This is actually a response to a posting I did under the Role of Political Reporters letters page. In the interest of having the last word, I've dropped it in here for greater visibility. So if you are passing by Kitt and Glenn, I appreciate your opinion, and respond to it thusly:

Kitt, what is the control to prove that polls are in fact correctly mirroring public attitudes? Most elections are won on the basis of one or two percentage points, not an overwhelming demonstration of the accuracy of polls as far as I'm concerned. Nor do I consider consistency a hallmark of accuracy, as anyone could imagine, if you are using consistntly using the same flawed set of assumptions, you will be wrong and consistently so. In this regard, I would point out that most nation and state wide polls are conducted by phone, in an age where fewer and fewer mainstream people have lan lines, or answer the phone without knowing who is calling. As I noted, other kinds of polling do seem to be based on more mainstream populations, i.e, exit polls. The polls conducted by the Desmoine Register, interestingly enough, contradicted most of the nation wide polls.

Finally, there is an unquantifiable head-nodding affect that emerges from our polling culture. In my honest, and unverifiable opinion, I sincerely think that most people don't bother to have an opinion about something until they perceive a groundswell of opinion, which is usually directed to them by the media and is based on polling. Everyone knows George Bush is unpopular, but what if his popularity is actually in single digits, not double? We'll most likely never know. I suppose I'm being extreme by calling for an end to polling, but it certainly has borne an uncontrollable monster that is threatening the idea of democracy, where ideas and candidates are dahsed to bits before they ever reach the ballot box, and winners are declared before there is even an election. I perhaps should have been more precise by calling for an end to phone based polling.

And Glenn, yes, you probably never did say that, though I had a feeling you had. Sorry,in any case, I don't have the time to research your previous blogs (or the desire, you do admit your mistakes which is why I continue to be a big fan of your blog).

Its nicer here when people actually have discussions, and don't find ways of tearing each other down instead.

I did do the research--Rasmussen, CBS, NYT, all these polls are conducted by phone. The obvious weakness here, as I've noted, is that people who pick up their phone and spend ten minutes speaking to a recording or a stranger, may not actually be representative of the population at large. There is a wealth of study by sociologists and pollsters in this regard already, considering the rapid change of communication technology in just the past decade. Google "cell phones" and "polling" and you'll find a lot of analysis by the polling community. John Zogby himself noted in 2004, "The reality is that polling on the telephone is becoming more difficult; caller id and the widespread use of cell phones are affecting response rates." (though he does stand by telephone polling). This was a serious issue in the last presidential election but seems to have dropped off from the left, now that polls show that people dislike Bush

(see this article, http://www.mysterypollster.com/main/2004/10/arianna_huffing.html

I also noted that exit polling was based on a different strategy and was relatively reliable, and that the Des Moine register had conducted its own polls outside the rubric of the national polls and come up with very different numbers, and representative numbers. In places where there are no phones, sociologists and pollsters get out and knock on doors, which takes longer but is a much more accurate way of assessing public opinion. The Jerusalem Media and Communication Center uses this method to results that are respected through out the international community.

Lastly, the Bush thing was a rhetorical question. And we will never know for the same reason that physicists don't know what electrons are doing unless they monkey with them first. Polling is all we got, but its far from accurate, and definitely not accurate enough to be used in the way it currently is in political discourse.

Lastly, and this is to everyone: Debate is healthy and text can be read in a number of ways, why not try to encourage discussion, instead of stifling it by indicating that people that don't hold your opinion are stupid.

You responded to my post, and your critique of my post was that it was unsubstantive; now you critique it again, dismissing my argument entirely using the rationale that I am wrong, but you don't feel like going into the reasons why. That's lame.

Everyone is accountable for their opinion, Kitt, once posted. If you can't defend your position--because you can't spare the time, apparently--then why post it to begin with? You've posted twice now about how you're too busy to refute my argument.

And that's as much as I need to know about your opinion, I suppose. Good night.