Lancet study finally mentioned

Last updated on June 10th, 2017 at 04:37 am

The Sydney Morning Herald’s Alan Ramsey gets it wrong, as usual:

On October 29 – just four days before polling in the US election – The Lancet published, on the internet, the result of a study carried out by a research team from Johns Hopkins University in Baltimore in association with researchers at Baghdad’s Al-Mustansiriya University. The study was based on a door-to-door survey in September of 8000 people in 33 randomly selected locations across Iraq and it concluded that a minimum 98,000 civilians had died in the violence, and its aftermath, since the US-led invasion 18 months earlier, in March 2003.

Here in John Howard’s Australia, as in official Washington, the study cut no ice whatever. Our newspapers have barely mentioned it.

He crazy. Meanwhile, reader Steve writes of Ramsey’s fellow plastic turkey conspiracist Phillip Adams:

Superb column by Adams in the Oz mag today. Not only does he build on the visionary content of Martin Luther King’s famous speech, but surpasses its oratorical elegance with his graceful and evocative writing.

King: “I have a dream …”
Adams: “Then, bugger it, the alarm clock went off.”

Ho, ho! Former Perth blogger turned big-media sellout (and WAFL umpire) Gareth Parker was in town last week for a conference, after which we went for a drive around Sydney. He was impressed enough by the Harbour Bridge and other standard sights, but these were mere preludes to the grandest landmark of them all. “And here, next to this hotel,“ I said, driving slowly along a Paddington street, “is the home of …�?

“No!�? Parker yelled. “Not the Dunny Lane Imperialist!”

Posted by Tim B. on 02/18/2005 at 11:40 PM
    1. You folks just had an earthquake about an hour and a half ago. 6.5 on the magnitude scale Earthquake list- USGS

      Posted by Mike H. on 2005 02 19 at 01:12 AM • permalink

 

    1. I’m sorry, it was just north of you.

      Posted by Mike H. on 2005 02 19 at 01:13 AM • permalink

 

    1. Any else here get the feeling that Ali bin Ramzi is way, waay behind the times. These dead camels he’s flogging are not just dead, they’re damned near skeletons.

      Posted by Boss Hog on 2005 02 19 at 01:53 AM • permalink

 

    1. The Lancet Study is going to be one of those Instant Truths of the Left, factual no matter how inaccurate.  Expecting it to go away is like expecting an Inquisitor to admit the Jews really weren’t poisoning the wells during the Black Death…

      Posted by richard mcenroe on 2005 02 19 at 02:05 AM • permalink

 

    1. Yes, I can see the Lancet getting the plastic turkey award along with Protocols of Zion.

      Posted by rog2 on 2005 02 19 at 02:14 AM • permalink

 

    1. Didn’t Phillip Adams have a similar WET DREAM back in 2004 or 2003, about “President Dean” (or was is Clark?) and “Prime Minister Crean” yakking on about all-mighty UN and the “unjustified” war on Saddam Hussein, etc.

      Posted by Richard_of_Oz on 2005 02 19 at 02:42 AM • permalink

 

    1. Even the most casual reading of the Lancet study shows that their figures are “a 95% chance of anywhere between 8,000 and 192,000”. This now becomes “a minimum 98,000 civilians”.

      Maybe he took the 9 from 95% and added it to the 8,000 minimum. One thing’s for sure: how come no editor checked up on this?

      Posted by Zoe Brain on 2005 02 19 at 03:02 AM • permalink

 

    1. aebrain,

      “a 95% chance of anywhere between 8,000 and 192,000�?.

      Isn’t that bureaucrat-ese for “no fucking idea”?

      Posted by Spiny Norman on 2005 02 19 at 03:24 AM • permalink

 

    1. Spiny: pretty much.

      I guess some are disappointed that the predictions of half a million dead civilians didn’t pan out so they need to seize at something that comes remotely close.

      Posted by Patrick Chester on 2005 02 19 at 03:28 AM • permalink

 

    1. I’d make a contribution to this conversation but I have to go. My neighbour is knocking on our common dunny lane access door, must’ve left the keys with his sick mum, better let him in.

      Posted by gubbaboy on 2005 02 19 at 06:40 AM • permalink

 

    1. So the Lancet “study” remains a viable rumor, huh?  It’s too bad that junk science lingers on like onions on your breath.

      Posted by The_Real_JeffS on 2005 02 19 at 06:50 AM • permalink

 

    1. The Lancet Study was indirectly not mentioned last night on Dennis Miller’s show, by Depak Chopra [sp].

      Chopra, frothing about “civilian deaths” in Iraq, menacingly asked Miller, “Do you know the figure?” Miller said simply, “Get me the figure.”

      “The figure” is now urban legand. Chopra did not know it or where it came from, existing as he does on a higher plane, that known as “frothing moronism”.

      The significant thing to me was the total failure of Chopra’s life. Follow that path? It made him rich, but nowhere. Why do I think immediately about Hollywood, the MSM, Saddam, Osama, Gore, Soros, Academia, Alterman, Jordan, Cole, Margo………?

      These kind of people are a disease asking for a cure, a cry for help, that of death.

      Posted by J. Peden on 2005 02 19 at 01:25 PM • permalink

 

    1. I bet twenty bucks Tim Lambert will make an appearance in this thread and scold us all for bagging out the precious report…

      Posted by Quentin George on 2005 02 19 at 06:19 PM • permalink

 

    1. I maybe tossing $20, but I think Andrea banned Tim Lambert, Quentin.  But go ahead and have a beer on me.  ;-D

      Posted by The_Real_JeffS on 2005 02 20 at 05:36 AM • permalink

 

    1. re. The Phat one:

      two comments;

      1) In your dreams!!!!!

      2) Whats with this desire for a good pussy whipping?

      Posted by LaVallette on 2005 02 20 at 06:17 AM • permalink

 

    1. Yeah Lancet is junk science alright.  Just like TV ratings and unemployment figures.

      That ‘dartboard’ critique has been dealt with:

      The confidence interval describes a range of values which are “consistent�? with the model1. But it doesn’t mean that all values within the confidence interval are equally likely, so you can just pick one

      The innumeracy of the commenters here is astonishing!

      Posted by dk.au on 2005 02 20 at 10:29 PM • permalink

 

    1. dk.au:

      The confidence interval is so broad – I mean, after all, that using the Lancet’s own figures there’s about* 1 chance in 20 that the total is either less than 8000 or more than 192000 – that it’s a dartboard.

      For example, would you buy a 100 page book if there was 1 chance in 40 of all but 8 of the pages being blank? So why buy a “study” whose 95% – not 99% or 99.99% – confidence interval is so broad? Read the original article, and you’ll see that using this methodology, there’s nearly a 1% chance that people were actually brought back from the dead due to the war.

      Given that all other studies show figures of between 5000 and 40000, and clustered around 13-18,000, I’d say the probability of the 100,000 figure, or anything remotely like it, is less than 1%. Of course this is a heuristic. To make it a “proper” methodology, I’d assume that all studies are random, equally likely and independent measurements, then come up with a figure and confidence margin. Would you care to make a bet what the result would be, based on these questionable assumptions? Because the Lancet study uses even more questionable ones. Like excluding the Fallujah figures, choosing random GPS co-ordinates and then fudging them, etc.

      OTOH your link and its many comments was to an excellent critique of the whole situation, and I’d recommend that anybody interested in the subject read it. I diagree with some of the conclusions: for example, that a Zero point less than 2.5% probable indicates that the study is sufficiently precise.

      I’m not questioning the study’s accuracy (though there’s questions over that). I’m saying the precision, or lack thereof, makes the whole thing almost completely useless. I’d personally be astounded if the “true” figure was less than 8,000, or more than 192,000 for example.

      The fact that applying the Lancet study’s methods to (pre-US-Attack) Fallujah leads to the conclusion that the most likely figure is that 1/3 of the population was already dead, and everyone else wounded – a figure greater than the Nagasaki A-bombing – would tend to call into question the methodology. Basically, it doesn’t pass Sanity checks, indicating that at least one of the assumptions was deeply flawed. If this isn’t “Junk Science”, I don’t know what is.
      * To quote your link :

      Stats teachers will tell you that it is a fallacy and wrong to interpret a confidence interval as meaning that “there is a 95% chance that the true value lies in this range”. However, I would say with 95% confidence that a randomly selected stats teacher would not be able to give you a single example of a case in which someone made a serious practical mistake as a result of this “fallacy”, so I say think about it this way.

      Posted by Zoe Brain on 2005 02 21 at 06:00 AM • permalink

 

    1. Oh yes: perhaps you would care to defend the Ramsay quote :

      it concluded that a minimum 98,000 civilians had died in the violence

      I think the figure of 8,000 is more accurate, no?

      Posted by Zoe Brain on 2005 02 21 at 06:02 AM • permalink

 

  1. No I certainly won’t defend Ramsay’s quote.  I don’t defend anyone that can’t read properly.  I certainly accept your claim that the confidence interval is wide.  Of course it is.  Good science, when human beings are involved, isn’t always about precision.  Statistics generally isn’t about precision.  It’s about the BEST estimate under a certain set of assumptions.

    I think it’s important that readers make up their own mind about whether this piece of research is worthy, keeping in mind it passed peer review in one of the most prestigious scientific journals in the world.

    People need to keep in mind what the study sought to do, namely, count excess deaths since conflict began and not attempt to classify them as combatants or civilians

    So don’t compare it to Iraq Body Count or direct empirical measures of deaths due to violence.  Keep in mind also the use of many hundreds of tons of Depleted Uranium which lingers.

    Posted by dk.au on 2005 02 22 at 01:29 AM • permalink