Talk:Lancet surveys of Iraq War casualties/Archive 1

2005 - July, 2006 Relevant commentary for editors to read: http://www.cjr.org/issues/2005/2/voices-guterman.asp?printerfriendly=yes

Wrong interpretation fixed
I just corrected the following as warned, due to lack of response:

As it is right now, this article claims that "The authors estimate that 98,000 more deaths than expected happened after the invasion outside of Fallujah, and estimate 200,000 if the outlier Fallujah cluster is included."

The part I have highlighted in bold is patently false. I have decided not to revise the article yet since there seems to be discussion in here and in the revision history about this figure.

According to page 5 of the report "[the data] indicates a point estimate of about 200 000 excess deaths in the 3% of Iraq represented by this [Fallujah] cluster."

Their point estimate showed that 200,000 would have occurred in only Fallujah's 1 cluster, not in all 33 clusters of Iraq.

Keeping in mind that 200,000 a point estimate, and not an interval estimate (like the figure for the other 97% of Iraq), if we tried to calculate the total estimate including Fallujah, we'd have to assume that it makes sense to add a point estimate and the most probable estimate in an interval estimation. Assuming it does make sense, then obviously the more correct figure would be 200,000 + 98,000 = 298,0000.

Notice in the report, the only mention of a total estimate including Fallujah is on page 1, and all it says it "and far more if the Falluja cluster is included".

I'm not sure it does make sense to add the two or use any other method to combine the two, so I'm leaving that for someone more knowledgeable to answer. But I do know that by my reading of it, 200,000 for all of Iraq is just wrong.

In any event, if there isn't an answer here in a reasonable time, I'm just going to delete it and leave the total figure question open.

A Reply to "Wrong interpretation fixed"... You are indeed correct, as the 200,000 figure is false and misleading. I have added the following more precise information: "Had the Fallujah sample been included, the survey's estimate would have been of an excess of about 298,000 deaths, with 200,000 concentrated in the 3% of Iraq around Fallujah (Roberts et al p.5)."

Reaction dot points
Below is some points an anonymous poster added in the "reaction" section. It might be useful as the basis of discussion, but it's way too POV and the style is all wrong. --Robert Merkel 23:51, 18 August 2005 (UTC)

The study randomized by area but projected by population, which vastly over-represents rural populations. In order to be accurate, a sample must representative of the population NOT the land! <--- I added a Lancet Publications Related to Criticisms section. the first item in the list deals directly with this insofar as it pertains to airstrikes and collateral damage

8 of the original clusters were deemed too dangerous to study and replaced with eight other clusters, invalidating the study.

NO clusters selected by area were allowed to be unpopulated. Nearby populations were chosen instead, using criteria not entirely obvious.

Only 6% of the deaths were verified by death certificate, lest the interviewed be "offended" by the request for documentation.

NONE of the Westerners spoke the native language, Two Iraqis were added to the masthead AFTER the initial publication.

Inclusion of the fallujah cluster makes the maximum study death toll 194,000.

Exclusion of the fallujah cluster makes the maximum study death toll 100,000.

Edsels happen because of this type of statistical research.

The Lancet study is probably more about politics and stem-cell research than mortality measurement.

Have you got a better analysis of this study's flaws?

why did you censor mine?

what a worthless group of propagandists you seem to be!

indeed you added 200,000 as a maximum, which is not even IN the study!

I can guess your politics...

are we going to get into reversion wars here?


 * These points are mostly inaccurate, and written in non-POV language. The 200,000 maximum is indeed in the study -- see page 5 of the PDF. One or two of the points are worth mentioning, and I will add them to the article in due course unless someone else does so first. You also seem to be under the mistaken impression that Wikipedia contributors are not independent individuals. Rls 00:31, August 20, 2005 (UTC)

this really need to be something other than a mindless parroting of an obviously flawed study!

relevant discussion to read.
Talk:2003_invasion_of_Iraq/archive, see also Talk:Current events archive 10. --Silverback 20:56, August 19, 2005 (UTC)

Statistical points
Silverback, please provide evidence for the following: Although, this study's statistical strength was much lower than the Lancet usually publishes

Also, please explain the following; it seems to be meaningless: It should be noted that the 8000 to 194000, 95% confidence interval is purely statistical, and does not include any adjustments for the several limitations noted by the authors. Rls 01:17, August 22, 2005 (UTC)


 * Can you name even one Lancet study with poorer statistical strength? Search on "limitations" in the article, you will find several mentioned.  The statistics don't add that uncertainty to the confidence interval.  Consider that only 2 deaths were confirmed at each cluster site, consider also that none of the attributions to air strikes were confirmed etc.  Consider the limitation due to known problems with human recollection as they admitted and as demonstrated by other studies, this uncertainty was not accounted for in the statistics.  The reason studies with this poor of statistical strength are not usually published, is that the ultimate confidence interval would mean that not just the magnitude, but the direction of the result is in doubt.  Note, that in Falluja, it may not have been safe to properly attribute the deaths.  The article states that the researchers did not feel safe asking some questions.   Note the discrepency in ratio of womem/children killed between Falluja and the rest of Iraq, this is not explained at all in the study?  Are there more women per children in Falluja?  Note the authors uncertainty about whether young males deaths were legitimate targets or non-combatants.  This uncertainty is not accounted for in the statistics or the confidence intervals.  The statistics were calculated as if all deaths reported as non-combatants really were.--Silverback 02:43, August 22, 2005 (UTC)


 * That's quite a few points. Some of them, such as confirmation of deaths and nautre of the casualties are certainly worth discussing in the article. However, these are mainly qualititive points which obviously cannot be represented statistically so they could not be taken into account. With the Falluja data, the authors note the problems and they are not included in their primary conclusion, so I think that's a red herring. As for other Lancet articles of a similar nature, I think the onus is on the person suggesting such a thing to provide evidence -- you can't prove a negative. Did I miss any of your points? Rls 12:15, August 22, 2005 (UTC)

Politics
However, the politics of the authors and reviewers is demonstrated by the interpretation that "air strikes from coalition forces accounted for most violent deaths"'', even though this conclusion is based upon a survey of laymen, when expert investigations would be needed to determine which deaths were caused by coalition munitions, and which from "friendly" fire of motars and anti-aircraft shells and missles. Iraqi weapons fire was notoriously and carelessly ineffective and inaccurate'' This statement is ridiculous. The report is pointing out that most deaths occured from air strikes, not ground fighting, which is an easy distinction to make. Rls 01:22, August 22, 2005 (UTC)


 * How about a compromise, something along the lines of: [It has been speculated by critics that an unknown proportion of deaths may have been caused by] "friendly" fire of mo[r]tars and anti-aircraft shells and miss[i]les. Iraqi weapons fire was notoriously and carelessly ineffective and inaccurate. ? Rls 01:28, August 22, 2005 (UTC)
 * That sounds good. But also eliminate the hyperbole about Lancet, they really lowered their standards on this one.--Silverback 02:30, August 22, 2005 (UTC)
 * OK, I'll add it. But the Lancet is a very prestigious journal -- certainly the top medical journal in the UK, and arguably the world. They were very keen to cover their backs on this paper and made sure it was heavily reviewed by a large number of experts in the field. Rls 12:19, August 22, 2005 (UTC)
 * Let's not use the phrase 'friendly fire' too much - if it was a US bomb-strike, then it was a US bomb-strike. One must not assume that it is in any sense 'friendly'.
 * Johnbibby 16:07, 29 October 2006 (UTC)

Statistics
Silverback, please show how you worked out that "there is an 8.8% probability that the number of excess fatalities were between 95,000 and 105,000" -- a quick back of the envelope calculation suggests that this is way off. Anyway, this is a strange range to pick. A more interesting statistic would be the range for 50% or 25% accuracy. Rls 19:34, 23 September 2005 (UTC)
 * I pick the range, because a lot of people using this study, just want to use the 100,000 figure, implying that the likely result is close to this figure. Not only is the 100,000 figure a poor representation of the result, even a range bracketing that figure is a poor representation as demonstrated by my calculation based on the normal distribution.  I calculated this probability in excel, the formulas follow:

=+NORMDIST(95,98,45,TRUE) =+NORMDIST(105,98,45,TRUE)

These return the cummulative probabilities at 95 and 105, which are .4734 and .5618 respectively. You difference these two to get the probability for the range: .0883. It is just as valid to discuss the probability of the middle of the range as it is at either extreme. I think someone earlier (was it you?) disputed the flatness of the bell curve, this calculation shows that it is much flatter that most people and evidently you also, would suspect from the way this study is presented.--Silverback 07:29, 25 September 2005 (UTC)
 * Where did the 45 come from? (Will be away from Wikipedia for a week or so, haven't got time to check your calculation yet) Rls 13:24, 25 September 2005 (UTC)
 * That's the standard deviation, half of the two SDs used for the original range.--Silverback 17:46, 25 September 2005 (UTC)

Justification
Although the death toll is light compared to the immense infant mortality that former U.S. secretary of state Madeleine Albright estimated was caused by the UN sanctions. -- this should not be in this article. It seems to be an argument over whether the invasion of Iraq was justified, which should not be part of this article. There are several others where this could be discussed. Rls 13:24, 25 September 2005 (UTC)
 * I think it is questionable, but it balances the immediately previous phrases which apparently were aimed at justifying opposition to the war. "There were apparent political biases at work as well. The results of the study were inconvenient to many supporters of the war, since such a heavy death toll could raise questions regarding the humanitarian justifications" Words like "heavy" are relative, and if the article could raise questions about the humanitarian justifications, those quite considerable humanitarian concerns should perhaps be included for balance.--Silverback 17:49, 25 September 2005 (UTC)

I have removed the section that compared the Lancet death count to the sanction death count for one very good reason - the Lancet study was trying to discover additional deaths. In other words, deaths from the usual infant mortality rate were factored in when they came up with the 98000 figure. --One Salient Oversight 07:47, 16 October 2005 (UTC)


 * I don't think the additional deaths applies here, because the Secretary's statements were from several years earlier. The Oil for Food program was tolerated despite its evident corruption because some assistance was reaching these people and reducing the infant mortality rate. There were vaccination programs etc.  The study also could not capture the episodic and geographically concentrated nature of Saddam's attrocities, since it did not look back far enough in time, and would have to sample geographically at a high enough resolution to put several samples in the specific areas where the mass murders took place.  Les Roberts admission of bias is quite remarkable, he wears it almost like a badge of honor.  This study's poor quality and biased rushed publication is a black mark on the the Lancet's reputation. Since the bias admissions are being cited in the criticism section, I've modified the language to make it clear that the Roberts quote is being offered in support of that criticism and is in fact an admission of it. --Silverback 08:30, 16 October 2005 (UTC)

Deleted criticism of baseline death rate figures
Hipocrite deleted a paragraph critiquing the study's calculated baseline of 5 deaths per 1000 people. The criticism exists, and to delete it was inappropriate. If Hipocrite believes the criticism is incorrect, or could be characterized better, then the appropriate thing to do is to edit to reflect the counter-argument, rather than to delete the criticism as if it doesn't exist. Here's the quote: The study, though, does have a fundamental flaw that has nothing to do with the limits imposed by wartime—and this flaw suggests that, within the study's wide range of possible casualty estimates, the real number tends more toward the lower end of the scale. In order to gauge the risk of death brought on by the war, the researchers first had to measure the risk of death in Iraq before the war. Based on their survey of how many people in the sampled households died before the war, they calculated that the mortality rate in prewar Iraq was 5 deaths per 1,000 people per year. The mortality rate after the war started—not including Fallujah—was 7.9 deaths per 1,000 people per year. In short, the risk of death in Iraq since the war is 58 percent higher (7.9 divided by 5 = 1.58) than it was before the war.

But there are two problems with this calculation. First, Daponte (who has studied Iraqi population figures for many years) questions the finding that prewar mortality was 5 deaths per 1,000. According to quite comprehensive data collected by the United Nations, Iraq's mortality rate from 1980-85 was 8.1 per 1,000. From 1985-90, the years leading up to the 1991 Gulf War, the rate declined to 6.8 per 1,000. After '91, the numbers are murkier, but clearly they went up. Whatever they were in 2002, they were almost certainly higher than 5 per 1,000. In other words, the wartime mortality rate—if it is 7.9 per 1,000—probably does not exceed the peacetime rate by as much as the Johns Hopkins team assumes.


 * I have clarified Fred Kaplan's argument, and provided the CJR rebuttal to it in my most recent revision. Hipocrite - &laquo; Talk &raquo; 18:50, 12 October 2005 (UTC)


 * The 10 biostaticians would most likely only confirm that the calculations and the conclusions were reasonable. There is no report that they attempted any independent verification or sanity checking of the data, which is the type of issue Kaplan is sourced for in this article.  It doesn't seem characterizable as a "rebuttal" at all.--Silverback 20:44, 12 October 2005 (UTC)


 * Did you read the CJR article? I'm happy to include more of it. Are you doing origional research on what Fred Kaplan's issues were? I'm happy to do origional research and tell you that he's wrong about death rates - a population that had a high death rate in the past tends to have a lower death rate in the present - because all of the weak and inferim were killed off in the years prior:
 * One columnist, Fred Kaplan of Slate, called the estimate “meaningless” and labeled the range “a dart board.”
 * But he was wrong. I called about ten biostatisticians and mortality experts. Not one of them took issue with the study’s methods or its conclusions. If anything, the scientists told me, the authors had been cautious in their estimates.
 * Hipocrite - &laquo; Talk &raquo; 14:05, 13 October 2005 (UTC)
 * Fred Kaplan said (1) the study was a dartboard; and (2) cited Daponte saying that the study's baseline calculations were inconsistent with accurate United Nations figures. The CJR article addressed Kaplan's criticism, and said absolutely nothing about Daponte's criticism.  -- FRCP11 14:18, 13 October 2005 (UTC)

I read the CJR article, and so no indication that the biostatisticians specifically addressed the prewar mortality issue.--Silverback 05:46, 14 October 2005 (UTC)

daponte and pittsburgh article
My proposed edit:"Fred Kaplan argued in Slate that the prewar mortality rate finding of 5 deaths per 1000 must have been low because Iraq's mortality rate in the period from 1980-1991 were all higher, ranging from 8.1 to 6.8, and that the study therefore overstated the increase in the mortality rate because of the comparison to incorrectly low baseline numbers. The source of Kaplan's mortality figures, Beth DaPonte, called 'the nation's leading authority on non-battle deaths in Iraq' by the Pittsburgh Post-Gazette, used numbers from United Nations statistics; coincidentally, DaPonte's study of mortality rates from the first Persian Gulf War were criticized as too high by the United States government." Hipocrit edit:"Fred Kaplan argued in Slate that the prewar mortality rate finding of 5 deaths per 1000 must have been low because Iraq's mortality rate in the period from 1980-1991 were all higher, ranging from 8.1 to 6.8, and that the study must therefore be wrong, as it found a 5.0 mortality rate. The source of Kaplan's mortality figures, Beth DaPonte, used numbers that 'are higher than those of other researchers,' according to the Pittsburgh Post-Gazette." Hipocrit's edit is a complete misreading of the Pittsburgh article he found. Daponte criticized the Lancet study because their baseline number was substantially lower than the 1980-1990 baseline number, and because all evidence is that mortality rates increased after the Persian Gulf War. The Post Gazette article (1) calls Daponte the leading American expert; (2) mentions criticism of Daponte's Persian Gulf War civilian and military casualty estimates by the US government; and (3) says absolutely nothing about Daponte's figures for civilian mortality for pre- and post-Desert Storm years, which is the crux of the criticism of the Lancet calculations. Hipocrit also deleted the text indicating that Daponte's 1980-90 figures come from the United Nations. Hipocrit's edit is either a misreading or misleading, but his cite doesn't correspond to the text he added, and his deletes are similarly inappropriate. He makes no attempt to justify his edit on the talk page. I thus revert. -- FRCP11 12:53, 14 October 2005 (UTC)


 * No, your edit is the one that is misreading. The 5/1000 figure is not a baseline, it was an experimental finding. This is the relevent quote:


 * In a subsequent 1993 study funded by Greenpeace, Daponte updated and publicly presented her analysis of the Gulf War, raising the total Iraqi death count to 205,000. She estimated that 56,000 Iraqi soldiers and 3,500 civilians were killed during the war, and that another 35,000 died as Saddam Hussein crushed Kurdish and Shiite rebellions that rose up after the United States stopped fighting. The largest number of deaths -- 111,000 -- Daponte attributed to "postwar adverse health effects."
 * Most of Daponte's estimates of Iraqi casualties are higher than those of other researchers. But National Defense University Prof. Judith Yaphe, a former CIA analyst, thinks DaPonte underestimated the number killed in the Kurdish and Shiite rebellions. Yaphe thinks at least 60,000 died in those uprisings, perhaps as many as 80,000 to 100,000.

Hipocrite - &laquo; Talk &raquo; 19:19, 14 October 2005 (UTC)
 * Precisely my point. The story is talking about DaPonte's estimates of 1991-92 casualties.  That doesn't change the fact that her repetition of the 1980-1990 mortality figures accurately reflect United Nations statistics.  You've given no reason for that deletion, so I've restored it.  I'm not sure why you think the fact that some researchers think DaPonte overstated 1991-92 casualties and another researcher thinks DaPonte understated 1991-92 casualties by a factor of 2 or 3 is so damning of her criticism of the study. -- FRCP11 19:56, 14 October 2005 (UTC)

The current article reads "some defenders of the Lancet study believe Daponte's figures to be unrealistically high". "Some defenders" is plural, so I would expect someone to cite two sources challenging Daponte's criticism if we're going to keep that text as it currently reads. Wikipedia editors otherwise unpublished don't count. -- FRCP11 20:17, 16 October 2005 (UTC)
 * Hearing no response, I will make an edit. -- FRCP11 00:44, 18 October 2005 (UTC)
 * Never mind, someone did it already. -- FRCP11 00:46, 18 October 2005 (UTC)
 * I was bold and just went ahead and did it. Hipocrite - &laquo; Talk &raquo; 13:56, 19 October 2005 (UTC)
 * Fair enough. Given that it was on the talk page, I wanted to put it up for discussion before I did it; a second editor agreeing with me and deciding to do it works also.  -- FRCP11 14:17, 19 October 2005 (UTC)

missed opportunity?
Some people have proposed that as a rule of thumb there are two to three times as many injuries/wounds as there are deaths in these kinds of conflicts. It occurred to me that the researchers missed an opportunity to perform a sanity check on their research results by not also inquiring about injuries and wounds, further more, they would not have to rely just upon recollection or the respondants interpretation, since many injuries and would leave traces such as scars and disabilities that could be confirmed.--Silverback 20:41, 12 October 2005 (UTC)

Tim Lambert attempts to control the page
I've recently added two things to the "Criticism" section, namely two facts in paragraph 2 about the way in which the "58 times more violence" and "84 percent caused by Coalition" claims have been made by the Lancet authors without noting that these require the re-introduction of Falluja data, which they claim to have discarded in the 100,000 estimate. The claims have been made without indicating this, which means without indicating that they are based on outlier data and therefore wildly uncertain, at least. This is factually accurate and a valid criticism. Tim Lambert deleted it twice because he doesn't want criticism to appear in the "Criticism" section. He wants only to have a "Refuted Criticism" section where everything can be presented as he'd like, meaning presented as if there are no valid criticisms of the Lancet study, and it is gospel truth. Since the two criticisms above can not be presented as having been refuted, they are sent down the memory hole by Lambert.


 * The "58 times" requires the inclusion of Falluja. But that's what the  study says.  And it's uncertain, but that's what the confidence interval tells you and that's in the study and in the wikipedia article.  The "Criticism" section is full of criticism. Including every single criticism of the study ever made and the responses to such criticism would result in a 50,000 word article.  This particular criticism is neither notable nor accurate. --TimLambert 03:27, 15 May 2006 (UTC)


 * Ok. I'm willing to go along with this one and not bother fighting it. That point does not need to be on the page, even though I think it should be. However, the remaining points, if the page is going to claim to ostensibly present Iraq Body Count's view of Lancet, it must do so accurately, and not falsely and completely misleadingly, as you would have it.

My second revision was to change an inaccurate description of the website Iraq Body Count, which refers to that website as using specific techniques of Marc Herold which it does not use. Its methods are not the same as Marc Herold's, just loosely based on them, and the description that was there (and that Tim keeps reverting to) is wrong.


 * So correct just that then, as a separate edit. --TimLambert 03:27, 15 May 2006 (UTC)


 * IOW: all changes or additions must be Tim Lambert approved.

Then, since the page refers to what IBC says of Lancet, I added the recent fact that IBC has analyzed the Lancet study, or one part of it, as probably being too high, based on comparison to the much larger UNDP/ILCS study. IBC has done the analysis three ways, all of which come to the same conclusion. Tim Lambert claims the first of these ways has a technical error which makes it too uncertain to draw this conclusion, a claim with which IBC disagrees, and Lambert has never addressed the other two corroborating analysis. But agree or disagree, IBC has made this criticism of Lancet, and the existing entry on the page does not accurately reflect IBC's view as it stands. But since Lambert now doesn't agree with that view, he again wants it sent down the memory hole so the page only implies that IBC thinks Lancet is right, which, in fact they do not.


 * The IBC's alternative methods all suffer from similar problems.


 * Thus spake Tim Lambert: self-appointed sole bearer of truth. That is a bald assertion of your opinion, and one that IBC obviously does not agree with. The page is not claiming to describe your - Tim Lambert's - opinion of the Lancet study, but rather that section is claiming to describe IBC's position on the Lancet study. The page as it stands - after your repeated suppression efforts - is false. IBC's view is the one you keep trying to send down the memory hole. It's not for you to decide whether that's right or wrong. It's for the readers of wikipedia, whom you obviously do not want to have that opportunity.


 * The uncertainty in the data doesn't allow them to draw the conclusions that they attempt to draw.


 * You're asserting your opinion as truth. The page is not claiming to describe Tim Lambert's opinion. It's claiming (falsely, thanks to you) to describe IBC's position.

The article cannot include every erroneous criticism of the study. --TimLambert 03:27, 15 May 2006 (UTC)


 * No, it can only include what Tim Lambert either agrees with or those criticisms that Tim Lambert can present a ready-made refutation for, obviously. Thus the page is nothing but a partisan and selective Tim Lambert editorial about the Lancet study, not an encyclopedic entry. So much for trusting wikipedia for an informative description of anything controversial if one partisan to the issue can just delete any facts or views on the issue that he doesn't like. If people want to read wikipedia for biased editorials by thoroughly partisan hacks, maybe your method will be useful to someone.


 * I didn't write the page and I've only made a few changes to it. The page is the consensus of the various people who have edited it.  If you don't think that partisans should edit it, why are you editing it? Or are you claiming to be impartial? --TimLambert 16:27, 15 May 2006 (UTC)


 * I'm editing because the previous page made a false claim about both the Lancet authors and the IBC authors, neither of which is sustainable given numerous recent comments by Lancet's lead author about IBC which asserted their numbers are inconsistent with Lancet, or IBC's response to these. It may have been plausible a year ago, when neither authors seemed interested in creating controversy with or nitpicking potential inconsistencies between the two, to interpret some statements of each that way. But it's hardly sustainable given the positions taken in the last year. And IBC's actual view of Lancet can hardly be "not notable" if the page is claiming to note it. What is "not notable" is Tim Lambert's view about IBC's view, and therefore is properly not noted in the page. Your last revision was at least better worded, and closer to accurate, but it still did not reflect the views of either the Lancet or IBC authors, which the page claims to be doing.


 * Evasive. Are you denying that you are a partisan IBC supporter? And will you please sign your edits on the talk page? --TimLambert 17:51, 15 May 2006 (UTC)


 * I'm evading your evasions, rightly. You're evading the factual issue: Does the page as it stands accurately reflect the stated views of the Lancet authors (lead author Les Roberts in particular) and IBC about each others' studies, as the page *claims* to do? The answer to anyone familiar with either is quite clear that it does not. Both sets of authors have published statements that contradict what was described as their positions on the page. Instead of addressing this, you want the discussion to be an ad hominem exploration of me. That's evasive.


 * An ad hominem argument about you was relevant however, because your claims are appealing to your own opinion about IBC's actual position, which, because you don't agree with it, you want suppressed. That is an appeal from your own supposed authority. And therefore, the validity of you claiming this authority and declaring yourself the arbiter is relevant. My argument requires no appeal to my authority for making value judgments about the validity of Roberts' or IBC's positions, as yours requires. It merely requires looking at their stated positions, and seeing the the current page very obviously does not accurately describe them.


 * The current edit is at least a little more accurate than those previous, but still falls well short of the truth. It's a waste of time trying to publish the truth of these positions when someone who is determined to suppress them has the power to do so at will. At least my dissent is noted here, and I may add to this discussion page all the relevant quotes backing up my points that the current page you insist upon maintaining does not reflect the stated views of either authors, as it claims to do.


 * And, for now at least, I'd prefer to post my argument anonymously, as is common practice for many of the world's leading epidemiologists to do.


 * I think we can guess why you want to be anonymous -- you are one of the authors of the IBC criticism and out of vanity want it mentioned on this page. Did I guess right?


 * No.


 * Oh, I think I did. It's Josh D, isn't it? --TimLambert 07:14, 16 May 2006 (UTC)


 * You've attempted to divert the issue to ad hominems again. Vanity makes people appeal to their own spurious authority to control public web pages, and makes them stand by risible howlers such as "So an independent study has confirmed that part of the Lancet study" (surprisingly, somehow the "uncertainty in the data" does not prohibit this comical assertion).


 * The page is not about the dispute between Les Roberts and the IBC -- that dispute belongs on the IBC page.


 * Why does it belong on the IBC page and not the Lancet page? In any case, certainly some random wiki reader is bound to believe Les Roberts' claims are "neither notable nor accurate" and "rubbish", and therefore they do not belong on the IBC page, right?


 * One criticism of the Lancet estimate is that it is much higher than the IBC number. The page correctly states that there is no contradiction between the two numbers because they are measuring different things.


 * No, the page claims to be describing the views of the Lancet and IBC authors about their respective studies "corroborating each others findings". An assertion that seems rather difficult to carry given many recent published claims.


 * The IBC criticism of the Lancet has nothing to do with this.


 * Ah, the quick shift from 'King Lambert decrees IBC criticism rubbish', to 'rational observer Lambert asserts that IBC's view of Lancet is irrelevant'. However, even this new line is dubious. The actual positions of the Lancet and IBC authors has quite a lot to do with what is "According to Lancet and IBC authors". But this is, of course, irrelevant because it's not what is "According to Tim Lambert", as you demonstrate yet again by putting back on the King Lambert crown below and decreeing who is right and wrong.


 * As well as being erroneous, it's not based on the IBC number. And will you please sign your posts on talk pages. --TimLambert 04:47, 16 May 2006 (UTC)


 * This is going nowhere and is a waste of time. As I said previously, I am not determined to fight your efforts at suppression and manipulations on this page, or try to do battle with your collosal ego. There doesn't seem to be any point in such a forum, and I am far less fanatic and controlling about this topic than you appear to be, so I would not have the stamina. So, essentially you are getting your way. Declare victory, and let's call it a day.
 * Are you JoshD? You certainly write like him.Felix-felix 13:34, 16 May 2006 (UTC)


 * Attacking character and questioning the identity of an individual does not win the argument, stating your position while rebutting the facts stated by the opposition does. I don`t edit anything on/in Wikipedia, only find humour in the vacuous arguments and childlike behaviour so consistently exhibited here.  Wikipedia, looking more and more like Youtube everyday. --Tunabun 22:26, 18 May 2006 (UTC)

Confusing paragraph
"The main debate in the media in the U.S. and U.K. focused on whether 98,000 (95% CI 8000–194,000) excess Iraqis were killed as a result of coalition intervention, and that statistic excludes all the Fallujah data. The authors estimate that 98,000 more deaths than expected happened after the invasion outside of Fallujah. Had the Fallujah sample been included, the survey's estimate would have been of an excess of about 298,000 deaths, with 200,000 concentrated in the 3% of Iraq around Fallujah (Roberts et al p.5). The study found that mortality rates had increased about 2.5 times since the invasion (with a 95% CI 1.6-4.2) including the Fallujah data, and 1.5 times (95% CI 1.1-2.3) excluding the Fallujah data (from 0.5% to 0.79% per year). According to the article, violence was responsible for most of the extra deaths whether or not the Fallujah data was excluded, with coalition airstrikes the main cause of those violent deaths."

The last sentence is particularly confusing. When "Falluja data was excluded" there were 21 violent deaths. Only 9 of these were attributed to the coalition. How could "coalition airstrikes" be the "main cause of those violent deaths" "whether or not the Fallujah data was excluded" when the coalition in its entirety, let alone just by airstrikes, didn't even account for half the deaths when Fallujah was excluded?

The article seems misleading itself on this point, so perhaps one could argue that it is true "according to the article" but just not according to the data. However, the paragraph is misleading in any case. 68.45.226.214 04:45, 17 July 2006 (UTC)

Also, is the paragraph below with or without Fallujah:

"Although forty six percent of the violent deaths involving coalition forces were men between ages 15 and 60, considered possible combatants by the authors, seven percent were women and fully forty six percent were children younger than 15, who they considered to be in all probability noncombatants." ???

Deaths under Saddam
The Lancet article states that the most common causes of death during the rule of Saddam Hussein were illnesses such as heart disease etc. What about the victims of Saddasm wars? and the tens of thousands of people found in massgraves? And all those still missing ? --Vindheim 11:05, 24 July 2006 (UTC)

Hi vindheim. I think the point is that the Lancet articles (both of them, I think) only use the year previous to the 2003 invasion as a baseline. I guess we assume that most of Saddam's killings were done during wars in the 1980s & the Gulf War, and defenders of the Lancet would say that he must have been behaving himself very well during 2002-3. Even still, looking at the figures for pre-war violent deaths, they do seem low.--Lopakhin 16:02, 19 October 2006 (UTC)


 * The average death toll during Saddam's reign, and counting only deaths directly caused by Saddam and his regime, was approximately 275 per day, over the entire twenty years that he was in power. The greatest number of his victims were Iraqis, followed closely by Iranians, plus smaller numbers of Kuwaitis and others.  That's just counting those who died in his wars of aggression, purges, genocidal mass murders, etc..  It does not include victims of common crims (which IBC includes in their counts), or other "excess deaths."  ~275/day is just the number of victims who were murdered by Saddam and his henchmen, or who died in Saddam's wars of conquest.  (The daily average is calculated by dividing the approximate total death toll, 2 million, by the numbers of days in 20 years.)


 * The truth is that the "excess death rate" following Operation Iraqi Freedom is negative, if compared to the average death rate during the 20 years before Operation Iraqi Freedom. NCdave 01:04, 25 October 2006 (UTC)

POV concern
I am concerned that this edit does not properly balance opposing points of view. I object to the wholesale removal of sourced statements crucial to the central points of the article without even copying them here to the talk page. Accordingly, I am adding a POV tag to this article. Please discuss the issue here, and do not remove the tag until a consensus is reached. Thank you. Starcare 19:31, 27 July 2006 (UTC)


 * I guess it's my fault for just trying to scrape by with an edit summary instead of taking the time to discuss it here. My main gripe with the paragraph was the "at least", which I tried to point out is meaningless in a statistical/epidemiological sense. Even if it were some sort of reference to the lower confidence limit, that's way down, the 100,000 and 285,000 are the most likely numbers, i.e. the centers of the confidence intervals, not the "at least". I think what was trying to be said was that the best estimate was being reported as 100,000, when actually that was a conservative best estimate, the actual best estimate without dropping Fallujah would be 285,000. No "at least" involved. But then the "at least" quote from Roberts was a problem, since he certainly knows his business, and it'd look a bit weird for me to be correcting him, so I went to the link on zmag which quoted him in the independent, so i went to the independent and searched for his letter they cited (geez they have a buggy search) as well as searching for it via google, in order to see exactly what he said and maybe it was misquoted, and got nowhere in either place, so I just gave it up and cut the paragraph since the "at least" bit was wrong, I'd stake my life that the Roberts quote with the "at least" quote is wrong, I'm in no position to be publicly contradicting Roberts even if I'd stake my life that his "at least" quote is wrong, I couldn't find the original quote to make sure that that's what he said anyway, and the whole Fallujah/conservative bit was discussed below, after all. So that's the story, no POV issues as far as I'm concerned as far as pointing out that the 100,000 was a conservative modification of the plain and simple estimate. Gzuckier 22:19, 27 July 2006 (UTC)
 * I don't see a problem with that-no matter what Roberts said-the 100,000 figure can't be at least in a statistical sense-even if common sense tells you that the true figure is very likely to be much higher. Thus I think that, regardless of quotes by Roberts (true or not) removing the at least is more encyclopedic.Felix-felix 07:55, 28 July 2006 (UTC)

This letter from the authors of the study defending their findings was published December 12 in the UK newspaper The Independent on Sunday:

Here is the letter: http://timlambert.org/2004/12/lancet11/

Iraq toll: at least 100,000

You reported (”Polish hostage held in iraq is released unharmed”, 21 November) the Foreign Secretary’s response to our study published in The Lancet of civilian deaths in Iraq. It is heartening that Jack Straw has addressed the topic in such detail. However, his response includes an apparent misreading of our results.

Our study found that violence was widespread and up 58-fold after the invasion; that from 32 of the neighbourhoods we visited we estimated 98,000 excess deaths; and that from the sample of the most war-torn communities represented by 30 households in Fallujah more people had probably died than in all of the rest of the country combined.

Fallujah is the only insight into those cities experiencing extreme violence (ie Ramadi, Tallafar, Fallujah, Najaf); all the others were passed over in our sample by random chance. If the Fallujah duster is representative, there were about 200,000 excess deaths above the 98,000.

Perhaps Fallujah is so unique that it represents only Fallujah, implying that it represents only 50-70,000 additional deaths. There is a tiny chance that the neighborhood we visited in Fallujah was worse than the average experience, and only corresponds with a couple of tens of thousands of deaths. We also explain why, given study limitations, our estimate is likely to be low. Therefore, when taken in total, we concluded that the civilian death toll was at least around 100,000 and probably higher, not between 8,000 and l94,000 as Mr. Straw states. While far higher than the Iraq Ministry of health surveillance estimates, on 17 August the minister himself described surveillance in Iraq as geographically incomplete, insensitive and missing most health events.

We, the occupying nations, should aspire to acknowledge the dignity of every life lost, and to monitor trends and causes of deaths to better serve the Iraqis, and in doing so, sooner end this deadly occupation.

Les Roberts, Gilbert Burnham Centre for International Emergency, Disaster and Refugee Studies, Johns Hopkins Bloomberg School of Public Health, Baltimore; Richard Garfield School of Nursing, Columbia University, New York, USA 68.45.226.214 10:19, 28 July 2006 (UTC)