Thursday, October 12, 2006

"Truth" - Brought to You by the MSM

Another political "study" on the death toll in Iraq due to the war. Amazingly, the numbers are fantastically larger than those in the last "study" that the Lancet published. I'm not a statistician, but the methodology of this study strikes me as frighteningly opportunistic.
BAGHDAD, Oct. 10 - A team of American and Iraqi public health researchers has estimated that 600,000 civilians have died in violence across Iraq since the 2003 American invasion, the highest estimate ever for the toll of the war here.

The figure breaks down to about 15,000 violent deaths a month, a number that is quadruple the one for July given by Iraqi government hospitals and the morgue in Baghdad and published last month in a United Nations report in Iraq. That month was the highest for Iraqi civilian deaths since the American invasion.

But it is an estimate and not a precise count, and researchers acknowledged a margin of error that ranged from 426,369 to 793,663 deaths.

It is the second study by researchers from the Johns Hopkins Bloomberg School of Public Health. It uses samples of casualties from Iraqi households to extrapolate an overall figure of 601,027 Iraqis dead from violence between March 2003 and July 2006.

And
The findings of the previous study, published in The Lancet, a British medical journal, in 2004, had been criticized as high, in part because of its relatively narrow sampling of about 1,000 families, and because it carried a large margin of error.

The new study is more representative, its researchers said, and the sampling is broader: it surveyed 1,849 Iraqi families in 47 different neighborhoods across Iraq. The selection of geographical areas in 18 regions across Iraq was based on population size, not on the level of violence, they said.

Not sure why the byline is from Baghdad. I'd guess that this report wasn't issued from there, but if it makes the report look more relavant to say it's from Baghdad, then Hey let's do it.

The error levels are frankly astounding. Research I did in college that had error rates that high were put into that class of data called garbage. The the reaction to the sampling criticism is just funny. Oh, 1000 isn't enough? Well, we'll sample 1849. That will make the error margin so much better.

Then there is this:
Statistics experts in the United States who were able to review the study said the methods used by the interviewers looked legitimate.

Robert Blendon, director of the Harvard Program on Public Opinion and Health and Social Policy, said interviewing urban dwellers chosen at random was "the best of what you can expect in a war zone."

But he said the number of deaths in the families interviewed - 547 in the post-invasion period versus 82 in a similar period before the invasion — was too few to extrapolate up to more than 600,000 deaths across the country.

Note that they don't name any of those statisticians who state that the methods "look" legitimate. If you don't have an actual name, you got nothing. This is worse than those anonymous sources, since this is an attempt at validation without explanation.

I also like the sampling coming only from urban dwellers. They then take those stats and apply them across all of Iraq. Funny, I would expect that would integrate a much higher error rate since your not actually applying the appropriate stats to the various populations. Of course, if that's all you can do in a war zone, then that must be something to justify poor methodology.

I had to look to the conservative MSM to find any real criticism.
Yesterday the study got generous coverage in The Washington Post, the Wall Street Journal and elsewhere. Why that happened is unclear, because the scientists -- the lead author this time is Dr. Gilbert Burnham, last time Dr. Les Roberts -- burned the media in 2004 by irresponsibly hyping a supposed death total of 100,000. The signs of politicization were clear enough: One author admitted to politically motivated timing. Lancet editor Richard Horton called the war "grievously in error."

And sure enough, as Slate's Fred Kaplan showed, the study actually proved no such figure. Sampling Iraqis around the country by interview, the authors' survey really determined that the death toll in 2004 was somewhere between 8,000 and 194,000. As Mr. Kaplan put it: "This isn't an estimate. It's a dart board." Every other credible group or analyst put the number in the range of 15,000-30,000. The 100,000 estimate was nevertheless declared to be "conservative" by these political scientists.
Burnham denies any political motivations, but that is rather odiferous.
Dr. Gilbert Burnham, lead author of the study, defended his findings and research practices, which he called standard among international health specialists.

The study was initially planned to be completed during the summer, he said, denying any political motive behind its release Wednesday.

Not that I doubt his veracity, oh, wait, Yes I do.

Hell, even Iraq Body Count doesn't report anything this outrageous. This report makes them look completely reasonable.

NewsBusters has an article on the findings here. They also have an article looking at the reporting by the MSM.
Despite how the estimate of 665,000 Iraqi deaths caused by violence since the war began -- a number forwarded in a new report from the Johns Hopkins Bloomberg School of Public Health -- represents quadruple the highest monthly rate as tracked by the UN and is 13 times larger than the total compiled by the Iraq Body Count group, CBS Evening News anchor Katie Couric set up a Wednesday story on the guesstimate by declaring as fact: --Now we're learning that the war has been a lot more deadly than we knew.-- David Martin proceeded to treat the number as perfectly reasonable as he put the blame on the U.S.: "A new and stunning measure of the havoc the American invasion unleashed in Iraq. A study published in the British journal Lancet estimates 655,000 Iraqis -- 2.5 percent of the entire population -- have died as a consequence of the war. To understand how large, consider this: The same percentage of the much larger American population would be 7.5 million dead."
Now tell me, is this responsible journalism? Does this method of reporting create truths that don't exist in reality? Does the end result do damage to the intentions of the US in Iraq and the Middle-East?

The fourth estate is again taking on the helm of the "Ministry of Truth."


1 comment:

Granted said...

I did a quick calculation. This if 500+ deaths a day, every day for the last 3.5 years. Am I alone in thinking that seems WAY high? I mean if 500 people were killed today, if only 400 were killed tomorrow, then 600 have to get killed the next day. It's an insane number. These guys clearly have zero credibility, but the MSM is just going to trot the stuff out for the next few days until the study gets well & truly busted, but then it'll be too late. What a crock.