Transterrestrial Musings  


Amazon Honor System Click Here to Pay

Space
Alan Boyle (MSNBC)
Space Politics (Jeff Foust)
Space Transport News (Clark Lindsey)
NASA Watch
NASA Space Flight
Hobby Space
A Voyage To Arcturus (Jay Manifold)
Dispatches From The Final Frontier (Michael Belfiore)
Personal Spaceflight (Jeff Foust)
Mars Blog
The Flame Trench (Florida Today)
Space Cynic
Rocket Forge (Michael Mealing)
COTS Watch (Michael Mealing)
Curmudgeon's Corner (Mark Whittington)
Selenian Boondocks
Tales of the Heliosphere
Out Of The Cradle
Space For Commerce (Brian Dunbar)
True Anomaly
Kevin Parkin
The Speculist (Phil Bowermaster)
Spacecraft (Chris Hall)
Space Pragmatism (Dan Schrimpsher)
Eternal Golden Braid (Fred Kiesche)
Carried Away (Dan Schmelzer)
Laughing Wolf (C. Blake Powers)
Chair Force Engineer (Air Force Procurement)
Spacearium
Saturn Follies
JesusPhreaks (Scott Bell)
Journoblogs
The Ombudsgod
Cut On The Bias (Susanna Cornett)
Joanne Jacobs


Site designed by


Powered by
Movable Type
Biting Commentary about Infinity, and Beyond!

« Realignment of Presidential Punters | Main | The Current State Of Anti-Aging »

Scientific Fraud

At The Lancet. This isn't really new news--anyone with half a brain who looked at the study carefully at the time (i.e., not all-too-credulous journalists) could see that it was a nonsensical statistical mess. But the case against it is looking even stronger now.

Of course, it fulfilled its political purpose--to damage the Republicans and the Bush administration in the 2006 elections. And when it comes to righteous moral crusades like that, accuracy and scientific integrity be damned.

Posted by Rand Simberg at January 04, 2008 06:09 AM
TrackBack URL for this entry:
http://www.transterrestrial.com/mt-diagnostics.cgi/8806

Listed below are links to weblogs that reference this post from Transterrestrial Musings.
Comments

I recall my thoughts when those inflated numbers were released.

Figures don't lie, but liars can figure.

And isn't it odd, that the same folks who support these kinds of inflated numbers, are also onboard with global warmming and zero sum economics?

Posted by Steve at January 4, 2008 06:53 AM

Unfortunately, the Lancet Iraq study is pretty much par for the course with regard to the methodology in even the most prestigious medical journals. I don't doubt there was some ideological bias among the authors (there always is), but the real problem is that most medical researchers don't use protocols that are known to filter out such bias. This is one reason progress in treating common medical conditions is so slow - the quality of the research is poor, and practicing doctors know it, so they treat based on common sense and their own experience instead of doing what "studies suggest".

Most (not all) of today's medical research goes like this: Researcher conjectures a relationship between two clinical observations. The relationship may be obvious or it may be far-fetched, but there is no attempt to develop a theory or mechanism for why the two observations might be related. Researcher collects data or mines data from other studies, then analyzes the data in a superficially rigorous way, with trends and p-values, that is never reviewed by a statistician. Then, if the analysis shows a "statistically significant" relationship between the two observations, the results are published. If no relationship can be shown after repeated attempts at analysis, the results are not published.

In the case of the Lancet study, the two observations to be related are the fact of the Iraq invasion and the number of excess deaths. There are so many ways for bias to creep into such a study that it is pointless to list all of them. The most telling deficiency is how grossly different the Lancet results were compared to all other estimates, yet there was no attempt to explain this, except in very generic terms (2nd full paragraph, col 2, page 6).

As doctors, the authors are taking the position that their job is to stop people from dying. On one level, it's hard to argue with that, but on another, political level, that is an essentially pacifist point of view when it comes to war. It's easy to say that if you give antibiotics to someone with pneumonia, and he recovers, you have done a good thing. But it's much more complicated to say that civilian deaths in a war are not worth the possible positive outcomes of the war.

Disclosure of bias: I have always been opposed to the Iraq invasion and I think it was a huge strategic mistake. But you don't mess with science.

Posted by Artemus at January 4, 2008 07:05 AM

For me the two factors that lead me to disregard the study was first, the breezy discussion of how they supposedly prevented fraud in the study. For example, there was no discussion of how to deal with forged death certificates or outside parties suborning the sampling process. Second, when I compared their figures to the Iraq Body Count (IBC) figures, car bombs and shootings both were magnified in the Lancet study by about the same factor. My take is one would expect car bombings to be better reported and hence more inline with the IBC numbers.

Posted by Karl Hallowell at January 4, 2008 07:18 AM

One insidious aspect of the Lancet report is the implicit portrayal of Iraq as some kind of primitive society, where a half-million or so people can get killed and buried without any record of it showing up in death certificates, police journals, hospital reports, newspapers, cemeteries records, or other regularly maintained documents.

It seems to me the entire premise of the study was that Iraq represents the Middle East equivalent of some recently discovered jungle tribe in the Phillipines that has somehow avoided contact with modern civilization for hundreds of years and doesn't understand what a pencil is for.

Posted by BD at January 4, 2008 09:19 AM

The Lancet study was a fraud. It does not change the fact that thousands of civilians, maybe tens of thousands of civilians, were collateral damage during the invasion and in the anarchy which has ensued since the invasion. You still have to weigh that against what may have happened to those civilians if we had not invaded.

Many would say that Saddam would have killed even more. I don't believe it. What Saddam may have done is a conjecture not backed up by any study.

Posted by Jardinero1 at January 4, 2008 09:57 AM

Saddam would have continue to do what he was already doing--terrorizing the populace, throwing people in prison and torturing them, randomly raping young women who took his sons' fancy and then killing or imprisoning them, starving children that weren't politically favored, destroying the marshlands and the way of life of the marsh Arabs, etc.

If you think that he would have somehow magically stopped doing all these things if we hadn't removed him from power, the burden of that extraordinary claim is on you. The continuation of those things would have been "collateral damage" of leaving him in power.

Do you really believe that there is a significant number of Iraqis who want to return to the days of Saddam? That's what is implied by your comment. If so, the answer is "no."

Posted by Rand Simberg at January 4, 2008 10:05 AM

I agreed the Lancet study was bunk. Where is the data that supports your statements.

I don't deny that what you state may have happened. But you can't say with any certainty that the Iraqis are better off today than they were under Saddam unless you have some real data about how many died, on an annualized basis, as a result of Saddam's rule. Personally, I think the invasion and the occupation have resulted in more Iraqi civilian deaths than would have occurred under Saddam. It has definitely caused 3000 more American deaths than would have occurred if we had not invaded; and created several tens of thousands more permanently maimed Americans than if we had not invaded.

Posted by Jardinero1 at January 4, 2008 10:36 AM

I agreed the Lancet study was bunk. Where is the data that supports your statements.

Actually, what made the Lancet study of interest at all was that its numbers of US killed dwarfed the numbers killed by Saddam. But here you go. It seems that instead of the US killing 2 or 3 times the civilians as Saddam, it looks more like 4 times the other way.

Now, for a moral retard, they might respond, "still the US is killing civilians". Ok, that is the case and a sad one for sure. However, it the US decided to not intervene then many more civilians would have died. Moreover, what kind of morals does it take for someone to elevate the number of deaths in order to justify an inaction to such brutality?

Posted by Leland at January 4, 2008 11:09 AM

Comparing life in Iraq today to life in Iraq under Saddam Hussein only makes sense if we assume Iraq has reached some kind of endpoint in its transition away from oppressive dictatorial rule. Obviously, they haven't reached that endpoint. To use the current situation in Iraq as the basis for comparison assumes that things will never get better for the Iraqi people than they are right now.

Also, if it's true that Saddam Hussein was the only thing keeping a lid on the underlying tribal conflicts that define Iraqi culture (which is quite a generous characterization of Saddam's role in Iraqi society), then much of the violence we have witnessed since the coalition invasion was probably inevitable. Saddam wasn't going to live forever. Unless you think Iraq's pre-war destiny was to exist under a perpetual police state, there inevitably would have come a time when the various religious and ethnic groups would have slipped the Baathist leash and started battling one another over longstanding grievances. George Bush gets blamed for a lot of things, but he didn't create the underlying sectarian divisions in Iraq.

Posted by BD at January 4, 2008 11:30 AM

George Bush gets blamed for a lot of things, but he didn't create the underlying sectarian divisions in Iraq.

Nor did he create Al Qaeda, in Iraq or elsewhere, which is responsible for most of the violence, either directly, or indirectly by fomenting those divisions.

Posted by Rand Simberg at January 4, 2008 11:38 AM

But you can't say with any certainty that the Iraqis are better off today than they were under Saddam...

Jardinero1,
you're right Rand can't make that call. But that's precisely what the Iraqi people say. I have this on good authority from my son, who was there, and his fellow Marines, who have been back there. If they think things are better, all the discussion here, or anywhere else is strictly an exercise of opinions and political points of view.

Posted by Steve at January 4, 2008 12:07 PM

Leland, Saddam killed more people, but he had a LOT more time to do so, making the rate lower than the wartime casualty rate.

Frankly, though, I don't give a s*** exactly what the difference in casualty rates is, for the same reasons as BD.

Posted by Math_Mage at January 4, 2008 01:20 PM

Math Mage,

I understand and agree with what your saying, but if you look at a set time frame equal to what the Lancet study was looking at, I'm not sure the rate is less. Alas, it is a stupid debate point.

Posted by Leland at January 4, 2008 06:08 PM

The League of Nations created the underlying sectarian devisions in Iraq when they gave the British a mandate in Mesopotamia at the end of WW1.
It was supposed to be divide and conquer folks, which is why the unstable mix of Shia and Sunni, Kurds and Assyrian Christians.

Posted by RKV at January 4, 2008 06:51 PM

I'm not sure the 'rate' is the proper metric to be studying.

Assume Hussein stayed in power indefinitely (ie Castro) and his 'death rate' stayed constant. Further assume that the UN sanctions were lifted in absence of any information from the regime that they had actually met the requirements of the UN. Assume, further, that either Uday or Qusay Hussein would assume power at Saddam's death/incapacitation and the death rate in Iraq remained constant for their regime - but magically changed to a normal rate at their death of natural causes at ~ 75 years of age. (~ 35 years or so from now.)

How does that differ from the fix that was (or is being) applied by US blood and treasure?

Posted by JAFAC at January 4, 2008 08:23 PM


Post a comment
Name:


Email Address:


URL:


Comments: