I am, too, skeptical about these statistical inferences - and I read in Danish media that statisticians were as well - but sadly I don't have time to go through them at this point. But right off the bat, I'm convinced that this bit

quote:

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.

is the standard mis-interpretation of confidence interval that journalists are so good at making. Given the methodology of the method, the lower bound on the range of uncertainty extends at least to (40*47=1880).

Anyway, no matter how you look at it the death toll is terrible. I tend to agree with pkc that

quote:

Originally posted by pkcRAISTLIN whenever there was a power vacuum in iraq (and it would have happened eventually) i dont see the figures being that much different.

quote:

Originally posted by MisterOpus1 I did not read anywhere that stated this as such. Rather, they were investigating and comparing the death rate prior to the war versus after the war, and they then separated out those deaths in accordance to the method of death (cancer, gunshot, etc.). No where did I see such a conclusion that such deaths were made by the Iraqi War itself on nonviolent deaths. Is there a quote I overlooked that points this out?

Page 1, the findings, copied here and highlighted:
Findings Three misattributed clusters were excluded from the final analysis; data from 1849 households that contained
12 801 individuals in 47 clusters was gathered. 1474 births and 629 deaths were reported during the observation
period. Pre-invasion mortality rates were 5·5 per 1000 people per year (95% CI 4·3–7·1), compared with 13·3 per
1000 people per year (10·9–16·1) in the 40 months post-invasion. We estimate that as of July, 2006, there have been
654 965 (392 979–942 636) excess Iraqi deaths as a consequence of the war, which corresponds to 2·5% of the
population in the study area. Of post-invasion deaths, 601 027 (426 369–793 663) were due to violence, the most
common cause being gunfire.

quote:

A closer look reveals your error - the survey has two timelines, pre-war and post-war. Pre-war examines the time between January 2002 and the invasion of March of 2003, versus the post-war period between March of '03 to July 2006. So that's 1 yr., 2 mo. pre-war versus 3 year, 4 mo. post-war. You are comparing the raw data between two very different time periods, which the latter (post-war) being >2x greater than the former (pre-war). Furthermore, look at the sample size that you're comparing: prewar: n = 82 postwar: n = 547 The more relevant data are the comparisons per 1000 people in the later Tables and Figures, which normalizes the data between the two variables considered.

Thing is I don't want to do the math to extrapolate to per 1000 comparisons, and at this point I don't think we need to.

First, it strikes me odd that they looked at only 1 year and 2 months before the invasion for there pre-sample, not the same period of 3 years and 4 months.

Second, if we do some very simple math: 80 (which repersents the number of non-violent deaths in 1 year and 2 months before the invasion) multiplied by 2.83333 (the number of years of the analysis after the invasion) we get 226. Which is a little lower than the 247 deaths they reported over a 3 year 4 month period after the invasion.

I believe when you get into it then, 21 deaths really to denote the whole difference of an extrapolation using two unbalanced samples you could potentially run into some areas of trouble.

Ok found a way to do a per 1000, according the sample the ~1850 or so people polled represent 11956 people. Plugging in the numbers above we get:
- 5.6 non-violent deaths per 1000 people a year for the pre war period
- 6.2 non-violent deaths per 1000 people a year for the post war period.

So thats with the non-violent deaths.

I find their violent death statistics uncomparable and therefore I don't look at them. For instance, in the pre-war period one person is reported to have died due to a coalition air strike. But I am unaware of any air strikes that were conducted before the invasion and if they can be usedas a repersentation.

Also, the complete unreporting of any violent deaths by crime or other circumstances before the pre-war period raises concerns on my part on how the survey was conducted.

___________________
SAVE ZIONIST MUSTARD: BUY ZIONIST KETCHUP!


Click here to support the free mustard alliance.

Just an interesting thing.
I was playing with the numbers now that I have the count of non-violent deaths per 1000.

According to the study Iraq has a population of 27,139,584.
Using the non-violent death rate of 6.2 per 1000 people a year, that means:
27139584/1000 = 27139.58
27139.58 * 6.2 = 168265.42 <-- number of non-violent deaths per year
168265.42 * 3.333 (timeframe of post-war study) = 560,828


So, if we are to assume that all the violent deaths are completely attributable to the war in Iraq, that would mean:

654965 (total death violent and non-violent over period) - 560828 (violent death over period) = 94,137 deaths.
Or 31,379 violent death a year. Still a staggering figure, but not as staggering for sure.

And much more inline with other estimates, such as Iraqiy (a source cited in the study which estimated 128,000 deaths) and the Iraq Body Count (~50,000). The Interior Ministry, as reported in the study predicted around 87,500 (75% higher than the bodycount).

___________________
SAVE ZIONIST MUSTARD: BUY ZIONIST KETCHUP!


Click here to support the free mustard alliance.

quote:

Originally posted by Yoepus Page 1, the findings, copied here and highlighted: Findings Three misattributed clusters were excluded from the final analysis; data from 1849 households that contained 12 801 individuals in 47 clusters was gathered. 1474 births and 629 deaths were reported during the observation period. Pre-invasion mortality rates were 5·5 per 1000 people per year (95% CI 4·3–7·1), compared with 13·3 per 1000 people per year (10·9–16·1) in the 40 months post-invasion. We estimate that as of July, 2006, there have been 654 965 (392 979–942 636) excess Iraqi deaths as a consequence of the war, which corresponds to 2·5% of the population in the study area. Of post-invasion deaths, 601 027 (426 369–793 663) were due to violence, the most common cause being gunfire.

I think you're ignoring the fact that non-violent deaths can be attributed to the Iraq war. Let's not forget that health programs and general well being (employment, wages, social stability, etc.) can be tied to deaths. The study correctly distinguishes between non-violent and violent deaths and truth be told even if we were to completely factor out the non-violent deaths that would only result in a difference of 50,000 or 60,000 people. Using cursory analysis, it looks like your math is wrong in your subsequent post about 6.2 deaths per thousand Iraqis because it looks like you're mistaking the absolute number of deaths with the excess number of deaths as a result of the Iraq war (which is what the study is attempting to segregate if I'm not mistaken). In other words, the standard Iraq death rate may be 600,000 but after the invasion, the death rate is 600,000 in excess of that. Please correct me if I misunderstood that.

I looked at your numbers for about 10 minutes so I could be wrong. However, I think that the polls findings are completely justifiable since all other liberal indiciations of death counts were based on media reports (Iraqbodycount.org, etc.) When you consider the media is concentrated around major cities, I'm sure a lot goes unnoticed around teh fringes.

[/QUOTE]

___________________
Retro ...

Nobody answered me, but - is there really a democracy in Iraq right now? Journalists are oppressed, violence, murder and crime are high, religious intolerance is growing, more Iraqis are unhappy with American regime ...

no doubt, even though it might not be 600,000, it has got to be a least a portion of that, at least 100,000 deaths. Thats a lot. Where's the value of human life? At this pace in 5-10 years there will be many more dead, as violence is on the rise. What about the civilian deaths in Afghanistan? Is there democracy there - no, the women are still haunted and threatened, prevented from reaching the same status as men, schools are closed because of sectarian pressure and fires, women still forced to wear those all-cover hoods, and what my Canadian peacekeepers are fighting for, is not democracy ... they're dying for nothing, for American agenda.

___________________
Whenever you go and buy something, you are affecting someone somewhere, be it environment, a person, or a community - you're making a statement with what you buy. So make it a smart choice ... Its a big picture

quote:

Originally posted by Renegade http://www.nytimes.com/2006/10/11/w...gin&oref=slogin The Lancet, if you remember, published a similar study a while ago placing the death toll at 100,000 and it was largely dismissed for its shoddy sampling and statistical inference. This study, however, seems to be much larger in scope and much better attested to by independently reached statistics: http://www.washingtonpost.com/wp-dy...s_world/mideast http://online.wsj.com/public/articl...ff_main_tff_top Here's the study in full if anyone with a better grasp of maths than me wants to poke holes in it: http://www.thelancet.com/webfiles/i...73606694919.pdf It goes without saying, but this is seriously disturbing if it's anywhere near accurate. Even Iraqbodycount.net places the civilian death toll at "only" 50,000: the minimum estimates for this study are 9 times that amount...

Well at least Bush has now got revenge for 3,000 deaths on 9/11

___________________

quote:

Originally posted by DJ Mikey Mike Social outcasts are often of the opinion that they must have a drink before being able to loosen up with their inhibitions, thus being able to have a good time. There's a word that sums up this sort of behaviour, and that word is 'reject.'

quote:

Originally posted by occrider I think you're ignoring the fact that non-violent deaths can be attributed to the Iraq war. Let's not forget that health programs and general well being (employment, wages, social stability, etc.) can be tied to deaths. The study correctly distinguishes between non-violent and violent deaths and truth be told even if we were to completely factor out the non-violent deaths that would only result in a difference of 50,000 or 60,000 people. Using cursory analysis, it looks like your math is wrong in your subsequent post about 6.2 deaths per thousand Iraqis because it looks like you're mistaking the absolute number of deaths with the excess number of deaths as a result of the Iraq war (which is what the study is attempting to segregate if I'm not mistaken). In other words, the standard Iraq death rate may be 600,000 but after the invasion, the death rate is 600,000 in excess of that. Please correct me if I misunderstood that.

No, I believe you are wrong. I am not exactly sure how the excess number of deaths is caclulated, but I believe it is something like absolute or near absolute. I am pretty certain they attribute 100% of the non-violent and violent deaths above the pre-war level to the Iraq war (whereas I don't because of the violent death discrepancy before the war).

But even when I assume there assumption there statistic is way off, let me demonstrate:

According to their Findings:
Findings Three misattributed clusters were excluded from the final analysis; data from 1849 households that contained
12 801 individuals in 47 clusters was gathered. 1474 births and 629 deaths were reported during the observation
period. Pre-invasion mortality rates were 5·5 per 1000 people per year (95% CI 4·3–7·1), compared with 13·3 per
1000 people per year (10·9–16·1) in the 40 months post-invasion. We estimate that as of July, 2006, there have been
654 965 (392 979–942 636) excess Iraqi deaths as a consequence of the war, which corresponds to 2·5% of the
population in the study area. Of post-invasion deaths, 601 027 (426 369–793 663) were due to violence, the most
common cause being gunfire.

So first of all the 40 month post-invasion period = 3.333 years
The population of Iraq according to the article is 27,139,584 (Table 1).

The difference in mortality rate between the periods is 13.3 - 5.5 = 7.8 deaths per 1000 per year.

So: ((27,139,584/1000) * 7.8 ) * 3.333 = 705,622.

So I'm assuming this is the how they got it, and they cut off some 50,000 deaths for this reason or anothe rto get to there figure.

However when you look at my complaints:
1) Why are there no violent deaths before the invasion?
2) Why was the pre-invasion period only 1 year and 2 months instread of 3 years and 4 months?
3) Using there data like infant births, old age births, one could as easily assume that the war increased the Iraqi population's health as decreased it, due to an influx of western medicine, food, highly trained specialist, clean water, equipment, etc.

___________________
SAVE ZIONIST MUSTARD: BUY ZIONIST KETCHUP!


Click here to support the free mustard alliance.

I haven't read the report yet, so I'm just basing this on the summaries I've read, but:

quote:

Originally posted by Yoepus So I'm assuming this is the how they got it, and they cut off some 50,000 deaths for this reason or anothe rto get to there figure.

I believe you're more or less right here: they are arriving at the figure of 654,000 by comparing pre-invasion mortality rates with post-invasion mortality rates. The "violent deaths" figure, however, doesn't seem to be an arbitrary "cut off" figure, at least as far as I can tell, but rather is based on information received from the death certificates they asked to see. In other words, presuming integrity of polling, a large percentage of the death certificates they saw would have listed the cause of death as violent in some way.

quote:

However when you look at my complaints: 1) Why are there no violent deaths before the invasion?

Good point, and accounting for this would probably reduce the figure somewhat, but even if all the deaths in pre-war Iraq were counted as violent deaths, there is still a large discrepancy. I don't have time (or the skill, really ) to do the maths, but I'd imagine that you'd still end up with a number of "excess" violent deaths far exceeding any of the previous body count estimates that have been thrown around before this one.

quote:

2) Why was the pre-invasion period only 1 year and 2 months instread of 3 years and 4 months?

Is this likely to make a difference? The pre-war mortality figures they're using are consistent with those reached in other studies (CIA, for instance, put pre-war mortality rates at 5.37 / 1000 - link) so do you have any reason to suspect that their pre-war estimates are somehow questionable, or serve to cast doubt on their post-war findings?

quote:

3) Using there data like infant births, old age births, one could as easily assume that the war increased the Iraqi population's health as decreased it, due to an influx of western medicine, food, highly trained specialist, clean water, equipment, etc.

Yes, but what statistics (or reliable reports, even) would indicate to you that "the Iraqi population's health" has improved since the invasion? Large parts of the country (including Baghdad for more than half the day) are without electricity (let alone access to drinking water, basic levels of sanitation or reliably delivered medical supplies) so I find it hard to believe that the hospitals are better equipped to deal with the sick, especially considering the numbers of violent casualties coming through, than they were before the war.

___________________
http://eschatonnow.blogspot.com/

Oh and here's a brief interview with John Zogby about it, if anyone cares (starts about 3 mins in):

http://www.rawstory.com/news/2006/V...re_of_1011.html

quote:

JOHN ZOGBY, ZOGBY INTERNATIONAL: The methodology of the survey, I think, from what I've seen so far is quite good, following all the rules of random sampling to a degree that it's possible in a country like Iraq, and cluster sampling. zeroing in on sampling points that are representative. I think where some of the disconnect may very well be is that this was indeed according to the methodology statement that I read a nationwide survey, including clusters of areas that are not within the daily purview of where the media are and where many public officials are who report those body counts. And so, I mean, translated, the media clustered in about five or six cities, and that's where much of the body count comes from. There is so much more to Iraq than just five or six cities.

Gotta kinda wonder when he says that "my company and others are able to call U.S. elections and European elections with pinpoint precision using a sample of a thousand [people]" though - wasn't it Zogby who predicted Kerry would win with 300+ of the electoral college votes in the days before the 2004 election? :-/

___________________
http://eschatonnow.blogspot.com/

15,000 people lost lives this month.

Just put in 15,000 people/graves at one place and try to pic how much that number is.

All for the fucking freedom formula of Bush, mission accompolished!

This wont be forgotten, Iraqies will not forget what happened to them.

And Magnetonium whatever you are saying has been said and done here couple of times before.. we have all gone through this again and again couple of times before.. just wait till someone comes in and says terrorist country Iran should be 'liberated' too; they got WMDs and no 'democracy'.

Edit: After posting just realised that the word 'Iraqies' just sound synonymous with 'terrorist' nowdays; how much this administration has misused this word.

___________________
___________________

Last edited by Purple on Oct-12-2006 at 16:34

quote:

Originally posted by Magnetonium Nobody answered me, but - is there really a democracy in Iraq right now?

Yea, they had a vote and everything...

___________________
"...End? No, the journey doesn't end here. Death is just another path...one that we all must take.
The grey rain-curtain of this world rolls back, and all change to silver glass...and then you see it...
...white shores...and beyond...the far green country under a swift sunrise."