As far as taking polls to be "the word of God", you should be aware that polling is a multi-million dollar business wherein accuracy and constant methodology-perfection means money, careers, and the difference between thriving as a polling entity and going bankrupt.

Coast-to-coast, political candidates everywhere on the spectrum are right now busy raising all the cash they can, in part so they can afford to pay pollsters large sums of money for the duration of their campaigns. What do the candidates get in return? Quantitative feedback on candidacy viability, message and issue strengths and weaknesses, who's likely to vote and who's not and what's important to them and what's not, etc., etc., etc.

Bill Clinton, it's said, once commissioned a poll to determine where he should vacation - and went with the voters' "choice".

Do candidates and their organizations regard pollsters' reports as the "word of God"? Absolutely. Are their reports the "word of God"? Well no - they're the word of the People. And in those few cases where pollsters get it wrong (outside the stated plus/minus), they're as good as out of business.

Now, you can bum about how polls show that your views are in a minority, or even on the fringe, but it is wishful thinking and pure folly to question the general viability of quantitative public opinion measurement.

And one more thing...

I fail to see where any of the data you provide in your post undermines the general viability of polling results (from established pollsters). Please explain.

While you're at it, please explain this from your post: "40% of Americans THINK there's strong evidence that Saddam Hussein was involved in 9-11, while only 36% KNOW there is little to none."

Why exactly do the "thinkers" merely think and the "knowers" definitely know? Do the knowers have evidence the thinkers do not, or vice-versa? Or, who characterized each side as either "think" or "know" - you or the pollster? What evidence do you or the pollster have to support that distinction?

Did 100% of the 36% said "I know" and 100% of the 40% say "I think"? If not, who decided what exactly who "knows", and what exactly who merely "thinks"? (Talk about playing God!)

Equally important, since you write that "36% KNOW there is little to no (evidence)," you'll have to define the meaning of "little" evidence. Why does "little" evidence put them among the 36% and not the 40%? And of the 36%, what percentage said "little" and what percentage said "none"?

Did the "little" portion of the 36% specifically say the "little" evidence they have is not enough to convince them to join the 40%?

And what exactly is "little" evidence, how much less is it than the evidence that leads some to "think", and why is it not enough to convince to go from "knowing" to "thinking", or from the 36% to the 40%?

You're in the world of quantitative political science now. I await your answers with great interest.

M-----, I think you are attacking a statement David isn't making.

No one is challenging the general viability of quantitative public opinion measurement as a science. There is no question that public opinion can be measured within a determined margin of error.

But there is also not much question that not all polls do accurately measure public opinion (whether my design or incompetence).

The two issues raised by David's post (and they are very old questions) are:

1. Not all poll questions ask impartial questions; and

2. Not all poll questions ask meaningful questions.

Phillip Morris is perfectly capable of wording a poll that will "prove" that Americans love smoking, wish kids would smoke, and hate lawsuits against tobacco companies. The American Cancer Society could get the same people to "prove" that Phillip Morris should be disbanded, its executives imprisoned, and all smokers shot. Which is "true?"

When one sees such polls - such as Poll A declaring that 90% of Americans don't support secret searches, and Poll B declares that 90% of Americans support the Patriot Act's authorization of secret searches, there are a limited number of conclusions that can be drawn:

1. One, or both, of the polls is not accurately measuring public opinon; or

2. People are stupid.

The usual reaction is to dismiss the poll with the result you don't like as biased or inaccurate, and support the results you do like as true and fair.

I could be wrong here, but I suspect David wasn't implying that pollsters are off the mark -- rather that they're accurate, and that he doesn't like what they reveal about the mind of the American public. Of course, No one in this world has ever lost money by underestimating the intelligence of the great masses of the plain people. Nor has anyone ever lost public office thereby." According to Mencken, that is.

Likewise, some polls measure useless things. A number of poll have shown that a majority of Americans believe in Angels. A larger majority believe UFOs have visited Earth.

While it is an interesting proof of (in my opinion), the stupidity of people, it is also a somewhat useless poll from a factual standard: Whether or not Angels exist is not open to a mass vote.

On the other hand, it may be useful to know that people believe they exist, because I can exploit the belief independant of the fact (for example, I could host a popular television show where I claim to speak to dead people, and expect it to succeed).

The Iraq/9-11 question is similar. Whether 1%, 50%, or 99% of America believes there is a link is irrelevant to the truth of the link - either Hussein was involved, or he wasn't. But it is very useful to know that people believe he was involved, because that belief (whether true or false) can be exploited.

Personally, I often measure a polls validity based on the pollster's stake: When a political candidate runs a poll to determine the viability of his/her message, I trust it more because that candidate has a strong personal interest in an impartial result. When Phillip Morris runs a poll to determine attitudes towards smoking, I am highly suspect because Phillip Morris has no interest in an impartial result, and a substantial stake in a particular outcome.

The problem is, the media doesn't do a good job of sorting out polls on this basis - pretty much all but the most obvious manipulations are reported simply as "polls show . . . ." This creates further incentive to create false polls, because you know you won't get called on it.

I don't see David's post as a rejection of polls in general. Almost the opposite - he's bummed that polls show that views he finds wrong are more popular than he would like.

One of his comments points out that some poll data seems inconsistent (people reject individual parts of the Patriot Act, but support the Act taken as a whole), but even that doesn't seem to challenge the validity of the poll, but rather observes the ability of pollsters to obtain irreconcilable results by changing the phrasing of the question.

Finally, I would strongly challenge your assertion that "in those few cases where pollsters get it wrong (outside the stated plus/minus), they're as good as out of business." There will always be a place for pollsters who are willing to do a pre-determined poll, just like there will always be a place for movie reviewers who give every film five stars: Some people are paying for outcome, not truth.

Did 100% of the 36% said "I know" and 100% of the 40% say "I think"? If not, who decided what exactly who "knows", and what exactly who merely "thinks"? (Talk about playing God!)

If David's referring to the poll that I think he is (which got a good argument going on my own blog...) the question has specifically to do with whether the 9/11 hijackers were Iraqi. A hell of a lot of Americans think they were. Since they're wrong, that's why they're only thinking, not knowing so.

Golly, you guys got me all wrong. Far from "attacking", I was just asking the questions a quantitative political scientist would ask of such a polling exercise, minus dozens more that he'd ask and I can't think of. Believe me, I've suffered through their classes and read their papers.

In a Roshomon kind of sense, it's funny you all saw the intent of his post differently from me. I can't know who's wrong and right. But his "word-of-God" lead-in seemed disapproving of those who place credence in polls (as we've recently discussed Bush-approval and Iraq war polls). And there's the "Moron Polls" header...

Meanwhile, if you go to the link, the writer of the polling data (the real culprit here) seems to use his poll's results to impugn other, more mainstream polls.

So was he relying on the validity of polls to argue the public is dumber than he'd like to think, while also casting doubt on the general polling process? I think so. Is that a contradiction? Sure, but that's how polls are sometimes used - to pick & choose and misinterpret to fit pre-desired conclusions.

Now, contrary to what Jimpy writes, behind the scenes the people who commission polls - tobbaco companies, candidates, etc. - generally face up to the hard numbers they get. What they report to the public ("We have momentum!") is another story. That is, they do "pay for the outcome", even though said outcome may not be what they tell you.

Also Jimpy, you wrote:

I would strongly challenge your (my) assertion that "in those few cases where pollsters get it wrong (outside the stated plus/minus), they're as good as out of business."

I'd cite last Fall's example of the Quinnipac (sp) College poll. Was it the NY governor's race? Can't recall, but they came out with numbers totally contrary to everyone else's. People went "Huh?!" When, with considerable embarrasment, they announced they'd found a flaw in their methodology specific to that poll, they suffered a much-discussed reputation reversal that would have just about sunk any non-educational, private sector pollster.

In the dog-eat-dog realm of political pollsters-for-hire, you're only as good as your last you-know-what.

I suppose we'll just have to wait for David to return and tell us what he meant.

He should be here.

He didn't say for sure he'd come.

And if he doesn't come?

We'll come back tomorrow.

And then the day after tomorrow.

Possibly.

And so on.

The point isÂ-

Until he comes.

Wow. Such a lively discussion in such a short time.

First things first--I'm not sure I get the prior post, Mike. Can you enlighten me?

Second, what I meant isn't, I think, entirely relevant, because all I meant to do was get this out there and foster some discussion. So it's not like we're voting on what I thought and someone has to win.

Basically, (most) everyone got what I was trying to say. First, that polls are phrased in manipulative fashion and then cited as hard evidence that "the people" want what we say they do, and we should be skeptical of them based on the constituencies involved. Second, that even the same poll can reveal that people are uneasy with the things they supposedly support overwhelmingly. And third, that people are pretty uninformed. Other than that, I find the results of the polls alarming, because I think that support for unprovoked, unproven attacks puts us in very bad historical company.

Waiting for Godot

This is America. Someone always has to win.

Seriously, DB, in the poll you cited, the "think" and "know" parts of the summary immediately red flag us to the pollsters' and/or poll summarizer's bias. If he wrote such a summary for the Gallup Poll, or for the quantitative professor I had, he would be raked over the coals with all the questions I raised and more. If he didn't he didn't have solid answers for his poll questions and/or interpretations, he'd be fired or handed a fat D for dunce.

Therefore, the poll, or poll summary you cite does support your contention that polls are subject to bias and manipulation, but not in the way you intended. You've cited a biased poll to prove that polls can be biased - and it did the job.

In further irony, your poll thus has far less credibility than mainstream polls that consistently showed a 75% war approval rating.

Once again, M-------beware of irony. You might've had a point until your last paragraph, o credulous one.

Actually, I prefer "o incredulous one."

Again I repeat: do you have "strong evidence that Saddam Hussein was involved in 9-11"? Do you? I'm not talking about "can you show any link between SH and OBL." I'm talking about 9-11, which is the question at hand. If you do have such evidence, please forward it to the White House. If you don't, please shut up.

"Shut up"? Jeepers. What's become of decorum? Civility? Good manners? R-e-s-p-e-c-t?

I'm not aware of anyone else here - even in the heat of losing a debate - telling someone else to "shut up." Now, I don't see civility on your blog as deteriorating, but if things ever do go that way such a sad pattern might be traceable to the moment when - in the midst of an otherwise perfectly civil discussion - you told someone to "shut up."

If this wasn't your blog, I'd ask for and expect an apology and retraction - and only so that things don't deteriorate in the future. But it is your blog, and I'm not questioning your prerogative to do what you want on it. Just don't be surprised if you ever reap what you sow, though I assure you I will never be a contributor to such a deterioration

That said, it's hard to not take "shut up" as a sign of weakness - frustration at an inability to defend the poll's unfounded and biased distinction between who "knows" and who merely "thinks." For specifically your edification, I will repeat the fact that such a distinction in a poll summary, absent crystal clear criteria for that distiction, does not meet mainstream industry standards and would render any such poll - as it does this one - non-credible.

Now that we've got that out of the way, you'll recall that Vaclav Havel's intel service reported that Mohamed Atta had a meeting with an Iraqi intelligence officer in Prague during the summer of 2001. I guess they must have talked about how the Yankees might fare in the playoffs. Since I neither "think" nor "know", I'm admittedly just guessing.

And that is strong evidence?

OK, so "shut up" was over the top. I'm just pissed off. You won't like me when I'm angry. Not that I turn green or anything.