A hidden epidemic of car bombs?

[Puppycow]

Split from this thread for scientific input

Reference

Quote:

More specifically, Lancet data suggests large numbers of deadly car bombings occurring on a daily basis, of which only a small fraction are ever reported (and whose victims, including injured, fail to be recorded by hospitals).

Lancet estimates 150 people to have died from car bombs alone, on average, every day during June 2005-June 2006. IBC's database of deadly car bomb incidents shows they kill 7-8 people on average. Lancet's estimate corresponds to about 20 car bombs per day, all but one or two of which fail to be reported by the media. Yet car bombs fall well within the earlier-mentioned category of incidents which average 6 unique reports on them.

'Baghdad-weighting' of media reports, even if applicable to car bombs, is unlikely to account for this level of under-reporting, as half of the car bombs IBC has recorded have been outside Baghdad. The Pentagon, which has every reason to highlight the lethality of car bombs to Iraqis, records, on average, two to three car-bombings per day throughout Iraq, including those hitting only its own forces or causing no casualties, for the period in question.

Is it possible that about 20 deadly car bombings occur every day in Iraq for 2 years, but that 90% go unreported despite causing casualties, while the few that do get reported generate an average of 6 independent reports?

[Puppycow]

The silence is deafening. No one has an opinion on this?

I submit that 99-100% of car bombs that cause mass casualties will be reported by somebody. The car bombs that do get reported average 6 unique reports.

[Gravy]

Not that I can provide it, but what's the scientific input you're looking for?

[PixyMisa]

If they claimed that only 90% of car bombs in Iraq are reported, I could believe that.

But claiming that 90% of car bombs go unreported? Maybe if they're homeopathic car bombs...

[Puppycow]

Originally Posted by PixyMisa

If they claimed that only 90% of car bombs in Iraq are reported, I could believe that.

But claiming that 90% of car bombs go unreported? Maybe if they're homeopathic car bombs...

And if that were true, would it be reasonable to assume that the ones that don't get reported would be as deadly as those that do?

I have a hard time believing that any car bombs that cause mass caualties would go unreported. Maybe duds or ones that only kill the suicider, but not ones that kill lots of people.

And if it follows that this claim is not credible, then isn't the whole study questionable?

Here is a new study that claims:

Quote:

According to the ORB poll, a survey of 1,461 adults suggested that the total number slain during more than four years of war was more than 1.2 million.

ORB said it drew its conclusion from responses to the question about those living under one roof: "How many members of your household, if any, have died as a result of the conflict in Iraq since 2003?"

Based on Iraq's estimated number of households -- 4,050,597 -- it said the 1.2 million figure was reasonable.

There was no way to verify the number, because the government does not provide a full count of civilian deaths. Neither does the U.S. military.

Both, however, say that independent organizations greatly exaggerate estimates of civilian casualties.

ORB said its poll had a margin of error of 2.4%. According to its findings, nearly one in two households in Baghdad had lost at least one member to war- related violence, and 22% of households nationwide had suffered at least one death. It said 48% of the victims were shot to death and 20% died as a result of car bombs, with other explosions and military bombardments blamed for most of the other fatalities.

The survey was conducted last month.

It was the highest estimate given so far of civilian deaths in Iraq. Last year, a study in the medical journal Lancet put the number at 654,965, which Iraq's government has dismissed as "ridiculous."

So 20% of 1.2 million is 240,000
Over 4.5 years that averages about 146 deaths by car bomb per day.
According to Iraq Body Count each vehicle bomb that does get reported generates an average of 6 independent reports, but numbers reported are nowhere near 146/day:

(Deaths/day by vehicle bomb)

[mhaze]

Lancet includes in their numbers what might be called "indirect" casaulties of the war, for example, heart and stroke attack victims that could not get to a hospital in wartime conditions, deaths due to bad water supply, etc.

Actual "war like events" like car bombs are intended to terrorize, not inflict mass causalties. Above a certain number, there would be no reason to increase the number of these style attacks further.

So yeah, the assertion makes NO sense.

[PixyMisa]

This series of studies is widely considered to be complete nonsense, because of the complete disconnect between the numbers produced and the number of actual bodies.

Another significant point is that you expect a ratio of dead to wounded of between 2:1 and 5:1 in a war, depending on the type of combat and the availability of medical care. So this study would indicate between 2.4 and 6 million wounded Iraqis... Which simply isn't true.

[Puppycow]

Originally Posted by mhaze

Lancet includes in their numbers what might be called "indirect" casaulties of the war, for example, heart and stroke attack victims that could not get to a hospital in wartime conditions, deaths due to bad water supply, etc.

Actual "war like events" like car bombs are intended to terrorize, not inflict mass causalties. Above a certain number, there would be no reason to increase the number of these style attacks further.

So yeah, the assertion makes NO sense.

Not to mention the sheer number of vehicles, bombs and suicide bombers it would take.

[Puppycow]

Originally Posted by PixyMisa

This series of studies is widely considered to be complete nonsense, because of the complete disconnect between the numbers produced and the number of actual bodies.

Another significant point is that you expect a ratio of dead to wounded of between 2:1 and 5:1 in a war, depending on the type of combat and the availability of medical care. So this study would indicate between 2.4 and 6 million wounded Iraqis... Which simply isn't true.

Thank you.
However, it seems that a lot of people are believing it, including some on this board.

[casebro]

And if true, there would be no cars left on the road.

I just can't see that taking a poll is the right way to find facts. Opinions maybe, not facts.

Then there is no corroborating data. Like, how many fewer cars on the road, average size of shoes being purchased is smaller due to adult male deaths. Lower falafel sales. Higher exports of scrap iron for recycling. Lower coffee sales. Higher dildo sales as women replace their missing men. Face it, a million dead has to effect the economy in some directly measurable way.

[RecoveringYuppy]

Originally Posted by casebro

Face it, a million dead has to effect the economy in some directly measurable way.

Well, two million people have left the country which would presumably have twice the economic impact of a million dead. I don't think you could rule out a million dead using that method.

[mhaze]

Originally Posted by RecoveringYuppy

Well, two million people have left the country which would presumably have twice the economic impact of a million dead. I don't think you could rule out a million dead using that method.

It's not my job to rule out all the possibilities.

It's the job of the Lancet to adequately substantiate their claim.

Or to put it more directly - if YOU want to assert something that seems a bit illogical (million dead) - I have no problem at all with it, but the burden of proof would be for you, right?

[PixyMisa]

Just to correct myself above (too late to edit) - when I said "the ratio of dead to wounded", I meant "the ratio of wounded to dead". Otherwise it doesn't make much sense...

[RecoveringYuppy]

@mhaze,

I was addressing casebro's idea that a million dead could be ruled out by looking for economic effects.

[fuelair]

Originally Posted by casebro

And if true, there would be no cars left on the road.

I just can't see that taking a poll is the right way to find facts. Opinions maybe, not facts.

Then there is no corroborating data. Like, how many fewer cars on the road, average size of shoes being purchased is smaller due to adult male deaths. Lower falafel sales. Higher exports of scrap iron for recycling. Lower coffee sales. Higher dildo sales as women replace their missing men. Face it, a million dead has to effect the economy in some directly measurable way.

I suspect they have to handcraft those dildos - given AlQuaedi in Iraq has already taken the ones they had so they can dream of Osama giving them his "blessing".

[FireGarden]

Do I have to repeat everything from the other thread?
The new poll is not of the same standard as either Lancet report.

Lancet-2 reports about 77 deaths per day from car bombs.

The British Ministry of defence's Chief Scientific Adviser, Sir Roy Anderson, on 13 October, states:

http://news.bbc.co.uk/1/hi/uk_politics/6495753.stm

Quote:

"The study design is robust and employs methods that are regarded as close to "best practice" in this area, given the difficulties of data collection and verification in the present circumstances in Iraq."

The methodology used is widely credited.
Les Roberts, one of the co-authors, has conducted many surveys. He says deaths are often under-counted in war-zones, and by a large degree.

He also gives a method whereby his data can be easily refuted.

The "official" crude death rate:
https://www.cia.gov/library/publicat...elds/2066.html

Iraq -- 5.26 deaths/1,000 population (per year)

Pop Iraq -- 26,783,383
http://en.wikipedia.org/wiki/Demographics_of_Iraq

5.26*26783 = about 141000 per year
Even the IBC total for the whole war only comes up to about half of that.

Lancet-2 recorded 547 post-war deaths, about 300 of them violent. (Amongst 12,801 people in the survey)

That is very, very different to the ratio you'd expect if IBC is more accurate.

And, while the counting of death certificates is something only a government ministry can do, faking the above ratio of violent deaths would be much, much harder.

Unfortunately, I can't find that data on the internet.

[FireGarden]

As for reporting car bombs...
How often do journalists go out? Do they rely on the phone?

This is the first reference I could find:
http://news.independent.co.uk/world/...icle548945.ece

Patrick Cockburn, May 2006.

Quote:

The sectarian warfare in Baghdad is sparsely reported but the provinces around the capital are now so dangerous for reporters that they seldom, if ever, go there, except as embeds with US troops.

Lots of opportunity to see every car bomb and get details of the casualties?

The journalists stay indoors most of the time and report what filters to them.

[FireGarden]

Here's an article from 2005, and remember that getting back to that level of violence is a sign of success.

http://www.news.com.au/adelaidenow/s...5-1702,00.html

Quote:

THE number of dead Iraqi civilians counted at the Baghdad morgue hit 1100 in July, the highest toll in recent history, a British newspaper reported today, blaming the daily violence.

[...] The death toll was up from about 800 in July last year and 700 during the same month in 2003, according to the left-wing daily.

By comparison, equivalent figures for July 1997, 1998 and 1999 - during the leadership of former Iraqi president Saddam Hussein - were all below 200, The Independent said.

From 200 to 1100.
The Lancet-2 report only reckons the death rate went from about 5.5 to 13.3

So how many morgues closed? Why this big increase in the arrivals at one morgue?

Quote:

Most of the victims were aged between 15 and 44. Most of them were male, the article said.

I wish they'd given a figure for that "most". And I wish they had told us how many died a violent death.

[Leif Roar]

Originally Posted by Puppycow

Split from this thread for scientific input

Reference


Is it possible that about 20 deadly car bombings occur every day in Iraq for 2 years, but that 90% go unreported despite causing casualties, while the few that do get reported generate an average of 6 independent reports?

I take it you're not familiar with statistics as a field. The Lancet study has a 95% confidence interval of 392 979 to 942 636 excess deaths. In other words, the width of the 95% confidence interval is a full 84% of the estimated number of deaths. That's for the total sample size of 1849 households. The precision goes down quickly (i.e. the width of the confidence interval increases) with reduced sample size.

What this means is that the precision for any subset of the reported deaths -- such as those deaths caused by car bombs -- is going to be wastly reduced compared to the precision for the whole study. The width of the 95% confidence interval for the number of people killed by car bombs is going to be a lot wider than 84% of the estimated number of deaths.

Just because the Lancet study makes a given statement about the total number of excess deaths, it does not follow that it make a proportionally equal statement about a subset of those deaths. That's just not how statistical analysis work. You must take into account how your change of sample size affects the confidence interval.


Now, the numbers from the Lancet reports -- particularly the mid to high range of the confidence interval -- strikes me too as unexpectedly high and I would not be surprised if it turns out that there is a problem with applying the used methodology to the situation in Iraq. However, you can not attack the methodology of the study based on hypothetical flaws. If you can not point to a real, actual flaw you will have to, tentatively, accept the methodology, if not the result, as sound.

Wether you accept the result or not, the Lancet study is the best there currently is. There are no other thourough studies (the IBC is a lower limit body count and not an estimate of total number of deaths) and the only bodies who seem capable of doing a more thourough job, the Iraq government and the Coalition forces, appear uninterested. (Personally I find it morally inexcuseable that the U.S. and U.K. military do not even attempt to estimate the number of civilian deaths.)

The Lancet study is not the final word in the matter, and the great difference between the estimate made it and the lower boundary counted by the IBC do call for caution and closer scrutiny. However, it is nonetheless a solid piece of work, and you can't just dimiss it simply because you don't like the conclusion it reaches.

[Puppycow]

Originally Posted by Leif Roar

I take it you're not familiar with statistics as a field. The Lancet study has a 95% confidence interval of 392 979 to 942 636 excess deaths. In other words, the width of the 95% confidence interval is a full 84% of the estimated number of deaths. That's for the total sample size of 1849 households. The precision goes down quickly (i.e. the width of the confidence interval increases) with reduced sample size.

What this means is that the precision for any subset of the reported deaths -- such as those deaths caused by car bombs -- is going to be wastly reduced compared to the precision for the whole study. The width of the 95% confidence interval for the number of people killed by car bombs is going to be a lot wider than 84% of the estimated number of deaths.

Just because the Lancet study makes a given statement about the total number of excess deaths, it does not follow that it make a proportionally equal statement about a subset of those deaths. That's just not how statistical analysis work. You must take into account how your change of sample size affects the confidence interval.

I am not an expert on statistics, but I do have a basic grasp, such as p values or confidence intervals. The confidence interval is a function of the population, and in the case we are discussing the population for car bombs is the same as the total population, therefore the 95% confidence interval should be the same.

Quote:

Now, the numbers from the Lancet reports -- particularly the mid to high range of the confidence interval -- strikes me too as unexpectedly high and I would not be surprised if it turns out that there is a problem with applying the used methodology to the situation in Iraq. However, you can not attack the methodology of the study based on hypothetical flaws. If you can not point to a real, actual flaw you will have to, tentatively, accept the methodology, if not the result, as sound.

Then let me point out an actual flaw, which is that a few individuals (respondants and/or data gatherers) acting alone by giving false answers can bias the data. And the bias is likely to be only in one direction. Its hard to imagine someone not reporting actual applicable deaths, but easy to imagine someone reporting non-existant or non-applicable deaths.

Quote:

Wether you accept the result or not, the Lancet study is the best there currently is. There are no other thourough studies (the IBC is a lower limit body count and not an estimate of total number of deaths) and the only bodies who seem capable of doing a more thourough job, the Iraq government and the Coalition forces, appear uninterested. (Personally I find it morally inexcuseable that the U.S. and U.K. military do not even attempt to estimate the number of civilian deaths.)

The Lancet study is not the final word in the matter, and the great difference between the estimate made it and the lower boundary counted by the IBC do call for caution and closer scrutiny. However, it is nonetheless a solid piece of work, and you can't just dimiss it simply because you don't like the conclusion it reaches.

I also find it morally inexcuseable that the U.S. and U.K. military do not estimate the number of civilian deaths, but whether I like the conclusion of the Lancet study or not has nothing to do with it. I am questioning the conclusions based on common sense. In particular, not only the car bomb component, but also the component of aircraft-dropped bombs are orders of magnitude higher than what the news media notices. Including Al Jazeera and many other anti-US news organizations that would love to report such news if it existed.

The flaws are not hypothetical, because the results disagree with observations and defy common sense.

[PixyMisa]

Originally Posted by Puppycow

Then let me point out an actual flaw, which is that a few individuals (respondants and/or data gatherers) acting alone by giving false answers can bias the data. And the bias is likely to be only in one direction. Its hard to imagine someone not reporting actual applicable deaths, but easy to imagine someone reporting non-existant or non-applicable deaths.

There's something interesting going on there, though. The Lancet researchers realised this was a problem, and asked to see death certificates. They reported that the great majority of respondents were able to produce death certificates matching their statements.

Which is good survey work.

The problem then is that the number of death certificates required to match the estimated "excess deaths" greatly exceeds those issued by the government.

The numbers generated from the survey do not match up to reality. The statistical methods used seem to be sound; that means that the data is bad. Why the data is bad I don't know, but it's the only reasonable conclusion.

[FireGarden]

Originally Posted by Puppycow

Then let me point out an actual flaw, which is that a few individuals (respondants and/or data gatherers) acting alone by giving false answers can bias the data. And the bias is likely to be only in one direction. Its hard to imagine someone not reporting actual applicable deaths, but easy to imagine someone reporting non-existant or non-applicable deaths.

In Lancet-2 (and possibly Lancet-1, also, I don't know for sure), families that reported a death were asked for death certificates. More than 80% provided one. Here's the kicker: they were asked at the end of the interview -- not the beginning. After they had already reported the death, they were asked for the certificate.

And for those that didn't provide a certificate, the numbers of deaths they gave matched with the 80%+ who did give certificates. (I don't know the extent of the match)

That, in my opinion, leaves only two groups of people you can accuse of fraud: the surveyors and/or those who collated their data.

So if you're going to say fraud -- say it clearly and name the real culprits. With luck, you might get sued.


ETA:
And I should add, you are not giving an actual flaw -- you giving what you think might have happened. It's still hypothetical.

[casebro]

Originally Posted by Leif Roar

... The Lancet study has a 95% confidence interval of 392 979 to 942 636 excess deaths..

A 95% confidence level of between four and nine. Sounds like a pretty sound estimate to me. A 240% range? Not anything I would bet on.

[PixyMisa]

Originally Posted by FireGarden

In Lancet-2 (and possibly Lancet-1, also, I don't know for sure), families that reported a death were asked for death certificates. More than 80% provided one. Here's the kicker: they were asked at the end of the interview -- not the beginning. After they had already reported the death, they were asked for the certificate.

Yep.

As I said, the problem is that this leads to an estimate that requires significantly more death certificates than were ever actually issued. (At least, at anything but the lowest point of the confidence interval.)

That doesn't imply fraud, but it tells us that the whole thing has to be screwed up somehow. Was the sampling skewed? Were the death certificates not checked properly? Was there after all some error in the statistical analysis? I don't know. But dead people get counted, even in Iraq, and the counts simply don't match the estimates.

[Puppycow]

Originally Posted by PixyMisa

Yep.

As I said, the problem is that this leads to an estimate that requires significantly more death certificates than were ever actually issued. (At least, at anything but the lowest point of the confidence interval.)

That doesn't imply fraud, but it tells us that the whole thing has to be screwed up somehow. Was the sampling skewed? Were the death certificates not checked properly? Was there after all some error in the statistical analysis? I don't know. But dead people get counted, even in Iraq, and the counts simply don't match the estimates.

It doesn't imply fraud, but neither is it inconsistent with fraud either, is it? You cannot rule out the possibility of fraud. (And if, as you suggest, the death certificates were perhaps not checked properly, that seems to approach fraud, too). I suppose a third party could in theory go to those same houses and ask to see the certificates again. I wouldn't go so far as to accuse them of fraud, but neither can I rule it out.

I wish the news media would do a better job of following up the story, instead of just reporting these studies without any attempt to confirm their results through investigative journalism. Because if the results are true, then there are majors stories out there just waiting for someone to report them. And if they are not true, it would be nice to know that, too

[a_unique_person]

Originally Posted by PixyMisa

There's something interesting going on there, though. The Lancet researchers realised this was a problem, and asked to see death certificates. They reported that the great majority of respondents were able to produce death certificates matching their statements.

Which is good survey work.

The problem then is that the number of death certificates required to match the estimated "excess deaths" greatly exceeds those issued by the government.

The numbers generated from the survey do not match up to reality. The statistical methods used seem to be sound; that means that the data is bad. Why the data is bad I don't know, but it's the only reasonable conclusion.

According to the reports I have read and heard, there is no actual functional government in Iraq. There is nominally something that exists on paper, but that's as far as it goes. It rarely meets, very few people attend when it does, it cannot pass the most basic legislation, let alone address the important issues.

I also cannot understand the fixation on car bombs. There are enough small arms in Iraq to arm a civil war, which is essentially what is in progress. If you want to know how bloody a civil war can be, history has plenty of examples.

[a_unique_person]

Originally Posted by casebro

A 95% confidence level of between four and nine. Sounds like a pretty sound estimate to me. A 240% range? Not anything I would bet on.

The new study confirms the two previous estimates. Polling is a very accurate and proven technology. All of a sudden, it doesn't work.

[FireGarden]

Originally Posted by casebro

A 95% confidence level of between four and nine. Sounds like a pretty sound estimate to me. A 240% range?

I don't understand this.
They needed to make the interval as wide as it needed to be to give us a 95% confidence level.

Of course, if they'd taken more samples the interval would have been smaller. But it would still have been a 95% confidence interval.

[FireGarden]

Originally Posted by PixyMisa

Yep.

As I said, the problem is that this leads to an estimate that requires significantly more death certificates than were ever actually issued. (At least, at anything but the lowest point of the confidence interval.)

More death certificates than were officially counted. I wonder how many were printed.

When a spouse dies, the other partner needs a death certificate for a few things by law. Ownership of land/houses, pensions, insurance.... There is a lot of pressure to issue death certificates.

But counting them is another matter.
Once they're issued, a copy or reference has to be sent someplace to be counted. That doesn't mean it will be counted, even assuming it gets there.

And all experience indicates that, as the violence gets worse and worse, a passive body count gets less and less accurate. You can't just wait for the news of each death to reach you.

It's not a huge surprise that deaths are undercounted in warzones.

Quote:

But dead people get counted, even in Iraq, and the counts simply don't match the estimates.

Kind of begs the question, "why do people bother with sampling if counting does just as well?"

After all, sampling has been standard in so many places. But suddenly it gets all screwed up in Iraq. More to the point, there are easy ways to "debunk" lancet-2. As I said above, just visit the morgues, cemetries, etc. What is the violent/non-violent death ratio amongst those getting buried?

Has there been a report from a morgue in Iraq along the lines of "We used to get 200 bodies a month, now we get 250"?

Or are all the reports like the one I quoted above? From 200 to 1100. Are there a quarter as many morgues as there used to be? And that was from 2005 -- the level of violence we'd all slap ourselves on the back for returning to.

[Cuddles]

Originally Posted by a_unique_person

The new study confirms the two previous estimates. Polling is a very accurate and proven technology. All of a sudden, it doesn't work.

It depends. Polls have also shown that 84% of people in the US believe that their government was responsible for the Twin Towers. Not just let it happen, but actually made it happen. Was this poll accurate? Of course, the Lancet study is likely to be a lot more relibale than this, but that doesn't mean you can just blindly say "Polling is reliable, therefore these figures are right". If the figures don't seem to match up, questioning whether the poll was valid is a perfectly sensible line of enquiry.

[PixyMisa]

Originally Posted by FireGarden

More death certificates than were officially counted. I wonder how many were printed.

Why should there be a difference? The only reason you are suggesting this is because the evidence is against the survey being correct. You assume the survey is right - why not assume that the count of death certificates is correct?

Quote:

And all experience indicates that, as the violence gets worse and worse, a passive body count gets less and less accurate. You can't just wait for the news of each death to reach you.

Except that government efficiency has steadily increased in Iraq over the past few years. There have been more civilian deaths after 2005 because AQ have largely given up attacking military targets and concentrated on murdering civilians

Quote:

Has there been a report from a morgue in Iraq along the lines of "We used to get 200 bodies a month, now we get 250"?

I believe there have been a lot of such reports - after all, if you look almost anywhere outside Baghdad, things are pretty quiet. But does a report that says "deaths are up from 200 per month to 210" get any attention whatsoever?

[FireGarden]

Originally Posted by PixyMisa

Why should there be a difference? The only reason you are suggesting this is because the evidence is against the survey being correct. You assume the survey is right - why not assume that the count of death certificates is correct?

There should be a difference because there is a different motivation between issuing the certificates and counting the certificates. And the different motivation explains the different efficiency.

There are so many things that a bereaved family member cannot do without a death certificate for whoever died. Can they even bury their loved one in a cemetry without a death certificate? Can you just turn up with a dead body and say "Here, would you bury this for me?"

Then there's insurance, inheritance, etc.
So the bereaved will make sure they get a death certificate. What happens after that? The counting isn't automated, as far as I know. And experience from other war-zones indicates that surveys get a better measure of death rates. Especially when the violence is high. Les Roberts quoted some figures in some of the interviews he gave.

Quote:

I believe there have been a lot of such reports - after all, if you look almost anywhere outside Baghdad, things are pretty quiet. But does a report that says "deaths are up from 200 per month to 210" get any attention whatsoever?

I agree, the other type of news is more likely to be reported. And we obviously need reports from a wide range of places.

Things weren't quiet outside Baghdad throughout the war, mind you.

[ben m]

If you did a survey of black eyes and bruises in the US, you'd "discover" an epidemic of bumping into doors and falling down stairs. Why? Because those are the "safe" answers which are used to cover up for domestic violence. Telling a surveyer "My husband gave me these bruises" may invite retaliation. In Iraq, telling a surveyer "My brother was killed execution-style by the Sadrists in the police brigade" may invite that same brigade to target you for non-cooperation.

Just a hypothesis, but this has been covered on several sites with good knowledge of the details:

http://www.samefacts.com/archives/th...09/1220580.php

http://scienceblogs.com/deltoid/2007...i_death_to.php et. seq.

[a_unique_person]

Originally Posted by PixyMisa

I believe there have been a lot of such reports - after all, if you look almost anywhere outside Baghdad, things are pretty quiet. But does a report that says "deaths are up from 200 per month to 210" get any attention whatsoever?

I don't know where you get that idea from.

[Mycroft]

Originally Posted by FireGarden

There should be a difference because there is a different motivation between issuing the certificates and counting the certificates. And the different motivation explains the different efficiency.

So what you're saying is you believe there is an Iraqi conspiracy to cover the true number of deaths?

[FireGarden]

Originally Posted by Mycroft

So what you're saying is you believe there is an Iraqi conspiracy to cover the true number of deaths?

Not necessarily.
Doctors have many important jobs, but they HAVE to give out the death certificates because the certificates will be asked for until received. This is not a task they can put off or avoid for long.

Is everything after that automated?
What are the responsibilities of everyone in the rest of the system? How do they prioritise the reporting and counting of death certificates?

[Dancing David]

Originally Posted by Puppycow

Split from this thread for scientific input

Reference


Is it possible that about 20 deadly car bombings occur every day in Iraq for 2 years, but that 90% go unreported despite causing casualties, while the few that do get reported generate an average of 6 independent reports?

What is your data concern here?

When you say 'unreported', did you read the lancet study?

How did they determine what you are asking about?

I am not saying that the Lancet is correct in every form of epidemiology they use.

I am willing to say that the level of data from a war zone is very, um, filtered. The Iraqi government is not the, uh, most stable and there are a lot of political reasons that there might be underreporting. Certainly the systems in Iraq are completely overwhelmed, using morgue statistics and media reports is not going to be accurate, nor are the Iraqi government statistics.

And most certainly the, um, data we receive here in the US is completely, um, filtered to make, um, certain people's cases without, um, accurate data to counter the claims they make.

I am not going to sit here and pass judgment, either way. Most of the epidemiology used by the Lancet is very accurate, how they came to the conclusion of twenty car bombs I would have to actually look at. The fact that we in the US don't even get any accurate data at all is rather telling. Compared to WWII and Vietnam there is an amazing control of the information we receive in the US.

So I will have to find out how they came to that conclusion before passing my limited judgment.

And if you consider that total control of the data stream that the media does receive, how do we get ‘independent’ sources?

The car bomb statistics do not obliviate the accuracy of the rest of the report.

[Dancing David]

I just read quickly the lancet article as reprinted on google
http://66.102.1.104/scholar?hl=en&lr...alityiraq.pdf+

Before I eveuate the accuracy of the statement that they say there were twenty car boms a day, I have to ask, how was that statement made?

Can you give me the data that suggest that The Lancet made this statement or how it was extrapolated from their data?

In my web search i found some great errors like someone on Wikipedia thinking that the population of Iraq was twenty thousand, and then people discussing the Iraq Body Count project as though it was meaningful

The lancet study is very upfront about how the data was gathered through reports of indivduals in interviews.

So how was the statement that is the basis of the thread arrived at?

Thank You.

[Mycroft]

Originally Posted by FireGarden

Not necessarily.
Doctors have many important jobs, but they HAVE to give out the death certificates because the certificates will be asked for until received. This is not a task they can put off or avoid for long.

Is everything after that automated?
What are the responsibilities of everyone in the rest of the system? How do they prioritise the reporting and counting of death certificates?

So it's not a conspiracy theory, but your theory is that Iraqis are just inept?

Other than supposition, do you have any evidence to back up this theory?

[FireGarden]

Originally Posted by Mycroft

So it's not a conspiracy theory, but your theory is that Iraqis are just inept? Other than supposition, do you have any evidence to back up this theory?

Inept compared to whom?
Lot's of governments had difficulty counting deaths in violent war-zones. That in itself is one piece of evidence.

A second is the Lancet Study.
Another is the ORB poll -- though a lower quality survey than Lancet.

So, basically, Lancet-2 tells us that what tends to happen in violent war-zones (under-counting of dead) is happening in Iraq.

If you don't like Burnham et al, get a better survey. Lobby your government to send a survey-team.