Thursday, May 31, 2012

On the Lancet's Claim of 1.4 Million Dead Iraqi Civilians

WE ARE LED TO BELIEVE that the Iraq Body Count has underestimated the number of Iraqi civilian deaths by 91.7%. WE ARE LED TO BELIEVE that the Associated Press has underestimated the number of Iraqi civilian deaths by 92.1%. WE ARE LED TO BELIEVE that the Iraq Family Health Survey has underestimated the number of Iraqi civilian deaths by 89.2%. AND WE ARE LED TO BELIEVE that the Wiki Leaks Iraq War Logs have underestimated the number of Iraqi civilian deaths by 92.2%. These assertions, folks, we are led to believe. Duh yuh think that there might be some politics involved here?

29 comments:

Dervish Sanders said...

Will: Duh yuh think that there might be some politics involved here?

Absolutely. By those who want to discredit the Lancet.

dmarks said...

If the same standards were applied to the Trayvon Martin case as Lancent applied to Iraq in there false 1100% over estimation of Iraq War deaths, Zimmerman would have been reported killed a dozen people. The one Zimmerman said he shot. Trayvon's dad's son. His mother's son. The guy heard with Zimmerman on 9-11. The four or five other victims made up using sampling just for the hell of it. 3 people who moved away from the gated community in the last 6 months that no-one can find a forwarding address for. Etc etc etc.

The Lancet "study" is full of bogus and deceptive tactics, such as generalizing based on deadly hot-spots such as Falujah and assuming the entire nation was like this.

The Iraq Body Count numbers are the most accurate: they count real deaths. There's no imagination at work. They don't automatically assume that refugees and people who quickly moved to other towns are war dead. They make sure one person is only counted once. They don't use statistical sampling and other flights of imagination to just make up stuff.

Dervish Sanders said...

Statistical sampling is NOT a "flight of imagination". This statement simply shows how ignorant dmarks is.

Lila Guterman, writing in a in a January 2005 Columbia Journalism Review article stated, "I called about ten biostatisticians and mortality experts. Not one of them took issue with the study's methods or it's conclusions. If anything, the scientists told me, the authors had been cautious in their estimates".

Also, while 1.4 million killed figure is an estimate, it is an estimate with a 95 percent confidence interval (Wikipedia: "In statistics, a confidence interval... is used to indicate the reliability of an estimate").

dmarks is one of those seeking to discredit the Lancet for political reasons.

dmarks said...

"Statistical sampling is NOT a "flight of imagination". This statement simply shows how ignorant dmarks is."

It is. It's all guesswork. And no matter how good the guesses are, it is never counting anything real.

Political fake lists discredit themselves. I stand with the accurate numbers from Iraq Body Count, which include no made-up persons.

Dervish Sanders said...

dmarks: It is. It's all guesswork...

The Lancet "guesses" have a 95 percent confidence interval. According to Wikipedia, "in statistics, a confidence interval... is used to indicate the reliability of an estimate".

A 95 percent reliable estimate is NOT a "flight of imagination". dmarks bogus claims of "imagination" and "guesswork" drive home just how ignorant he is.

dmarks said...

The Lancet wild guesses (no quotes needed) are off by a factor of 10. More than 91% of their claims of deaths are fabricated; fictional. Such a whopping huge error rate is more than a "flight of imagination". It's a fevre dream.

My pointing out the fact that the fake study is almost all imagination and guesswork stands strong. Being "ignorant" of events, deaths that never occurred is a strength.

What a weird world you must live in, WD. To be haunted so by over a million deaths which never occurred.

dmarks said...

And check here. The facts. Counts made by counting.... imagine that. Not by people sitting in offices with no idea what happened in reality.

Dervish Sanders said...

dmarks: Not by people sitting in offices with no idea what happened in reality.

This proves you have no idea what you're talking about. The Lancet sent people out into the field to conduct their surveys. They did not "guess", they did a sampling and extrapolated from that. It's the way all polling is done.

We already know how ignorant you are on this subject dmarks. There is no need for you to go on and on to further prove it.

dmarks is the one who has no idea what happened in reality.

dmarks said...

"They did not guess, they did a sampling and extrapolated. from that"

Sorry to break it to you pal. That's a type of guessing. And real bad guessing, too: as it created a "body" of deaths the vast majority of which were imaginary.

I can't wait until these imagineers conduct a US census, and we find ourselves living in a nation of 4 billion people.

"We already know how ignorant you are on this subject dmarks."

I know a lot more than you on it, as has been proven.

Dervish Sanders said...

dmarks muzzily said: And real bad guessing, too...

The Lancet "guesses" have a 95 percent confidence interval. According to Wikipedia, "in statistics, a confidence interval... is used to indicate the reliability of an estimate".

You have to imagine real hard to believe a 95 percent confidence interval is bad.

dmarks said...
This comment has been removed by the author.
dmarks said...

No imagining is necessary when the vast majority of the deaths "found" never occurred at all.

As for the laughed-at "study" from 2006, it is a textbook example of bad polling and violation of media ethics. In fact, it was honored with a 2010 “Top Ten Dubious Polling” by the watchdog Stinky Journalism, due to reasons which included "non-disclosure of essential information on the survey’s methods".

The smoke-and-mirrors of Lancet's sloppy and unethical "methodology" are well documented, which is why no one except those on the fringe take it seriously.

And then there is the fact that no other study (including the competing ones which unlike Lancet have valid and open methodology) even comes close to the Lancet novel.

This matters if you are concerned with science and facts, where such conclusions are repeatable and supported, as opposed to only appearing in a puff of pixie dust from one dubious organization which treats documentation like alchemist guild's secret.

But if you have an uncritical mind and aren't concerned with science, of course you will accept Lancet as gospel.

Stephen Ssoldz at the hard left "Z Communication" readily rips the flawed ideology of the bogus studies also.OF course, most on the left AND right reject the fake flawed studies.

Dervish Sanders said...

The Lancet "guesses" have a 95 percent confidence interval. According to Wikipedia, "in statistics, a confidence interval... is used to indicate the reliability of an estimate".

You have to imagine real hard to believe a 95 percent confidence interval is bad.

Will "take no prisoners" Hart said...

Confidence intervals are notoriously unreliable, wd. What you really need is a credible interval to adequately address the sampling reliability.

dmarks said...

There's no quotations needed around guesses. 90% of the death claims from them are nothing but guesses when you get right down to it (the difference between the actual death count, and their tenfold numbers).

Will "take no prisoners" Hart said...

It's clear that the Lancet people really needed to consult some professional polling people. Every other reputable source says between 100 and 150,000 and they come up with this 1.4 million figure. Very strange. I think that you were right, dmarks, when you said that they probably included a lot of displaced people in their documentation.

Dervish Sanders said...

Even if dmarks could be convinced that the the Lancet numbers were accurate I doubt he'd care. The only reason he argues against the truth is because it looks bad for the war criminal president he defends.

dmarks said...

Why not just call George W. Bush a "poopy pants"? You are using the term "War Criminal" entirely without reason, and from you it only says how much you hate Bush, really... and says nothing about Bush. Only you.

So considering this, any lame insult will do.

dmarks said...

Will said; "Every other reputable source says between 100 and 150,000 and they come up with this 1.4 million figure"

Which brings up a very important point. I've looked at the lists of different death toll claims. The Lancet numbers, with their strength of having 90% imaginary deaths, stand way above the rest of them.

Will, is it quite obvious to you that WD only chose to stick with the Lancet numbers because they are the highest... and not because of any consideration to validity or methodology?

I stick with "Iraq Body Count" because they are the most accurate: they only count actual deaths, and even though they think there are more, they don't conjure imaginaries as Lancet does. There are lower estimates coming from the military and others. But these aren't as well documented, so I reject those.

dmarks said...

And Will, if Lancet were put in charge of the US Census the next time around, you'd find yourself living in a state with a 600 million population.

Dervish Sanders said...
This comment has been removed by the author.
Dervish Sanders said...

dmarks: Why not just call George W. Bush a "poopy pants"? You are using the term "War Criminal" entirely without reason...

Why would I do that? I don't know, or care if bush pooped his pants. But there is a mountain of evidence to suggest that bush is a war criminal. But of course you'd term looking at the evidence to be "entirely without reason".

bush approved torture in violation of the Geneva Conventions of his own admission (he admits to approving torture in his book).

Secretary-General of the United Nations Kofi Annan, speaking on the Iraq invasion, said, "I have indicated it was not in conformity with the UN Charter. From our point of view, from the charter point of view, it was illegal".

Conclusion: It is completely illogical to say calling bush a war criminal is "entirely without reason". I just gave you two reasons (and there are many more). Personally, I think dmarks is a poopy-head.

dmarks: ...is it quite obvious to you that WD only chose to stick with the Lancet numbers because they are the highest... and not because of any consideration to validity or methodology?

It's obvious to me that you're rejecting them for that reason. Iraq Body Count admits their numbers are wrong! Your own comment says so.

I'm sticking with the Lancet because they are attempting to account for the bodies that go missing... instead of pretending those people didn't die.

Jerry Critter said...

The Iran Body Count is low if we are to believe dmarks. They missed dead people. The Lancet number is higher, which is in the correct direction from the Iraq Body Count number. Lancet says there number is correct with a 95% confidence level. To attack the Lancet number you must attack their methodology and show that their confidence level is too high.

To just say the Lancet number is too high because it does not agree with anyone elses is a bogus argument.

Will "take no prisoners" Hart said...

Confidence intervals are often quite meaningless when doing mass multiple surveying like this. If the Lancet had said that they had a 95% credible interval, that would have been significantly more convincing.

Jerry Critter said...

"Confidence intervals are often quite meaningless when doing mass multiple surveying like this."

Then show what they did wrong. You just saying so does not make it true.

Will "take no prisoners" Hart said...

Jerry, the fact that they used a confidence interval and NOT a credible interval is what they did wrong. Confidence intervals are totally inadequate when you're doing multiple samplings.

Jerry Critter said...

You have moved beyond my level of expertise and interest. I will take your word for it. I am not familiar with credibility limits, but I'd they should have used them then they screwed up.

dmarks said...
This comment has been removed by the author.
dmarks said...

erry said: "...The Iran Body Count is low if we are to believe dmarks"

Careful. That's a war that hasn't happened yet.

"To attack the Lancet number you must attack their methodology and show that their confidence level is too high."

There are plenty of other reasons to "Attack" it. Such as the fact that they have chosen to hide the "evidence" which caused them to come to their conclusions. This link reveals the smoke and mirrors and deception and fraudulent methodology Lancet attempts to hide itself behind. "Iraq Body Count" is completely open about everything. In contrast, the Lancet study was so bogus that the principle investigator, Gilbert Burnham, was suspended due to the problems with the Lancet study. It's basically a fraudulent work, and its author was caught and busted over it.

No wonder he had something to hude: "Burnham had revealed that the survey used a sampling methodology which differed from the published account1. When researchers requested details, all were refused"

Sorry, that sort of secrecy plays in the Vatican, but not in science.

Forget discussion of confidence intervals.... when there's a con artist involved with it. A con artist who got fired due to the problems with the Lancet study, which he authored.