Last month I wrote a satire called Superior Carbon Neutral Humans Discovered. It was based on a BBC article that claimed that an undiscovered tribe was threatened by illegal logging. The joke was that a fictional scientist admires a the tribe's tiny c...

Apparently a high priest of global warming caused by man, has made a new pronouncement....

AKINOLA REVOLTSThe Archbishop of Nigeria is revolting, not only in his practices but against the Archbishop of Canterbury. The American conservative Anglican community, who has chosen to worship Peter Akinola instead of the Holy Trinity of the Fa...

Today we received an entirely new type of spam EMAIL.The gist of the message revolved around a doomsday predicted at the end of 2012. After several paragraphs about voodoo, magic spells, black holes, and ancient prophecies, there was the following......

Even Obama's slogans and campaign slogans bear an uncanny resemblance to McGovern's pledges of a new course and change.

WHEN CONSERVATIVES GO REALLY BAD"...The Holocaust was not a cause of the war, but a consequence of the war. No war, no Holocaust...."Patrick J. BuchananPat Buchanan's latest Townhall column has been quite revealing, proving what I've been saying fo...

But it's just a bad question, isn't it? Because the meaning is open to personal interpretation, it's really a bunch of questions and you have no way of knowing which one any given respondent decided to answer.

Exactly so.

I suspect, incidentally, that the same is true of the much more direct "Are you a feminist" question. I'd probably say No, given the baggage the term has. But I'd say Yes to questions about whether women should get equal pay for truly equal work and so forth.

Of course, the heck of it is, approached that way, a one-third yes rate looks really low. On one level, this suggests that TWO THIRDS of respondents are denying a universally true fact because they think the pollsters want No answers. On the other hand, we have no reason to believe the respondents are trying to fool the pollsters - they could just be fooling themselves.

But it's just a bad question, isn't it? Because the meaning is open to personal interpretation, it's really a bunch of questions and you have no way of knowing which one any given respondent decided to answer.

More generally, "people tell the pollsters what they think they want the pollsters to hear" seems like a pretty durable truism. But what backs it up? What did the late polls in 1980 and 1984 look like, frex? (A genuine question for the guy with closer ties to the polling community than me.)

Jim, A fair question.

As I noted in a post yesterday, recent evidence seems to suggest that the so-called "Wilder effect" or "Bradley effect," where the poll numbers for black candidates is substantially overstated in comparison to the eventual vote, seems to be dissipating.

I don't have the older poll numbers handy but the "definitely vote" numbers are way, way higher than in previous cycles. the ABC/WaPo numbers for 6/11/2000 showed 49 percent saying they'd definitely vote and 36 percent on 7/17/95 (which isn't directly comparable, obviously, since it's a year earlier in the cycle).

In the case of voting turnout and "following closely," it may just be a conventional wisdom thing. Still, I'm rather sure that, despite almost everyone surveyed claiming that they're paying close attention right now we'd find that the vast majority of people don't know where the candidates stand even at a high level on their signature issues. And, while I expect 2004's high turnout to be replicated or even exceeded, I don't think we'll see 71 percent of eligible adults showing up. Or, indeed, 71 percent of registered voters.

My guess is that the admission of racial prejudice figures were so high because of question ordering. It came after a series of other questions on the subject which allowed people to get more comfortable. They were able to say that they though race relations were a big problem, that they themselves had good friends of other races, and so forth before being asked, essentially, "C'mon, admit it, you have racially insensitive thoughts every now and again even though you're not really an evil racist, right?"

More generally, "people tell the pollsters what they think they want the pollsters to hear" seems like a pretty durable truism. But what backs it up? What did the late polls in 1980 and 1984 look like, frex? (A genuine question for the guy with closer ties to the polling community than me.)