Iraq death toll

[Judge Hades]

If you go with your gut would you be using your stomach instead of your heart?

[taks]

colbert is certainly a funny guy.

 

stats are very easy to pervert, with or without intent.

 

in the end, however, "going with your gut" is what happens once media (of any format) present a story to the public. both in the way the reporter writes the piece, and in the way the public interprets it.

 

taks

[Judge Hades]

If you pervert stats I don't think you are using either your gut or your heart.

 

Ew.

Edited October 12, 2006 by Judge Hades

[kumquatq3]

So far as I know this group is the only group actually trying to find out the effects of the war.

 

Not even close

 

They range from the website Iraq Body Count to Various independent estimates that seem to use similar process (This one here, albeit with less total people, but with a similar cluster strategy. At least, on the surface) and military estimates

[taks]

If you pervert stats I don't think you are using either your gut or your heart.

<{POST_SNAPBACK}>

uh, not in that sense...

 

taks

[Judge Hades]

Oh, that's right. We don't have a democrat in the presidency. :D

[Colrom]

By the way, ignore that stuff I posted about 1/(1.6^2). I am too simple for fancy stuff like that.

 

They say right out what they mean.

[Craigboy2]

http://news.bbc.co.uk/1/hi/world/middle_east/6040054.stm

 

I can't see anything much wrong with their methodology. But I do think there must be something to explain the massive discrepancy between this 650,000 number and all the others. Anyone know more about this? I haven't had time to look into it in more detail.

<{POST_SNAPBACK}>

I don't think it's completely accurate. Kind of like those smoking Ads with really high death rates every year of the people who have died from smoking. But sometimes they just look at anyone who has died of influenza and count them as a victim of a smoking related death. I think it

[Colrom]

This study is certainly different than a body count study. It will include indirect deaths and stress deaths. For example if a person dies of infection from stepping on a nail because they could not get to a hospital because it was destroyed they might add to the pool of "excess deaths" relative to before the invasion. If they die because they get a disease from water that is no longer clean because the purification technology is destroyed they might add to the pool too, although those deaths would likely show up pre war as well because we destroyed their water system and their water purification capabilities pre war.

 

Certainly the US military is well familiar with these types of casualties. They figured them into the estimates that were made for likely casualties from nuclear exchanges.

[Tigranes]

The sampling bias is definitely there and that is one of the reasons people should trust the 600,000 figure as much as they should trust 30,000. this is like trusting Arrian when he says Alexander defeated a million Persians with 30,000 troops.

 

If we accept 30,000, for about... 2 years,

 

THat's 41 deaths a day. Not a very high figure at all for a country in turmoil.

 

And it doesn't count military deaths.

 

And the initial period of the war must jack up the figures.

 

Then, we must look at the entire 2003-2006 period (though of course events during that period are not consistent).

 

All that considered, even the Iraqi Body Count, in the full period and including military units, would go near 6 figures. And since they only count bodies (how stupid is that), it's certainly lower than it should be.

[Colrom]

Sorry.

Edited October 13, 2006 by Colrom

[alanschu]

All that considered, even the Iraqi Body Count, in the full period and including military units, would go near 6 figures. And since they only count bodies (how stupid is that), it's certainly lower than it should be.

 

Err, I am kind of confused, but how else do you determine the number of people dead, if counting the number of people that have died is "stupid."

 

It's a lower estimate (since people have probably died that you didn't see), but anything higher than that is just guessing.

[Colrom]

All that considered, even the Iraqi Body Count, in the full period and including military units, would go near 6 figures. And since they only count bodies (how stupid is that), it's certainly lower than it should be.

 

Err, I am kind of confused, but how else do you determine the number of people dead, if counting the number of people that have died is "stupid."

 

It's a lower estimate (since people have probably died that you didn't see), but anything higher than that is just guessing.

<{POST_SNAPBACK}>

 

Iraq body count does not count the number of people who have died. It counts a heavily filtered subset of that. Like I said elsewhere it is a miracle that they count anybody at all given all their constraints.

 

The Iraqi University, Johns Hopkins MIT procedure does count a sample of the number of people who have died.

 

Interestingly a subset of their results covering just the period before the war (to provide a baseline) matches the US CIA estimates for Iraqi mortality in that period.

 

I remember when the genocide in Cambodia was in process and the reports came out how most people really didn't want to hear about it.

 

I gather that it was that way in regard to reports of the holocost too.

 

People don't want to believe it.

 

It would mean that some people they like are covered in blood.

 

It would mean they really ought to do something.

 

But they don't know what to do.

 

So many deny it and many push it aside.

 

Not a pretty picture.

Edited October 13, 2006 by Colrom

[Tigranes]

As Colrom explained at more length earlier, the Iraqi Body Count is just that. It counts bodies.

 

A cluster sample is infinitely more accurate, even when there is sample bias.

[taks]

A cluster sample is infinitely more accurate, even when there is sample bias.

<{POST_SNAPBACK}>

based on what? your assumption that all of the procedures and statistics are valid? which is, again, based on what? infinitely even in the presence of sample bias? that statement alone refutes the concept of any sort of statistical accuracy, which is what is being used.

 

when there is bias in any sample, it cannot be trusted and is therefore, inaccurate. floor to ceiling confidence intervals are not "infinitely more accurate" than a body count. if anything, they are infinitely less accurate*. i.e., if there is sampling bias, your confidence intervals cannot be calculated (properly) and your range for the estimate is therefore anywhere from 0 to the max that could have died (population of iraq + insurgents + military, etc.). an answer cannot be less accurate than "somewhere within the total population."

 

taks

 

* certainly the word infinitely w.r.t. the word less is not the best use of grammar, but it suffices to make the point.

[Judge Hades]

Okay, I think trying to come up with an exact total is rather silly. In a war torn area such as many areas are in Iraq making an accurate assessment of deaths are next to impossible. Lets just say many many many many many people have died. Many many many many many many many many more will die. It just seems to me that the death rates in Iraq is the same now as it was with Saddam's rule.

 

No worse but no better.

[taks]

more sense from the usually senseless. there is hope, hades, hope...

 

presonally, however, i do not think it is impossible to uncover the true (er, nearly true) number. just very very very difficult (a fourth very might well be impossible).

 

taks

[~Di]

People will believe, and justify that belief, in whichever figures most suit their own political bias, in my view. Those who hate the Iraq war will leap upon the larger figure as a way to demonstrate why their disapproval of the war is righteous. Those who support the Iraq war will probably dismiss the larger figure, since it would lend credence to those who think the war is a really bad thing.

 

I think the war is a really bad thing. That said, I really think this figure is grossly exaggerated. I'm sure everyone would agree that using 39 households in the Sunni section of Baghdad would garner ten times the deaths that using 39 households down in the Shia stronghold of Basra, or up in the Kurdish area would garner. Color me suspicious, but this seems to me to be a bit of a propaganda tool by the anti-war group... of which I am one, but still... According to what I read on another forum, for this figure to be correct 722 people would have to have died PER DAY for the past 2.5 years. Do y'all really think the global media would have missed that?

 

I'm more inclined to believe the IBC page because it includes all proven deaths, both combatants and non-combatants, that have died war-related deaths and supports those deaths with detail... how, why, when, where and in many cases, who.

Edited October 13, 2006 by ~Di

[taks]

that's exactly what i mean by sample bias, di.

 

taks

[Judge Hades]

Then I got to work on that. There has to be absolutely no hope for me, otherwise i wouldn't be in character.

[alanschu]

Iraq body count does not count the number of people who have died. It counts a heavily filtered subset of that. Like I said elsewhere it is a miracle that they count anybody at all given all their constraints.

 

You didn't really clarify things.

 

What exactly is the body count? It seems to me to be counting the number of people that have died. But you say that it doesn't count the number of people that have died. So what exactly does the body count count (and please don't say bodies).

[Colrom]

Iraq body count does not count the number of people who have died. It counts a heavily filtered subset of that. Like I said elsewhere it is a miracle that they count anybody at all given all their constraints.

 

You didn't really clarify things.

 

What exactly is the body count? It seems to me to be counting the number of people that have died. But you say that it doesn't count the number of people that have died. So what exactly does the body count count (and please don't say bodies).

<{POST_SNAPBACK}>

 

Here is the Iraq Body Count description of their methodology with text bolded in areas that relate to your question. I will talk about it more in a separate post. This post is intended just to establish a background for answering your question.

 

IBQ Methodology

 

1. Overview

 

Casualty figures are derived from a comprehensive survey of online media reports and eyewitness accounts. Where these sources report differing figures, the range (a minimum and a maximum) are given. All results are independently reviewed and error-checked by at least two members of the Iraq Body Count project team in addition to the original compiler before publication.

 

 

2. Sources

 

Our sources include public domain newsgathering agencies with web access. A list of some core sources is given below. Further sources will be added provided they meet acceptable project standards (see below).

 

ABC - ABC News (USA)

AFP - Agence France-Presse

AP - Associated Press

AWST - Aviation Week and Space Technology

Al Jaz - Al Jazeera network

BBC - British Broadcasting Corporation

BG - Boston Globe

Balt. Sun - The Baltimore Sun

CT - Chicago Tribune

CO - Commondreams.org

CSM - Christian Science Monitor

DPA - Deutsche Presse-Agentur

FOX - Fox News

GUA - The Guardian (London)

HRW - Human Rights Watch

HT - Hindustan Times

ICRC - International Committ of the Red Cross

IND - The Independent (London)

IO - Intellnet.org

JT - Jordan Times

LAT - Los Angeles Times

MEN - Middle East Newsline

MEO - Middle East Online

MER - Middle East Report

MH - Miami Herald

NT - Nando Times

NYT - New York Times

Reuters - (includes Reuters Alertnet)

SABC - South African Broadcasting Corporation

SMH - Sydney Morning Herald

Sg.News - The Singapore News

Tel- The Telegraph (London)

Times - The Times (London)

TOI - Times of India

TS - Toronto Star

UPI - United Press International

WNN - World News Network

WP - Washington Post

 

For a source to be considered acceptable to this project it must comply with the following standards: (1) site updated at least daily; (2) all stories separately archived on the site, with a unique url (see Note 1 below); (3) source widely cited or referenced by other sources; (4) English Language site; (5) fully public (preferably free) web-access.

 

The project relies on the professional rigour of the approved reporting agencies. It is assumed that any agency that has attained a respected international status operates its own rigorous checks before publishing items (including, where possible, eye-witness and confidential sources). By requiring that two independent agencies publish a report before we are willing to add it to the count, we are premising our own count on the self-correcting nature of the increasingly inter-connected international media network.

Note 1. Some sites remove items after a given time period, change their urls, or place them in archives with inadequate search engines. For this reason it is project policy that urls of sources are NOT published on the iraqbodycount site.

 

 

3. Data extraction

 

Data extraction policy is based on 3 criteria, some of which work in opposite directions.

 

Sufficient information must be extracted to ensure that each incident is differentiated from proximate incidents with which it could be potentially confused.

Economy of data extraction is required, for efficiency of both production and public scrutiny.

Data extraction should be uniform, so that the same information is available for the vast majority of incidents. This is best guaranteed by restricting the number of items of information per incident to the core facts that most news reports tend to include.

The pragmatic tensions in the above have led to the decision to extract the following information only for each incident:

 

Date of incident

Time of incident

Location of incident

Target as stated by military sources

Weapon (munitions or delivery vehicle)

Minimum civilian deaths (see Note 2)

Maximum civilian deaths (see Note 2)

Sources (at least two sources from the list in section 2 above) Reliability of data extraction will be increased by ensuring that each data extraction is checked and signed off by two further independent scrutineers prior to publication, and all data entries will be kept under review should further details become available at a later date.

 

Note 2. Definitions of minimum and maximum

 

Reports of numbers dead vary across sources. On-the-ground uncertainties and potential political bias can result in a range of figures reported for the same incident. To reflect this variation, each incident will be associated with a minimum and maximum reported number of deaths. No number will be entered into the count unless it meets the criteria in the following paragraphs. This conservative approach allows relative certainty about the minimum.

 

Maximum deaths. This is the highest number of civilian deaths published by at least two of our approved list of news media sources.

 

Minimum deaths. This is the same as the maximum, unless at least two of the listed news media sources publish a lower number. In this case, the lower number is entered as the minimum. The minimum can be zero if there is a report of "zero deaths" from two of our sources. "Unable to confirm any deaths" or similar wording (as in an official statement) does NOT amount to a report of zero, and will NOT lead to an entry of "0" in the minimum column.

 

As a further conservative measure, when the wording used in both reports refers to "people" instead of civilians, we will include the total figure as a maximum but enter "0" into the minimum column unless details are present clearly identifying some or all of the dead as civilian - in this case the number of identifiable civilians will be entered into the minimum column instead of "0". The word "family" will be interpreted in this context as meaning 3 civilians. [Average Iraqi non-extended family size: 6. -CIA Factbook 2002.]

 

4. Data storage

 

Although it is expected that the majority of sources will remain accessible on the web site from which they were drawn, the project will create a secure archive of all original sources (in both electronic and paper form). Where judged appropriate by the project team, this data may be released to bona-fide enquirers, for verification purposes. At an appropriate juncture, the entire archive will be passed to an institution of public record (such as a University or National Library) for permanent access by bona-fide researchers. The copyright of original sources will remain with the originators. The copyright of the Iraq Body Count data extraction remains with the named researchers on the project (see About us).

 

5. Publication of data (including conditions of use)

 

Once verified through the processes described in section 3 above, each new incident will be added as a new line on a spreadsheet database which will be updated regularly (at least daily) on the www.iraqbodycount.org site. The total minimum and maximum deaths will be automatically updated, and will feed through to all remotely positioned web-counters donwloaded from the site.

 

Permission is granted for any individual or agency to download and display any of the web counters available on this site, provided that the link back to the www.iraqbodycount.org site is not disabled or otherwise tampered with when displayed on a live interactive web-site. Permission is also granted for cut-and-paste downloads of the spreadsheet database listing each incident. All press and non-commercial uses are permitted. Other commercial uses are prohibited without explicit permission (contact info@iraqbodycount.org).

 

We request that you acknowledge any use of the Iraq Body Count data base or its methodology by mentioning either the project name ("Iraq Body Count") or the url (www.iraqbodycount.org) or the names of the principal researchers, Hamit Dardagan and John Sloboda.

 

 

6. Limitations and scope of enquiry:

 

Any project has limitations and boundaries. Here are some FAQs about this topic and our answers to them.

 

Why don

Edited October 13, 2006 by Colrom

[Colrom]

IBQ really doesn't count bodies or deaths. They count civilian deaths reported by selected news media on the internet for newsreported events.

 

If fifty people are brought to a morgue with gunshot wounds they will not be counted by IBQ unless they are reported over the internet as deaths resulting from one or more well defined events (which, most recently, can be just being found together at a location) reported by two approved news agencies.

 

If ten civilians are killed and forty people are wounded at an event reported by two approved news sources the ten civilians will be counted by IBQ but no record of the forty wounded will be recorded. If fifteen of those forty die in the next few days but those deaths are not reported in association with the causative incident by two approved news sources then those fifteen deaths will not be counted by IBQ.

 

By the way I can't figure out some of the news symbols in the latest listings and some of the events list news organizations not given in their methodology.

 

It is hard to dive into the details of their data and check anything.

 

Does that help?

Edited October 13, 2006 by Colrom

[alanschu]

Not really, because it seems as though we aren't talking about the same thing.

 

I'm not talking about the "Iraq Body Count" website or whatever, but rather your comment about:

 

"And since they only count bodies (how stupid is that)"

 

 

Given the bouncing around between capitalization of "Iraq Body Count" being referred to as "Iraq body count," I'm not really sure when people are talking about the website or if they are talking about the actual body count of Iraqi people.

 

I would never dispute that Iraq Body Count could be wrong. I was curious why counting the number of dead bodies is a "stupid" form of assessing casualties. I know it wouldn't be accurate, but at least it wouldn't be purely guessing.

[Tigranes]

When talking not about the website but about the methodology of 'counting bodies';

 

It's a way, but not a good way. Its not as if you have a team of people scouring the countryside for.. uh, bodies. If they rely on public reports, some people will not report deaths. Some bodies will be 'taken care of'. Some bodies will be deep in the countryside or in mountainous areas. If you dont know about a death and the body isn't easily found then you would completely miss that, etc, etc.

 

Cluster samples using people with death certificates run into a similar set of problems regarding 'hidden' deaths, but they draw somewhat closer , if not by much, by getting away from the needless obsession with the physical body.

 

As I say, both ways will be biased. And if faced with 30,000 and 600,000, it will be somwehre in the middle. Probaly 100,000-200,000.