Weapons Of Math Obstruction
We’ve been hearing ah-oogahs, boings, and sawing noises coming from Malkin’s place lately. Let’s see what it’s about.
Document drop: A new critique of the 2004 Lancet Iraq death toll study
By Michelle Malkin • July 25, 2007 11:01 AMOne of the most useful roles of the blogosphere is its service as an open-source intelligence-gathering medium. You can draw on the expertise of people around the world at the touch of a button. We saw this with typography experts during the Rathergate scandal; Photoshop experts during the Reutersgate debacle; and military experts during the Jesse Macbeth unmasking.
Oh, it’s that whole thing again. Indeed, let’s sum up their accomplishments of the past three years:
1) Having apparently been spoon-fed information by Republican campaign operatives, they announced that several documents obtained by CBS regarding George Bush’s military service were inauthentic — and have been bwaa-haa-haaing ever since, claiming this as proof that the mainstream media is hopelessly corrupt and obsolete, and that the right-blogosphere represents the future of journalism.
2) After showing that a photographer crudely added extra smoke to a Reuters photo of the bombing of Lebanon, and after Reuters summarily fired the guy, they whooped and yelled endlessly that the wire services are in league with Arab terrorists, that any photograph representing something they don’t like is potentially a conscious forgery perpetrated by enemies of America, and that the right-blogosphere represents the future of journalism.
Above: Slightly-More-Smokegate
3) Some sucker MC claimed to be an Iraq War vet when he really wasn’t, and told some hair-raising stories of atrocities, not only in several alternative-media videos, but also via a number of Kos diaries. Having unmasked him as a fraud, they integrated that fact seamlessly into their fulsome years-long victory dance that everyone is a liar except them, that everything is phony except stuff they agree with, and that, oh right, wait, now that you mention it, the right-blogosphere represents the future-uture!-uture! Of journimalism-alism!-alism!
Because when you have one fool trying to solve a problem, foolishness results. But get 10,000 idiots on the case, and somehow one of Rupert Sheldrake’s morphic-resonance fields comes zonking into the is-ness, radiating quantum smartrons.
…As, you know, their track record amply demonstrates, with these three monumental achievements towering out of the daily whooping wrongness and mortification of everything else they ever say, practically down to the level of words such as ‘and’ and ‘the.’
Now, it’s the statisticians and math geeks’ turn.
Yippee!
Remember that massively-publicized 2004 Lancet Iraq death toll study? It was cited in nearly 100 scholarly journals and reported by news outlets around the world. “100,000 Civilian Deaths Estimated in Iraq” blared the Washington Post in a typical headline.
But since 100,000 dead civilians would make America, i.e. Conservatism, i.e. Michelle Malkin (i.e. America) look bad, this must by definition be a false and dishonest study outrageously concocted by blatant enemies — who will now be justly exposed for the good of America (i.e. Michelle Malkin). They just haven’t found out how yet, is all.
There were attempts made by lay journalists to debunk the 2004 study (as well as the 2006 follow-up study that purported to back up the first). But none of those dissections comes close to a damning new statistical analysis of the 2004 study authored by David Kane, Institute Fellow at the Institute for Quantitative Social Science at Harvard University.
Michelle then admits that she doesn’t understand the math (but likes the conclusions), and plops the whole paper, equations and all, in front of her readers for an evaluation. These readers are, as we have seen many times, a self-selecting group of extraordinary gullible cranks who can be counted on to know less about any given thing than a comparable group of pikers in the buffet line at Foxwoods.
See, now that would be some comedy. What I was going to do was surrender to the joy of it, pronouncing our own judgment on the paper (with accompanying diagrams) roughly as follows:
The printing looks funny.
It has extra smoke.
The author of the paper is actually a phony masquerading as a paper-author.
But in my happiness I went over to Tim Lambert’s place, where Malkin first found out about the paper, to see what the scientists might be saying about it.
And dag! Here’s Daniel Davies ruining all the fun. The ‘David’ he’s addressing is the author of the paper, David Kane, who’s also in comments there:
David, at present your entire paper rests on your assumption that the reported confidence intervals are based on parametric estimates from a unimodal distribution. The paper says pretty specifically that they were calculated by bootstrapping from a bimodal empirical distribution. This is a big problem for your paper. The nature of the dataset makes it very clear that the bootstrap was the correct way to calculate the confidence interval for the risk ratio. That is another big problem for your paper. Finally, your paper has the implication that it would be sensible for a statistician to conclude that the discovery of mass deaths in Fallujah is evidence in favour of the proposition that the death rate had fallen. That’s the really, really big problem for your paper.
Posted by: dsquared | July 26, 2007 03:48 AM
It would be useful at this point to do a Shorter Davies. This is something of a hall-of-mirrors moment, since Davies is the inventor of the Shorter.
Shorter Daniel Davies
Above: Also known for other things
- David, you dolt, the Lancet study clearly says they used apples, and you keep being all like, ‘but if they actually used oranges, then maybe Iraq is a magical pony kingdom.’ Cripes, are you completely retarded?
‘Shorter’ concept created by Daniel Davies and perfected by Elton Beard.
Meanwhile, in a move that surprises no one, Malkin has already updated the post, anointed a champion, and declared victory:
Update 9:30pm Eastern. Shannon Love at the Chicago Boyz blog called foul on the Lancet 2004 study early on and, with vindication, reacts to David Kane’s new analysis of the 2004 Lancet Iraq death toll study: “Kane shows that if the Falluja cluster is included in the statistical calculations, the confidence interval dips below zero, which is a big no-no. Since the study’s raw data remain a closely guarded secret, Kane cannot be absolutely certain that the inclusion of the Falluja cluster renders the study mathematically invalid…but that’s the way to bet. In science, replication is the iron test. I find it revealing that no other source or study has come close to replicating the original study. All my original points still stand. Ah, vindication is sweet.”
In other words, he sums up the claims in the paper and says ‘yay.’
Malkin and her sort know that it isn’t necessary, generally speaking, to win many arguments in order to control the debate around an issue. It’s often sufficient simply to sow confusion, to blow so much fog around that no one knows, anymore, which side is right or wrong.
Look for new claims by the usual right-wing foundation hacks that the Lancet study has been ‘definitively debunked,’ and chalk up another petty, rigged ‘victory’ to the distributed intelligence of the WingNet.
Shorter Malkin: Now I don’t understand the math in this paper, and I don’t understand the math in the original paper, but one of them re-asserts my pre-conceived notion of what happened, so I’m going with that one.
If you don’t stop bootstrapping from a bimodal empirical distribution, you’ll go blind.
You’re going to need to scrub much harder than that, Lady Malkin.
Um, so what does this superfantastic debunking mean? Does it mean that there are no civilian deaths in Iraq and it’s all happy magical pony kingdom instead?
Yay! AMERIKA1111!!!!WE WIN!
Shorter Gavin M:
Michelle Malkin… the pundit whose name is the punchline.
Daniel Davies is smart and good-looking and red-headed and all, but if he created the ‘Shorter’ concept, who’s going to explain it to these people?
“In science, replication is the iron test.”
I love that sentence.
In sports, winning is encouraged!!
In business, making money is fun!!!
thus proving that a double distillation further concentrates teh funny.
I’m sure all the dead Iraqis will be very happy to know that their deaths are no longer operative.
Snark is one thing, but the sickening thing is that these people are dancing over the bones of the civilans that have died as the result of Bush’s invasion. The Lancet study points out that the numbers they represent are in excess of the mortality rate undeer Hussein, so in any case, America has managed to increase the kill ratio…
And these moral black holes are giddy becasue one off the wall number cruncher came up with a way to view the data that shows a slightly lower rate than previously thought.
Or does the new Magic Pony Report also magically revive all the war dead?
g did a shorter billy pilgrim, and got it in before me.
well done g.
Question: WHY does the Lancet hate us for our freedoms???
Me, I’m done reading it, for its obvious liberal slant and total blindness to Freedom On the March, the Sacrifice Of Our Brave Men and Women In Uniform, the Great Purple Finger of Democracy and so on.
Peer review means nothing, now that the “peers” are a bunch of liberals!
When will the *Discovery Institute* start publishing a medical journal? I want to read about the efficacy of prayer in end-stage cancer; and of the elevated morbidity/mortality risk associated with being Muslim and hating us for our freedoms, etc…
Teh stupid – it BUUUUURNNNNS!!
I am honored, billy.
Sadly for Malkin, to quote Mr. Kane, when asked how many deaths he thought had been caused by the Iraqi invasion, he responded by saying:
“the more that I study this topics, the more convinced I am too (sic) trust Jon Pedersen’s judgment: 100,000 excess violent deaths.”
So, Shorter Malkin: “I’m wrong, but I might be wrong in an entirely different way, so I win!”
Wait–“Rupert Sheldrake” is a real person? I just Googled. I could have sworn he was a character in a movie played either by George Sanders or Peter Sellers.
Come on. Seriously?
I notice in MM’s list of blogspheric triumphs, she conspicuously leaves out the “still standing” Mosque. Huh. Funny that.
For those of us who can’t really follow the maths very well, I suspect this comment by SG in the Tim Lambert blog sums it up:
And, in a triumph for Sadly, No!’s Shortenings, Daniel Davies has really gone with the ‘not apples, but apples and oranges’ metaphor, to try to explain to Kane in simple terms why he’s wrong.
A character in a movie played either by George Sanders or Peter Sellers.-
Are you thinking of Col. Mandrake in Dr. Strangelove?
I’m not very good with the math stuff but is MM saying that if you change the dataset then the original study, done with a different set of numbers, can’t be replicated?
I find it the height of scientific rigor that Michelle has promoted a paper as disproving a published, peer reviewed article, before said paper is: 1) peer reviewed or 2) published.
I find it the height of scientific rigor that Michelle has promoted a paper as disproving a published, peer reviewed article, before said paper is: 1) peer reviewed or 2) published.
Fair use, my friend, fair use all the way!
Kane commenting on an EphBlog post on the subject (my emphasis):
Sadly, No! at your service, Mr. Dr. Purple Cow.
Shorter Malkin-hyped paper: “but…but…but…what if there are ZOMBIES in Iraq!!!”
As a commentator points out regarding the paper she is hyping, “The Figure 2 diagram actually has 1.34% of its probability mass represented by states in which Iraqis were being resurrected from the dead.”
Basically, the other uses a technique which states that if you include a Falluja type death cluster, you must assume that an opposite Falluja is also possible. Thus, by increasing how many people died in Iraq, its possible less people actually died (because you have the negative deaths from the opposite-Falluja).
Thus, death rates in Iraq are actually lower than before the war assuming you have a 1.34% chance of rising from the dead.
I think the important question is, did Michelle post a video giving a cheer of vindication?
“Kane shows that if the Falluja cluster is included in the statistical calculations, the confidence interval dips below zero, which is a big no-no.”
Google search results for “big no-no”:
* Cyber-Pet “Chlorinated Water…A Big No-No”
* Sharpton to TMZ: “Roboho” a Big No No
* Ambien + Phenergan + Driving = A Big No-No
Apparently there are much bigger no-nos. The Internet contradicts Kane!
One fine day, about midnight
Two dead boys got up to fight
Back to back they faced each other
Drew their swords and shot each other
A deaf policeman heard the noise
He went and killed those two dead boys.
I hope that clears it all up for you…
mikey
“Kane shows that if the Falluja cluster is included in the statistical calculations, the confidence interval dips below zero, which is a big no-no.”
Seriously, this is some stupid “crossing the streams” = “total protonic reversal” bullshit. But that’s what Michelle deserves for getting statistics assistance from a bunch of homies calling themselves “Chicago Boyz.”
If Kane’s confidence interval dips below zero, it would just mean that his model isn’t very good — mostly because the model Kane is fitting — a theoretical normal distribution with long positive and negative tails — is inappropriate for data which must be positive (death counts). It’s even worse because the data is clearly bimodal (Fallujah).
Surprisingly, the Lancet authors were aware of this. Harvard’s Kane seems to have screwed it up big time.
pfffffft-BWA-Ha-HAAAH-hahahaha. Oh holy crap. Oh my God. Wow. I had no idea JC Christian was ghostwriting Malkin’s columns these days.
Oh god, Mikey, I had a friend in school that used to love to repeat that little singsong. I’d forgotten all about it…
I may not be a mathemetician, but I’ll bet that I’ve had more training in math and statistics than Malkin; and from what I read at the time, it seems to me that the Lancet study authors did excellent work in using statistical modeling to correct for the error factors inherent in what they were trying to estimate. and being mathemeticians, not political operatives, they did it with an end result in mind: getting as close to reality as possible. The reviews of the study bear this out, and it is only among US based conservative oriented ‘Think Tanks’ that any questions about the methodology have arisen.
Of course, MM and fellow-ghouls, being political operatives, have no interst in the truth. Never have, never will. The truth only exists to be manipulated into ideological supports.
It’s been said many times, but only because it gets truer every single day: For these people, 1984 was a blueprint, not a cautionary tale.
Why is it the war-bloggers obsess about anything but coming up with a concrete plan to actually win the damn war?
“Kane shows that if the Falluja cluster is included in the statistical calculations, the confidence interval dips below zero, which is a big no-no
Ho-kay I am no good at the maths, mostly because, while yet a young gatito, the extra toes led me to start counting from base twelve rather than base ten. (stupid dewclaws)
However I remember hearing on the BBC een 2004 that they threw out the Fallujah numbers because they were so high that to include them would have skewed the dataset.
¿Ees no that what statistical studies do? ¿Throw out the very highest and lowest outliers so they can show a more accurate result?
Por favor, I am curious, I would like to know.
(Joo know how we gatos get, eh?)
so.
This comment by “sortition” says it all.
“dsquared writes:
In general, there’s a couple of points in this thread (“measure the crude mortality rate using a normal distribution” being a prime example) where you seem to be quite confused about the difference between estimating a quantity of interest and estimating the confidence interval of the estimate [should be “of the quantity of interest” – Sortition].
Kane answers:
You are correct to note that my terminology in this thread is a bit sloppy. Apologies. Do you see similar mistakes in the paper? I hope not! If so, please point them out.
As I pointed out several times, you make that mistake repeatedly throughout your paper. You keep attributing a distributions to your parameters. This is simply wrong.
As dsquared mentions: CIs have a distributions since they are random, parameters do not since they are not.”
Daniel Davies looks disconcertingly like Joss Whedon.
Because remember: it’s only the EmmEssEmm that’s biased. We report; you decide.
And here’s the money quote from the comments on Malkin’s site:
We all figured the Lancet Iraq Death toll study was bogus. It’s nice to have a scholarly paper to help validate our doubts.
“We all assumed that the study was bogus, because it made our desire for war seem like a bad thing, what with all that extra death and stuff. So we rejected it and called the writers treasonous liars and all that jazz. And now someone wrote a science-y paper that might support our beliefs? BONUS!”
I wonder if Malkin’s commenters’ heads would explode if they went over to Deltoid and read the colloquy. My bet is that they would immediately accuse everyone who is arguing against Kane’s paper over there of being leftists, traitors, etc.
The funny thing is, none of this would be necessary if Malkin’s heroes, those fine humanitarians bringing rich, creamy freedom to Iraq, hadn’t said: “We’re not even going to count how many civilians we kill; that’s how much we don’t give a flying fuck.”
EGN:
In this case, they threw the high outlier out. As you can see here, the average yearly mortality rate of everything but Fallujah looks to be about 15 people per 1000(I think the Lancet study put it at 12.8 per 1000). Fallujah is somewhere around 200 people per 1000. That is definitely an outlier, and the Lancet study did not include it.
What Kane’s paper is saying is that if you include the outlier that screws your data up, your data are screwed up. It’s pretty damn misleading.
In other words, [Shannon Love] sums up the claims in the paper
Would that he did. Although I frankly don’t rate that paper at all, and I am *really really annoyed* that David Kane was soliciting comments on it while simultaneously distributing the unaltered version, David Kane does at least have the baseline of statistical knowledge that would prevent him from saying that “the confidence level dipping below zero is a big no-no”. That’s pure S. Love.
After reading the Deltoid thread —
David Kane is far from suggesting that the 2004 Lancet study was “bogus”. He notes that the more that I study this topics, the more convinced I am [to] trust Jon Pedersen’s judgment: 100,000 excess violent deaths. He is arguing, in good faith, that the confidence interval was too narrow.
Some things should not ever be mentioned in the same sentence, so it is necessary to write
Malkin (“damning new statistical analysis”).
Good faith
on separate lines.
In science, replication is the iron test. I find it revealing that no other source or study has come close to replicating the original study.
— Except (alas) the Lancet 2006 paper.
More from Kane in the comments to the EphBlog post I linked to above (emphasis his):
Malkin has linked to scienceblogs.com twice: yesterday and back in March 2006, to a P.Z. Myers post on Ben Domenech. Sure, she reads Deltoid.
your entire paper rests on your assumption that the reported confidence intervals are based on parametric estimates from a unimodal distribution. The paper says pretty specifically that they were calculated by bootstrapping from a bimodal empirical distribution
I like kittens. And bunnies.
Me too. Bunnies particularly, in a hot sauté with peppers and wild mushrooms, served over rice. Kittens need a longer method, a stew or at least a braise, with potatoes and root vegetables to smooth out the “tang”…
mikey
Well I guess the Iraqi Board of Tourism has their new slogan, thanks to Ms. Malkin:
Vacation in Iraq…
You’ll come for the casualties but stay for the ponies.
It’s funny to read that Deltoid thread and see how Kane’s tone alters the minute dsquared gets into it. Within a few replies he’s squirming out of his early certainty like snake from its skin.
My eyes crossed reading this, but unless I’m mistaken, it’s all about how to handle small sample sizes when you discover, after the fact, that you’ve got a gorilla-sized outlier on your hands. Having done this kind of modeling once, I recommend the following: throw out Mitch Richmond’s stats–they’ll just hurt your regression of how your 1993 hoops fantasy league values its players.
I have a headache now.
The Deltoid thread is comedy gold.
Tim Lambert: “David, do you really think that there is a non-zero probability that the post-war CMR in Iraq was less than 2.2?”
David Kane: Yes!
As Robert pointed out in the thread, this would imply truly astronomical life expectancy rates which far exceed the ages of the oldest people who have ever lived.
Internet Explorer may not respond when you click Print or Print Preview in the File menu.When you try to connect to Web folders, you may receive the following error message:The current operation could not be completed because an unexpected error has occurred.
Here’s Kane’s bio:
David Kane is the CEO of Kane Capital Management, a quantitative, market-neutral, global equity hedge fund. He has a Ph.D. in Political Economy from Harvard University and a BA in Philosophy and Economics from Williams College. David served as an officer in the United States Marine Corps from 1988 to 1991. He was a lecturer in Government at Harvard in the fall of 2002. David’s research interests involve applied empirical finance using open source tools, especially R, the use of which he introduced to the Government Department. David is a member of the Eliot House Senior Common Room.
What I’d like y’all to notice is that Kane’s qualifications are in Political Economy, Philosophy, and Economics. Not, as you might imagine, maths or stats.
Jebus, some folks are stupid. Speaking as someone with a maths degree, I spit upon such noodles. I turn my furry behind to their faces and waggle my tail. For anyone to suggest that some goddamned economist (a profession I consider only slightly less icky than prostitution) has the right to dispute a paper by a statistician, published in the fucking Lancet, no less, fills me with righteous feline ire.
And to then seize upon his paddlings and claim that his is just as good as theirs, even though they’ve subjected theirs to rigorous peer review (and even to be allowed to submit something to the Lancet, you’ve got to be pretty damned speccy) and he hasn’t…well, I think I’ll just cough up a furball in disgust.
Here’s an equally hilarious thread about Lancet, but this time the other way around…
http://scienceblogs.com/deltoid/2007/03/lancet_post_number_135.php#comment-389419
Still the same Pulitzer prize waiting to be collected for the first journalist anywhere to write the story that supports what Lancet’s estimates suggest….
http://www.iraqbodycount.org/press/pr14.php
It appears the media conspiracy on Iraq is real and every reporter, from every country on the planet and every political persuasion is in on it.
Or, you know, those Lancet estimates deserve the criticism they receive.
Well yes that is funny. Especially considering that pretending you believe in clearly BS death rates is a prerequisite for defending the legitimacy of those Lancet estimates.
Lancet states their surveyed deaths correlate with official death certificates to a 90% degree. Those estimates are the ones that suggest only 1 in 10 violent deaths in Iraq is recorded by any means at all, let alone official records.
The only explanation other than the Lancet estimates being a deliberate falsification is that the pre-war death rates they used as the basis for the whole study are BS and far too low which accounts for both scenarios.
I cannot honestly say whether it is funnier to watch people pretend they see no problem with claiming a 90% corroboration with a source when there’s only 10% or that this isn’t a significant different, or watch them choosing to highlight clearly BS death rates to embarrass a random critic of these Lancet estimates when this also happens to invalidate each Lancet estimate as ridiculously overinflated.
It all looks like comedy gold from here. Hilarious eh ?
What’s really hilarious, Kilo, is that the British government’s statisticians warned Blair not to challenge the Lancet estimates because they had independently concluded they were almost certainly good estimates.
Of course, your complete incomprehension of the details of the math (combined with a “knowing” pomposity) is fun too.
What I’d like y’all to notice is that Kane’s qualifications are in Political Economy, Philosophy, and Economics. Not, as you might imagine, maths or stats.
But it would be sad if you were not allowed to argue about possible weaknesses in someone’s research unless you held actual qualifications in that field. That rule would force me to keep my mouth shut more often than is good for me.
—————————-
Here’s an equally hilarious thread about Lancet,
I disagree. The Deltoid thread was sadly lacking in snark. The participants took Kilo seriously, and wasted their time making civil responses when withering sarcasm would have been more appropriate. A squandered opportunity.
Qetesh and Herr Doktor Bimler: If you haven’t already seen it, this Crooked Timber discussion is worth a perusal as well.
I don’t think you’ve quite understood the reasons for the references to humor here. Nor what you are referring to.
The British advisers supported the methodology used in these estimates, something that both I and Lambert do.
So who’s the joke on ? Yourself apparently.
The point you’ve missed is that we would probably also agree on the way to calculate the circumference of a circle. Yet this methodology would still produce a bad result where inaccurate data was used for such a calculation on which all conclusions relied.
Which, as I mentioned, is what would account for the estimate contradicting itself and producing the anomalies seen.
And much like questions over the methodology used, it would appear that once again there is no disagreement between yourself, Lambert and I that these contradictory figures are indefensible.
See, you’ve got to make it funny like I did there, it doesn’t just happen on its own.
You can’t just say “A man walks into a bar hahahaha”. It doesn’t work.
Yeah you tend to get a lack of whithering sarcasm when getting dressed down for defending something you dedicate the other 364 days a year criticising.
For those who are still interested — the Deltoid thread in question was mainly concerned with the discrepancy between the incidence of violent deaths in Iraq according to official government records, and the incidence estimated by cluster sampling. Government figures were systematically underestimating the death toll… or the sampling systematically over-estimated it… or both. I say “systematically” because the figures correlate well, despite differing by an order of magnitude.
Highlights of the thread include comments #29 and #34, pointing to other surveys of the Iraqi population which confirm that government figures massively under-report the amount of violence. Several commentators point out that it would be easier to believe in the sincerity of the Lancet study’s critics if they cared about the Iraqi population enough to set up a better study.
In comment #26, near the end of the thread, Kilo requests that defenders of the Lancet study should make “even a single suggestion (we don’t even have to start with a plausible one) for how Lancet estimates are not ridiculously high if they suggest a rate of violent deaths 10 times higher than figures based on the number of Iraqis recorded as killed” (ignoring previous comments such as #13). Abe G. responds in comment #30 by listing 9 suggestions. At this point, Kilo packs up and went home.
Kilo then turns up here to suggest that the Lancet researchers simply fabricated their data. Also to argue that these figures for violent death don’t matter anyway, because the figures for excess deaths also depend on the estimated pre-invasion death-toll — which are apparently “BS” (repeated three times), despite coming from official CIA sources IIRC.
What do you mean still ? Has someone here shown any interest here in validating or invalidating such discrepancies so far ?
That order of magnitude being the rather large one mentioned, which nobody here nor there wants to been seen pretending they believe is possible, even when posting anonymously on the interweb.
Let’s all try and guess why.
If finding news media surveys which support an underreporting of violence then presumably the low-light would be other news media surveys and audits of hospital and ministry of health records which find there’s nothing to support the Lancet’s credibility gap on those same records. Only if you mention it of course. Otherwise you get the confirmations sought. Such is the efficiency of cherrypicking and exclusion.
I don’t recall that argument being made and it sounds quite ridiculous on the face of it. There were several Deltoid readers who suggested a statistical survey with robust methodology would be easier to believe if the authors cared more ?
Indeed he did. Along with zero suggestions for anyone believing these were the cause, himself included.
Add to that number the amount of people who not only believe such reasons (ie. inadequate training for hospital staff in identifying live vs dead bodies, 9 out of 10 Iraqis not wanting to keep their dead husband’s house or claim death benefits) could account for such systemic underreporting, but simultaneously account for such excellent reporting that they validate 90% of the Lancet’s deaths…
and you still have zero.
No, I suggested that apart from fabricating their data there was only one explanation that would account for such correlation with current death figures, yet such massive divergence from the same data you claim corroboration with when extrapolating these estimates.
There’s been no disagreement with this assessment nor alternate explanation offered to date, so I’m unsure why you mention it.
Yeah. For the reason specified. Again, something you nor anybody else appears to disagree with.
Okay, Kilo.
Here’s the challenge for your peanut brain.
You live in an outlying suburb of Baghdad. Your child dies. Do you…
a) Cart his body across town, through several roadblocks, to the morgue, to get a piece of paper, or…
b) Bury him
Let’s see if you can figure it out. Okay?
Lancet states their surveyed deaths correlate with official death certificates to a 90% degree. Those estimates are the ones that suggest only 1 in 10 violent deaths in Iraq is recorded by any means at all, let alone official records.
Here’s the difference between a survey and a complete tally: you don’t have to catch everything in a survey; that’s what makes it a survey. The numbers for the percentage of recorded surveyed deaths and the percentage of the total recorded deaths do not have to be equal. At all. Ever.
It appears the media conspiracy on Iraq is real and every reporter, from every country on the planet and every political persuasion is in on it.
Every country? The Iraq Body Count website is based off of the figures reported in the Western media. “There is a world elsewhere” and when it’s in one’s backyard, the reporting tends to be a bit more detailed. While the war was on, and thereafter, I read articles in the Arab language media which were not recorded in the Iraq Body Count website, to such a degree that I simply gave up on the site entirely. I have no idea what its current “count” is, but you can be damn certain that it’s underreported.
To use their website, given where they get the information from, would be like claiming that there were practically no deaths in Fallujah during the major bombing campaign in that city because almost all reporting teams, except for (IIRC) Al-Jazeera, were turned out from there.
Okay, Kilo.
Here’s the challenge for your peanut brain.
You live in an outlying suburb of Baghdad. Your child dies. Do you…
a) Cart his body across town, through several roadblocks, to the morgue, to get a piece of paper, or…
b) Bury him
Let’s see if you can figure it out. Okay?
Especially considering that janazah should be accomplished as soon as possible (this refers to the whole process of burial) in order to avoid embalming or otherwise disturbing the body. When it’s 40 or 45 degrees C outside, unembalmed bodies do rather tend to decay rapidly. It’s not nice.
Has someone here shown any interest here in validating or invalidating such discrepancies so far ?
In a word — No.
In four words — No, quite understandably not.
The key issue is whether the Iraqi government or a network of researchers has more credibility. To address this concern, we ask ourselves questions like
(1) Which source of information is propped up by an occupying army?
(2) Which source of information is cobbled together from a tangle of opposing interest groups, ranging from the merely venial to those that operate torture chambers as part of the Ministry of the Interior?
IMO, there are sufficient reasons not to take official Iraqi govt. death statistics seriously. I am not saying that they are completely divorced from reality, just that they are undergoing a trial separation.
Nullifidian appears to not know what he/she is talking about:
“The Iraq Body Count website is based off of the figures reported in the Western media.”
Checking the sources brings up:
Al-Zaman
Al-Zaman
Al-Alam TV
Al-Alam TV
Al-Arab
Al Arabiya TV
Al-Furat
Al-Furat
Al-Ittihad
Al-Ittihad
Al-Jaz
Al Jazeera (Web)
Al-Jaz TV
Al Jazeera TV
Al-Shar
Al Sharqiyah TV
Al-Taakhi
Al-Taakhi
ALB
Al-Bawaba
Arab N
Arab News
Arabic N
Arabic News
ArN
Arabic News
AS
Asahi Shimbun
Asharq Al A
Asharq Al Awsat
ASB
As-Sabah
AST
Asia Times
AT
Arab Times
etc. etc. etc.
I don’t know what Null thinks constitutes the “Western media”, but it seems to not be what anyone else would mean by the term.
“I read articles in the Arab language media which were not recorded in the Iraq Body Count website, to such a degree that I simply gave up on the site entirely.”
Yeah, right.
Kilo, otoh, seems to have a brain in his head:
“Lancet states their surveyed deaths correlate with official death certificates to a 90% degree. Those estimates are the ones that suggest only 1 in 10 violent deaths in Iraq is recorded by any means at all, let alone official records.”
Exactly. The Lancet numbers are obviously horse-hockey unless you buy a huge conspiracy to suppress and hide all this paper trail (and all those bodies), and which has completely fooled everyone everywhere except the Lancet. Please. The Lancet is a fluke or a fraud.
Okay, Kilo.
Here’s the challenge for your peanut brain.
You live in an outlying suburb of Baghdad. Your child dies. Do you…
a) Cart his body across town, through several roadblocks, to the morgue, to get a piece of paper, or…
b) Bury him
Let’s see if you can figure it out. Okay?
If the 90% death certificate rate in Lancet is right, people do “a” about 90% of the time.
Some reasons why so many people would tend to do “a” can be gleaned from this:
http://www.unhcr.org/cgi-bin/texis/vtx/home/opendoc.pdf?tbl=SUBSITES&id=462cbf672
Death certificates
Death certificates, needed in order to obtain retirement benefits for the spouse or children as well
as for inheritance purposes, are issued by the MoH Births/Deaths Administrative Offices located in Public Hospitals. Death certificates are usually issued within two days.
Required documents are:
Medical report;
Civil ID card or food ration card of the deceased person;
Residency card of the deceased person.
The issuance of a death certificate is free.
If the 90% death certificate rate in Lancet is right, people do “a” about 90% of the time.
Actually, too generous. I should have pointed out that the “a” in Null’s false dichotomy was just wrong to begin with. The above indicates that you don’t need to “cart the body” around to obtain the “piece of paper” that Lancet says people do obtain for 90% of deaths. You need a medical report and the person’s id documents.
Yay, Kilo has a playmate!
Thanks for summarizing the Deltoid discussion, Herr Doktor. dsquared has a new post at Crooked Timber if you’re interested and haven’t seen it yet.
Nullifidian appears to not know what he/she is talking about:
“The Iraq Body Count website is based off of the figures reported in the Western media.”
Well, it appears the Iraq Body Count site agrees with me:
Out of all of these, only five are media from the Arab world. The inclusion of two Indian newspapers doesn’t change the overall Western thrust of their count, and of all the sources they have listed, they’re all in English or also available in English.
I don’t know where you got your list, but I can see why you didn’t cite it. It seems like you just compiled it out of thin air from a listing of Arabic language media.
As for your “Yeah, right.”, how do you know what I have and haven’t read?
Although not entirely Arabic-language media. Asahi Shimbun, for example, is definitely not a Mid-East paper. It also has nothing to do with the Iraq Body Count, as any search of those two terms together will reveal.
[…] the David Kane paper we mentioned that supposedly refuted the Lancet study? The one that Malkin published and did that little dance […]
Null claims I made up the list I posted. I copied-pasted it from the list of sources that is linked to on their database:
http://www.iraqbodycount.org/sources.php
Take a look at it. I just copied-pasted a bunch of the obviously non “Western” sources listed near the beginning of that list.
Your list must be way out of date, and it says in your link that “Further sources will be added”.
As for your “Yeah, right.”, how do you know what I have and haven’t read?
I guess this because you obviously don’t know what sources they use, and say false things about it. Given this i find it hard to believe you’ve been reading the Arabic press and then meticulously sifting that big database cross-checking them, let alone to this large “degree” you say. I think you’re making all this stuff up just to trash Iraq Body Count because you wan’t people to disregard it and believe the Lancet study, iow, lying.
Null claims I made up the list I posted. I copied-pasted it from the list of sources that is linked to on their database:
http://www.iraqbodycount.org/sources.php
Take a look at it. I just copied-pasted a bunch of the obviously non “Western” sources listed near the beginning of that list.
Your list must be way out of date, and it says in your link that “Further sources will be added”.
Okay, I’ve taken a look at it, and it doesn’t seem as if this list is established as the sources for the Iraq Body Count, particularly since they explain that they go to sources in English (Al-Taakhi is in Arabic, for example) and that they use sources which are consistently available via the internet (which is not the case for Al-Taakhi either). I don’t really have to explain what it’s doing there, because it’s clear enough this list includes sources which, by their own stated standards, they wouldn’t use.
I guess this because you obviously don’t know what sources they use, and say false things about it.
But I do, and I provided the list. And even if your argument were accurate, what does that have to do with what I read in the Arabic language press?
Given this i find it hard to believe you’ve been reading the Arabic press and then meticulously sifting that big database cross-checking them, let alone to this large “degree” you say.
Strawman alert! In fact, all I said is that I had been reading the Arabic-language press, would find things that were not included in the Iraq Body Count, and would then file that away mentally as a personal data point against them, which led me to stop following the website very early in 2004. Perhaps they’ve improved their research since then, taking in non-English and non-internet resources, but even so that doesn’t make their numbers in the least bit reflective of overall reality. For one thing, it doesn’t account for the deaths that don’t get reported in the press, and it doesn’t overcome their initial deficiency in reporting numbers. At best, it simply establishes a lower limit for the deaths in Iraq. Using it to critique the Lancet study, therefore, is an exercise in stupidity.
I think you’re making all this stuff up just to trash Iraq Body Count because you wan’t people to disregard it and believe the Lancet study, iow, lying.
Well, you’re wrong, and you’re an idiot if you think so. In fact, that’s a bit superfluous–you’re just an idiot.
If I had wanted to try to convince people to believe the Lancet study, then I would have gone into a great deal of discussion about the robustness of their methodology and how it’s the same well-established one that has been used to estimate deaths from genocides in Rwanda, Sudan, Kampuchea, etc.
The fact is that I simply do not trust the Iraq Body Count as a resource, and was saying why.
“Okay, I’ve taken a look at it, and it doesn’t seem as if this list is established as the sources for the Iraq Body Count,”
Um, it’s on the iraqbodycount database page, under the heading “sources”. The list you posted is obviously not an exhaustive one of everything they use, and it even says so. It was probably put there in 2003 when they first wrote up the notes you drew it from.
“particularly since they explain that they go to sources in English (Al-Taakhi is in Arabic, for example) and that they use sources which are consistently available via the internet (which is not the case for Al-Taakhi either).
Who’s says sources in Arabic don’t put out stuff in English as well? And what about translation services? The fact that something is primarily in Arabic doesn’t mean that it can’t also be made available in English by various means.
I did a search for the term on the db and found this:
k2852
4 Apr 2006 -5 Apr 2006 –
Basra
professor Salah Abdul-Aziz
gunfire 1 1
AP 05 Apr
Al-Taakhi 06 Apr
That’s the only one i saw there with that term, but clearly it’s been used as a source. I also doubt the listing on their page lists every one of their sources that reports each incident. They seem to list two most of the time, but I’m sure that for many of these things there’s a lot more. For example, they list the notorious Haditha massacre as k2171, but they list just two sources: NYT and WP, but that has been talked about all over place, and by most of the sources they use a lot elsewhere. But they still list just two for it.
It’s pretty clear you don’t know what you’re talking about.
No shit. You do however have to bridge those two poles with an explanation when your source samples show one extreme and your extrapolations of those samples show the opposite extreme.
When you are estimating something as important as the number of people dead and your estimate suggests there’s 10 times more dead people than anyone can find, you’d wanna make that explanation a good one.
So far, zip. Nothin. That sound reasonable to you does it ?
Bullshit. All they do is collate deaths info from newspapers. WTF kind of odds did you think the odds were that after 4 years they hadn’t figured out the arabic for “dead” and “killed” and been using local papers ?
Based on what can I be certain of this ? Your assurance that you read something once as opposed the the fucking endlessly repeated statements to that effect by IBC ?
You remember IBC…. it was that source I didn’t refer to.
Neither. Iraqi hospitals aren’t under the control of the US military or the Iraqi government.
As the guy pointed out above, the primary source for these records and the incentive for them to be kept are the family of the deceased.
So while you’re asking yourself questions, ask whether you think that Iraq-wide (as this survey was) that there’s not only “opposing interests” among the family members of the dead who are claiming bodies (that the police find and are delivered to hospitals) to this degree that it could account for such an anomaly, but that it has also done so in completely the opposite trend that the survey itself suggests.
Your answer ?
Yes, that’s right, you don’t buy that bullshit either.
Except that we’re not talking about “official figures” from the government.
And that places like the LAT have surveyed regional hospital records vs those reported by the government and found no such massive manipulation.
Clearly you’re not going to bury anything until you collect the body.
The 9/10 figure I’ve been referring to is violent deaths. Are you still not bothering to go to the hospital for that same piece of paper that entitles you to compensation, death benefits and estate claims if that’s where your dead child is located ? Where you need to go to collect it ?
You know what looks good when you try to look smart, fail miserably and succeed in only making yourself look stupid, twice. Talking about how small other people’s brains are. Really.
Stat boy 1 : It looks like we killed a vast number of innocent people here guys…
Stat boy 2 : No, I reckon we just killed a big number of innocent people here, don’t get carried away.
SB1 : I’ve done some checking and seriously, vast…
SB2 : No, I reckon you’re still wrong, it is just a bit huge, not vast, nowhere near vast. Why not check again?
SB1 : Well, I would go back and check again but I’d be afraid of being killed by roving gangs of kidnappers, murderers, Islamist nutjobs, Militias, Trigger happy occupiers, or just pissed off relatives.
SB2 : Well, you’re in a pretty weak position then aren’t you. I’m going to have to declare your conclusion invalid.
*Meanwhile in Iraq*
Newly minted Iraqi orphan : Who’s going to look after me now?
Islamist militia : Come here little one.
Enraged Iraqi Ex-Dad : While you’re here, can I get some explosives and perhaps maps of US convoy routes?
IM : No problem, plenty for everyone.
*Meanwhile*
George Bush : Watch this drive!
No shit. You do however have to bridge those two poles with an explanation when your source samples show one extreme and your extrapolations of those samples show the opposite extreme.
No, one doesn’t. You see, this is because the papers were presented in the context of a medical journal, where it is assumed that the readers are capable of thinking.
When you are estimating something as important as the number of people dead and your estimate suggests there’s 10 times more dead people than anyone can find, you’d wanna make that explanation a good one.
So far, zip. Nothin. That sound reasonable to you does it ?
Yes, it does. One is not likely to miss deaths due to natural causes, particularly when there are doctors around to sign a death certificate. However, you are not often going to find death certificates for the victims of sustained bombing campaigns, because the deaths cover a much broader area and any doctors which have not already been killed or incapacitated are going to want to be as far away from where the bombs are falling as possible. Therefore, in any situation inside a war zone suffering from heavy bombing, the deaths by natural causes are going to be overrepresented compared to the deaths by bombings in those cases for which documentary evidence (i.e. the death certificate) exists, and the heavier the bombing, the greater the degree of divergence. Any competent statistician could have come up with that.
What I really love about the arguments here is the light it shines on wingnuts’ unwillingness to engage in any substantative way with the data on hand. Instead, it’s all insinuation, arguments from ignorance, and a panoply of embarrassing gaffes (like the assumption of a unimodal model for assessing death rates, which leads to the zombie threat in Iraq) that would make even the most minimally self-aware being hang its head in shame.
Who’s says sources in Arabic don’t put out stuff in English as well? And what about translation services? The fact that something is primarily in Arabic doesn’t mean that it can’t also be made available in English by various means.
“Who is says?” Apparently wingnuts don’t even put stuff out in English.
Note that I never said that Arabic language sources are always not simultaneously, to some degree, available in English in some respect as well. Do you need me for this conversation, or are you content to respond to things I’ve never said?
All I said is that I doubted it was a source, since when I was following the IBC, they refused to use non-English sources, and they didn’t use sources not on the internet. If your citation is accurate, then it was being used a single time as a source two full years after I stopped following it, so I accept on this that I was wrong. However, as to what they’re now doing, I don’t know and neither do you, so all this talk of translation services is so much chaff and frankly irrelevant.
What’s more irritating is that you consistently miss the point: using this source to critique the Lancet study is misapplied because a) they have undercounted Iraqi deaths, regardless of how carefully they’re doing it now and b) this undercounting was most in evidence when the war was at its height and c) therefore their numbers can, at best, only be a minimum level for Iraqi war dead. That is a minimum, not maximum.
A few years back there was a Dare to be Great school of financial management. The idea was you attended a seminar, then put up $5000, then watched your investment pile up as others came along and dutifully added to the pile. At the seminar you watched a guy at a blackboard with racing speech who scribbled rising numbers as he shreiked”… 20…40…80…100…140,000 Dollars, ladies and gentlemen!”. That was to motivate you and free up your $5000.
The leading figures in Dare to be Great were taken away to Federal Prison in due course , but their style is very definitely with us as right wing looney tunes scrape the barrel seeking some good news. ANY good news. What I notice, that seems to be their hallmark, is that they start with some nugget they have stumbled across, pronounce it valid by some connection to a name or a claim of science, then rapidly zoom off into space with wild conclusions. Yes indeed, “140, 000 Dollars, ladies and gentlemen!”
Time to call for the Feds, or the guys with butterfly nets, or maybe even Deltoid.
Do these murderous creeps think there were fewer than normal deaths in Iraq when the USA was claiming it was bombing large populated areas and hosing the streets with machine gun bullets? Do they suppose there were only a teensie bit more deaths than normal. Do they suppose no one is watching.
The moral implications of their conduct in attacking the Lancet study are atrocious.
You’re so cool! I don’t think I have read through a
single thing like that before. So great to find somebody with some genuine thoughts on this subject.
Seriously.. thanks for starting this up. This web site is
one thing that is required on the web, someone with a little originality!
streetdirectory – http://www.streetdirectory.com/businessfinder/company/8482/Singapore_Escort/ –