Now it's about Henry Kaiser and his decedents. Please seize on a subject and stick with it. I promise to listen even if it's just you doing the talking.

... you should re-read IBC's statement on the paper, and then moderate your comments better:

We would hope that, before accepting such extreme notions, serious consideration is given to the possibility that the population estimates derived from the Lancet study are flawed. The most likely source of such a flaw is some bias in the sampling methodology such that violent deaths were vastly over-represented in the sample. The precise potential nature of such bias is not clear at this point (it could, for example, involve problems in the application of a statistical method originally designed for studying the spread of disease in a population to direct and ongoing violence-related phenomena). But to dismiss the possibility of such bias out of hand is surely both irresponsible and unwise.

Mr. Gabriel,

And you know it. This has nothing to do with science. You are taking a deplorable path in this argument.

you now wish you could take back. If you think it true and meaningful, your credibility just took a nosedive in my eyes.

Mr. Aaron,

What is almost certain (though not specified, one can reliably assume it) about the study is that the Iraqi participants are politicized, and odds are that persons of particular sectarian/political alignment were selected to conduct the surveys in specific areas. Otherwise they would be running extreme risks.

It is no accident that the highest mortality rates reported are, apparently, among Sunni Iraqis. Most Iraqi academics are Sunni.

What does this tell you about the overall situation in Iraq ? That a minority is profoundly aggrieved ? That there are segments of the population enthusiastically waging war ? This is not news. There is a litany of cases of the same sorts of persons fabricating news stories, conducting sabotage (such as the recent mass poisoning of policemen) and undermining the Iraqi government. Total war is total war, and respect for science cannot compete with that imperative.

That they would skew data to support their political opinions is a minor act in comparison to what so many of their bretheren are willing to do.

Let's say you're a poor family, part of a tribe living in some rural Shiite area. Several relatives, close and distant have been killed. A son, a brother, some cousins, a daughter's brother in law, etc. You hear of killings occuring almost daily, and the rest of the world seems not to care. The TV stations talk about fighting the occupier, and turn a blind eye to the kidnappings and suicides attacks that you have to deal with.

One day, some professional looking people wearing nice clean suits show up with clipboards, representatives of a distant world of civilization, peace, security and science. They're there to ask you about people in your family who have died. They say they want to produce an accurate count of the dead.

Now, technically, the only one that you're legally allowed to count is the son, since he's the only guy that actually was living in the house. But to you, the deaths of the brother, the cousin, are just as important. The daughter has moved back home and is still in mourning, and she doesn't think her husbands death is unimportant either.
Maybe you don't even under stand the point of only listing the son's death. You don't understand statistics, you don't see how anyone can count accurately without listing every death.

How are you going to respond to that? Are you going to sit there and abide strictly by the rules and only have them list the son's death? Or are you going to take your opportunity, a rare opportunity for someone to hear about your suffering, and try to mention as many deaths in your family as you can, regardless of whatever silly rules the scientists have?

http://www.news.com.au/dailytelegraph/story/0,22049,20588297-5006003,00.html

busily defending North Korea just last week.

Please explain how such a study would in any way affect the outcome in Iraq. Seriously.

I have no problem with open and honest discussion, I'm all for it in fact. I suggest that denying perfectly legitimate discussion from others because it does not fit your views of the relevance and validity of the study is not particularly conducive to that.

I've already acknowledged the essential validity of the math. As for your slow reading, lighten up -- my "slam" is no worse than was your "30 minutes to read" slam. A sense of humor is mildly helpful. :)

I did misstate this "You reveal your own prejudice by defending the methodology" -- should have specified that the methodology was not the issue, the timing, purpose and condition under which that methodology were used is, IMO, an issue. I apologize for misstating that.

OTOH, I am not dismissing the study because I dislike the results, those results are questionable in my mind because they are at odds with everyone else's numbers and because the timing of the release makes them suspect. Nor do I consider the study to be odious -- just suspect.

As for your quote, contemplate that the examples he cites are human acting counter to their own interests...

I'm not trying to be entertaining, nor to argue that human behavior is beyond study. I am definitely saying that sampling is subject to error and that most studies of any kind have a premise -- and they will frequently bear out that premise in spite of conflicting data. Nor am I willing or desirous of dismissing whole schools of thought. I am using some hyperbole to get you to realize that such surveys, studies and polls that produce 'metrics' deserve a fairly large amount of skepticism -- skepticism is not dismissal. I have not dismissed this study though I am very, very skeptical of the numbers.

I think we're at a standoff. You're asking me to ignore 45 years of experience -- the last third of them as a moderately senior user of the results of polls, surveys and studies in the implementation of policy -- which I'm not going to do and I'm asking you to contemplate things that are beyond your experience -- specifically, the difficulties of doing anything in a combat zone and the social habits of people from and in the Middle East (different than those exposed to the west) -- which you cannot do thus this is becoming rather pointless. Nothing pejorative in that, it's just a simple statement of fact.

The only reason to conduct such a survey is to serve a political aim; it has no practical value I can see. I fully understand your stated rationale for it but I cannot agree. It is unlikely to have any practical effect though it can admittedly have a political effect. I believe that such effect will be quite minor, you seem to disagree and that's okay.

You and the authors admit the data is the best that could be obtained -- that is, it contains imperfections. The question is how much did or do such imperfections affect the result. I submit the answer is a great deal. Aside from the vagaries of war having an obvious impact, the survey was conducted in the Middle East, an area of the world where escaping responsibility, obfuscation and bargaining are a way of life and where there is an overweening desire on the part of most to tell anyone asking a question what it is thought the questioner wishes to hear; an area where a guest in the home is accorded, to western mores, excessive deference. You can, as an Academic, dismiss that. I believe that would be a big mistake.

You called upon Livia, Mr. Alegria and I to present factual data to challenge the survey. Disingenuous at best; you're asking us to prove the negative. We have no obligation to do that -- regardless of your assertion that we do -- and, as I'm sure you realize, we could not produce it if we were inclined to do so. Good academic and rhetorical ploy but of no practical merit. If the study were as well done as you and Gabriel, a former third world pollster and thus both believers in the methodology seem to think, it would not be challenged by anyone. That it is challenged by two working Engineers and a former Government employee who worked policy issues should at least give you slight pause. Regardless, to say that it is a valid survey on mathematical grounds and that it uses a methodology used by many including the US Government is totally true as nearly as I can tell.

That does not address the fact that the probability of human foible and error due to location and circumstances adversely affect the accuracy and thus the reliability of the report.

You note that the Coalition was assigned responsibility for a higher number of deaths than you expected. That comes as no surprise, as you also noted, most people answer surveys fairly honestly. I agree and would add "in accordance with their perceptions." You also note that perception is reality, true again. Put those two together with the proclivities of people in the Middle East and the possibility of skewed data -- compounded by "no-go" neighborhoods and certainly selective picking of cluster locations for a variety of reasons and the entire survey has to be considered skeptically. As you did. And as I did -- the difference is just in the degree of skepticism accorded.

Like you and as you know, I was a Political Science major as an undergraduate. Unlike you, I did not pursue an advanced degree in the subject (and congratulations to you
for doing so) I instead went to work for a great many years in an environment that required application of my studies. During that time, I was bombarded with a number of
surveys and polls and I made two discoveries. The first was that there was always a reason for the effort -- and that reason almost invariably affected the design and thus the
outcome of the effort. The second fact was that the outcome was far more often than not called into question (usually to a greater rather than a lesser degree) by later events, be those events an election or more concrete facts later ascertained. I'd recommend that you bear that in mind as you later proceed to apply your own knowledge.

I can believe the possibility of *some* error for the reasons you mentioned. But you *do* need to provide a better argument than that if you want me to believe that there is *overwhelming systemic error* - a trend towards overreporting that can explain the numbers being more than a magnitude of order larger than you likely are willing to accept.

Your criticisms are valid, though they apply to almost every statistical study ever, as nobody is ever able to fully vet their data in the way you are demanding and ultimately you do have to just trust your respondents. Plus, as I have argued, measuring perception is arguably just as important as measuring reality, if not more so. So, your conclusions (that is, your desire to dismiss the study offhand) do not follow, again unless you are able to provide an argument for *overwhelming systemic overreporting*.

Can you point to it? Because the researchers say it's not there. The evidence is that passive methods undercount by up to 10 times. You have evidence this is false?

Becasue this is the core of the matter. You claim you now more about how this methodologyshould be aplied than the researchers but you provide nothibng to back it up.

that claims hundreds of thousands of Iraqis died as a direct result of the sanctions levied against Saddam isn't what I qualify as an appeal to emotion (which was the standing pre-invasion policy), nor is citing the policy before that (no sanctions and Saddam run amok), but whatever.

no amount of PHDs in Saddam's Iraq changes my mind that the situation in Iraq did not lend itself to an American invasion in order to establish forward bases of operation within the larger Middle East.