently as he disagrees).

Some of the rest of his points (biased questioning, etc) are valid and true, but they add weight to existing beliefs, they don't create them out of nowhere.

There is in a lot of media a desire to go against internet polling generally, and the linked article in the bit of his I quoted is both incredibly snobby (and typical), and lumping in the type of unsampled internet polls that do offer rewards with no real methodology with much more reputable and effective methodology.

Panel polling, where you invite people to a specific poll from a much larger group of volunteers, and weight the invitees, can be very effective.

ABC pays money to traditional, expensive, face-to-face pollsters. They don't report internet polls, even from Harris, because doing so is to acknowledge that their expensive polls might be just a waste of money.

Yes, the poll is flawed, but it's nowwhere near as flawed as that guy makes out (it is, however, commissioned by an author trying to plug a book).