How is it possible that this hasn't been out there sooner?

I have to agree with Leslie on this one. I could care less who Bush gives the finger to or who Cheney tells to frick off on the Senate floor. I do care about the 100,000 civilian Iraqi deaths since the invasion (Estimate released 10-28-2004 by John Hopkins University), all the Americans and coalition troops who have been killed or wounded and the hostages that have been murdered. There is a lot of blood on that mans finger.

Oh my goodness! Does this mean the president may even have used dirty words! Get a grip, people.

On the serious topic raised here, Pedro, I'm sure it gives you great satisfaction to quote that study. It ought to give fans of the Lancet cause to regret that publication's decent into partisan politics. The study was admittedly rushed through by the Lancet's editor -- an action that he defends in language that exposes the political motivation of that decision. I don't know whether the Lancet has a written mission statement, but it's hard to imagine that this kind of activity falls within whatever such a statement would include.

This would all be bad enough if the study were statistically sound. Tim Worstall, writing at Tech Central Station argues that it is anything but sound.

One should reflect that politicized research bodies could have attacked the American and British governments during WWII with claims of even greater mayhem visited upon civilians. Pedro, I suspect, leaps on the claims of the cited study in the most uncritical manner, without a thought for either the motives of the study's authors and publishers, or the methodological problems of the study, or of the fact that the coalition forces are fighting against an enemy that dresses as civilians, that kills civilians deliberately for effect, and that uses civilian populations for cover.

I forgot to post the link to Worstall's piece:

http://www.techcentralstation.com/102904J.html

While, "The scientists who wrote the report concede that the data they based their projections on were of 'limited precision,' because the quality of the information depends on the accuracy of the household interviews used for the study." It still should be noted that "Even though the sample size appears small, this type of survey is considered accurate and acceptable by scientists and was used to calculate war deaths in Kosovo in the late 1990s." And "Richard Peto, an expert on study methods who was not involved with the research, said the approach the scientists took is a reasonable one to investigate the Iraq death toll."

And, as true objective scientists do, "The researchers called for further confirmation by an independent body such as the International Committee of the Red Cross, or the World Health Organization." These quotes are from the AP story, here.

It's a little strange to point us to a clearly biased site for an "analysis of the study." As a statistics nut, I can also say that a 95% confidence interval (CI) is pretty common in studies whose sample size cannot be large enough to safely encompass greater accuracy. And, to summarize, the formal definition of 95% CI: "The interval computed from the sample data which, were the study repeated multiple times, would contain the true effect 95% of the time." The author of the article you like states this as "a one in twenty chance that the effect simply does not exist." That's a bit miscontstrued.

Admittedly, these scientists are not saying their figures are 100% accurate. However, any clear understanding of these type of studies informs us that, indeed, if the figure is not 100,000 deaths, it can safely be determined to be close. I encourage you to read the study here. And, don't blame the messengers just because you don't like the message.

Check out this version of the salute piece.

http://www.dailyrecycler.com/blog/2004/10/choice.html

Meh. Kinda cheap.

The national guard quote is also taken out of context. Bush says when asked about Quayle and "There were lots of routes into the military at that time. Sometimes I guess if you wanted to serve in the Guard people made a few calls. I think Quayle should be proud of his service."

Bush never said he did not make calls. He never said he did. All he said in respect to his own service is that he served, proudly, in the National Guard.

Non issue folks.

"On the serious topic raised here, Pedro, I'm sure it gives you great satisfaction to quote that study."

Anthony please tell me how anyone could get great satisfaction from finding out about the magnitude of the slaughter of thousands of civilians?

I remember how I felt in the months that followed 9/11/2001. 3000 died during the attacks in New York, Pennsylvania and Washington DC. While time has partially healed the feeings, the horror never goes away.

Try to imagine how the people of Iraq feel. It's a much smaller country and they have lost a far greater percentage of their population than we did. Now I'm not talking about armed gunmen or terroists here, I'm talking about old folks, women and children who have not taken up arms against us. Even if the war against Saddam Husein was justified, the collateral damage of this magnitude is not.

Even if the John Hopkins Study over estimates the civilian deaths in Irag by a factor of 2 to 1, it is still an outrage. George W. Bush is directly responsible.

Pedro,

I don’t insist the slaughter of civilians itself makes you happy, but I find it hard to doubt you enjoy a certain surge of self-righteous justification in the opportunity to attribute evil to the current president. Hence your eagerness to uncritically affirm the study's findings.

Feeling horror about the victims of 9/11 is good, as far as it goes but it’s not enough. Action is required. Unfortunately, in taking action the President of the United States can always count on there being plenty of people to find fault.

Also, the practitioners of terrorism can always count on their being plenty of people to serve as tools toward their purpose. They create as much mayhem as they can, particularly among civilians, hoping that enough people like you will be horrified enough to sap the national will to fight. If the American electorate consisted entirely of people like you, despite their limited resources, the terrorists could always count on manipulating public opinion to their ends.

The Americans have sought to minimize civilian casualties as much as possible, while their opponents have sought to maximize them. Those enemies, in addition to directly murdering large numbers of civilians, have put them in the crossfire by using tactics forbidden by the acknowledged rules of war. Can there be any doubt, given the state of American weapons and tactics that at least a large number of civilian deaths – whatever the correct number is – are attributable to the insurgents use of civilian populations for cover?

Forget about what, say, the Geneva Convention says about responsibility in such cases. For you, “George W. Bush is directly responsible.”

Jesse,

You write that it’s “a little strange to point us to a clearly biased site for an ‘analysis of the study.’” I’m not sure what those quotation marks mean, because that is not the language I used. However, what is wrong with sending you to a site that tilts one way or another? BlueOregon.com is “clearly biased”; does that mean everything that appears there is inadmissible nonsense? The readiness with which you resort to fallacious argument doesn’t inspire confidence in your competence in judging methodological purity.

If the bias of Worstall’s site makes you uneasy – however it is to an evaluation of his claims -- the survey’s accuracy has also been challenged at a site with the opposite bias. You can read that opinion at the following URL, in an article by Fred Kaplan entitled “100,000 dead or 8,000? How many Iraqi civilians HAVE died as a result of the war?” [I use capitals to render italics.] http://www.slate.com/id/2108887/

You claim that “if the figure is not 100,000 deaths, it can safely be determined to be close. Is that what “limited precision” means? Kaplan offers argumentation that calls this claim of yours very seriously into question. Not being a “statistics nut,” like yourself, I won’t venture to adjudicate between your and Kaplan.

I’m not inclined to blame messengers simply because I don’t like the message. You may note that my criticism was directed at the editors of the Lancet, not the study team. I have no idea of the motives of the Johns Hopkins team, and I have no argument against their methodology (not yet, anyway, given my current state of ignorance). It seems clear, however, that the motives of the Lancet’s editors were political, and to such an extent that they departed from their normal timetable – which, one might say, “as objective scientists,” they normally do not

Anthony,

Good point, sorry.

But I hadn't wanted to call the article's methodology into question, just to point out it's political slant which most likely lead to this "This would all be bad enough if the study were statistically sound. Tim Worstall, writing at Tech Central Station argues that it is anything but sound. It seems that you believe Worstall, who is biased. I'm glad you're not calling the study biased as I thought your posting seemed to support the opposite. But Worstall chose to highlight only that there was a one in 20 chance that this range was totally wrong. It's a bit of a difference than if the study were repeated 100 times, 95% of the time the same range of 8,000-198,000 deaths would occur. Their methodologies were statistically sound.

Then you chose to point out the range of deaths--a worthy note. Using the article you mentioned, "This isn't an estimate. I It's a dart board. I agree. I think that's what the scientists meant by limited precision. And that's why I also think they called for peer review. They're scientists.

I do conceed that, "the lead researcher said he wanted it [published before the election." I'd also like to point out that, "The Lancet routinely publishes papers on the Web before they appear in print, particularly if it considers the findings of urgent public health interest."(--from the yahoo article I referenced earlier.)

And yes the reporter wanted it out before the eletion, but I see no deviation in timetable. The report was finished at the end of September. It wasn't published for a month. The editorial board said it's an accurate study with limited precision--once again reiterated in many articles I've read. Perhaps you take issue in thinking either a) they should wait longer than a month before publishing findings they consider urgent public health interest or b) the media should write scientific articles differently. I agree with b) but not a).

Thanks.

Jesse,

Thanks for the reply.

I brought up Tim Worstall as an example of someone who found problems with the study. I’ve suggested also Fred Kaplan’s piece. The bias of either of these commentators is irrelevant (Kaplan has political leanings opposite those of Worstall, but agrees that the survey is lousy). What matters is the quality of their arguments.

You suggest (between your recent post and the previous one) that Worstall is somewhat misleading in focusing on the fact that the study’s range has a one-in-twenty chance of being totally wrong. You prefer to focus on the fact that if the study were repeated 100 times, 95% of the time it would have the same range of 8,000 to 198,000. But then you acknowledge that this range is a large one. So without even quibbling with the survey methodology, the best one can say is that discounting the one-in-twenty chance that the range is completely bogus, the casualties might be anywhere between 8,000 and 198,000. I’m no mathematician, but “limited precision” seems wildly euphemistic for so wide a swath.

I’m not sure what you mean by the “reporter”; my complaint is with the editors of the Lancet. I certainly wouldn’t expect news reporters to wait. But I would expect the Lancet’s editors to follow their own timetable, which they have presumably adopted for science-related reasons. The Lancet has basically acknowledged it did what it did for political reasons, and to call this a matter of “public health interest” is simply wrong. It’s a political commentary, designed to have political ends. To use a scientific journal this way would be highly questionable even if the study were sound.