600,000 killed by Bush's war in Iraq

[Tingkai]

Originally posted by Straybow
Sorry, you can't extrapolate a survey of a couple thousand families into nationwide figures.

Uhmn, actually, statistically you can. A sample size of 18,000 families is huge, statistically speaking.

[Chemical Ollie]

Hmmm, what claims have the highest credibility: a study published in a peer-reviewed, non-political scientific journal, or some words by a politician who is well-known for lying, and who rely on body-count figures from the Pentagon, an organization that "doesn't do body-counts"?

[Straybow]

Originally posted by Tingkai
Uhmn, actually, statistically you can. A sample size of 18,000 families is huge, statistically speaking.

1,849 Iraqi families

[Victor Galis]

Originally posted by Straybow

Originally posted by Tingkai
Uhmn, actually, statistically you can. A sample size of 18,000 families is huge, statistically speaking.

1,849 Iraqi families

That's still bigger than the average survey. Margin of error should be 3% or so unless they really left out specific types of people.

[Dinner]

Originally posted by Chemical Ollie
Hmmm, what claims have the highest credibility: a study published in a peer-reviewed, non-political scientific journal, or some words by a politician who is well-known for lying, and who rely on body-count figures from the Pentagon, an organization that "doesn't do body-counts"?

Given the extremely small sample size of just 1,800 families most of which are located in just the major urban areas... Neither.

[chequita guevara]

Originally posted by Kuciwalker
It's impossible to determine which deaths are attributable to US action.

No, the study attributes 1/3rd of those deaths to direct American actions.

[chequita guevara]

Originally posted by Oerdin
Given the extremely small sample size of just 1,800 families most of which are located in just the major urban areas... Neither.

It's not that small for a country that size. Statisticians are saying this is a well done study.

[Ramo]

The most interesting thing about the study is that the mortality rate pre-invasion was .55%/year, increasing to 1.32%/year. There are huge error bars (as one would expect), but even a low estimate would imply a nearly doubling of the mortality rate. While all sorts of things may have happened in an alternate universe where the invasion did not occur, it's hard to imagine a nearly doubling to trebling of the mortality rate.

[chequita guevara]

The statitistical sampling method used is the same method that private marketting research companies use here in the United States. It is extremely accurate. Creepily so.

[Kuciwalker]

Originally posted by chegitz guevara
No, the study attributes 1/3rd of those deaths to direct American actions.

Reread my post.

[chequita guevara]

The way you wrote it, you are saying you cannot determine which deaths were caused by "U.S. action" which means the ones where we are directly responsible. Your intent may have been to inculde indirect causes, but that would require changing "action" from a verb to a noun.

The people who did the study are saying that the U.S. is responsible, directly or indirectly, for all of those 655,000 deaths, because we created the conditions in which tey occured. Crime is included because Hussein pretty much had crime under control, which is one of the major factors. Also, the total collapse of the health care system and other forms of infrastructure have lead to a much higher mortality rate than previosuly existed. Like, if I cut your telephone line, and you have a heart attack and are unable to phone for an ambulance, your death is my fault, because you couldn't get help. You might have died anyway, but I took the other possibilities away from you.

[Tingkai]

Ok slightly wrong.

The study surveyed 1,849 households containing 12,801 people in 47 clusters.

My point remains valid that the study surveyed a very large sample size.

[Kuciwalker]

Originally posted by chegitz guevara
The way you wrote it, you are saying you cannot determine which deaths were caused by "U.S. action" which means the ones where we are directly responsible. Your intent may have been to inculde indirect causes, but that would require changing "action" from a verb to a noun.

I'm not going to get into a semantic argument with you. My point was simply that there are many, many deaths for which you can't necessarily hold us responsible - or absolve us. These range from the neutral Iraqi civilian killed because there were insurgents in his home to a sick child who doesn't have access to adequate medical care to the man killed in a dark alley for his wallet because there weren't enough police on the streets.

[Chemical Ollie]

Originally posted by Oerdin


Given the extremely small sample size of just 1,800 families most of which are located in just the major urban areas... Neither.

I haven't read the actual report, but according to what I remember from the radio news where I heard it first, the special thing about this study was that it was done on a mixture of urban and rural families, whereas all other body-count studies have been done in urban morgues, where only major media go. Or do I recall wrong?

[Straybow]

Unfortunately, reliability of sampling also depends on the quality of the samples. If the samples are accustomed to exaggeration or to concealing information then the data is suspect. Here in America we get exit polls that differ from the actual voting results by enough to know that poll respondants will lie.

If even a small percentage of the samples return exaggerated results it can throw the whole off when we're talking about extrapolation of three orders of magnitude. And we know that a significant percentage of the Iraqi population have their own sectarian or partisan or anti-US/Western agendas.

Now the study purports that every US soldier over there is directly responsible for two deaths? I've known too many who spent their 18-24 mo and never fired a weapon except at a shooting range.