LA Times - "Poll: Civilian death toll in Iraq may top 1 million"

[SAJ]

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

Link;http://www.latimes.com/news/nationworld/world/la-fg-iraq14sep14,1,3979621.story?coll=la-headlines-world&ctrack=1&cset=true
Link to ORB'S site http://www.opinion.co.uk/our-clients.aspx

 

[CptStern]

strange, the latimes article asks you to register but this link seems to work

http://www.latimes.com/news/printed...ory?coll=la-news-a_section&ctrack=1&cset=true


here's another source

http://www.knoxnews.com/news/2007/sep/14/survey-puts-iraqi-toll-at-12-million/

pretty horrific with no end in sight

 

[The Monkey]

The numbers seem pretty inflated, but regardless of the specific number, it's still far too much.

 

[SAJ]

The numbers seem pretty inflated, but regardless of the specific number, it's still far too much.

Inflated how?
Is that a criticism of the methodology, the organisation , your preconceptions or "common sense"?

Elaborate please.

 

[Ennui]

The household survey studies are the most methodologically secure, reliable estimates of civilian death toll; they also usually give the most appalling results.

 

[Gunner]

Bill O'Reilly will probably say that the poll was commissioned by Iran.

 

[No Limit]

Why would our government lie when they say the number is under 100K? :dozey:

 

[Mr Stabby]

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.

How does a survey give a death toll?

 

[Raziaar]

While the death toll is probably high, I find a million to be a great exaggeration. And based on a poll of only 1461 people? That's not really accurate I would think in determining something such as widespread deaths in the country. I mean for crying out loud, there's only 27 million people in the country.

 

[No Limit]

How does a survey give a death toll?

Through simple statistics methods. I always sucked at math so don't ask me about the details but its pretty simple, its the same way that a poll of 1000 americans can predict how the public will vote in a presidential election. Throughout the last couple of decades these polls have not been off by more than 5% I think.

 

[CptStern]

While the death toll is probably high, I find a million to be a great exaggeration. And based on a poll of only 1461 people? That's not really accurate I would think in determining something such as widespread deaths in the country. I mean for crying out loud, there's only 27 million people in the country.

well the lancet study from last year put the death toll at 655,000 ..up from 100,000 2 years earlier ..so this seems right ..oh and the margin of error is 2.4%

 

[Raziaar]

Through simple statistics methods. I always sucked at math so don't ask me about the details but its pretty simple, its the same way that a poll of 1000 americans can predict how the public will vote in a presidential election. Throughout the last couple of decades these polls have not been off by more than 5% I think.

I don't think close to 1500 people is nearly reliable enough to determine something as unpredictable as 1 in 27 people in iraq having been killed. 1 in 27!

 

[CptStern]

I don't think close to 1500 people is nearly reliable enough to determine something as unpredictable as 1 in 27 people in iraq having been killed. 1 in 27!

why not take it up with them? ..here are their figures:

http://www.opinion.co.uk/Documents/TABLES.pdf

here they explain their methodology:

http://www.opinion.co.uk/Newsroom_details.aspx?NewsId=78

 

[Raziaar]

why not take it up with them? ..here are their figures:

Well, because I'm not particularly interested in disproving them. It doesn't fit my mo.

 

[Atomic_Piggy]

Too many have died in a ludicrous and pointless war.

 

[CptStern]

Well, because I'm not particularly interested in disproving them. It doesn't fit my mo.

lol, your "mo"? .."method of operations"? "modus operandi" ..heh just find it funny for some odd reason

 

[No Limit]

I don't think close to 1500 people is nearly reliable enough to determine something as unpredictable as 1 in 27 people in iraq having been killed. 1 in 27!

But this is a scientific method, this isn't something that they just pull out of their ass. We use the exact same system to figure out how 60 million people will vote in our elections and to my knowledge its never been off by more than 5%. So even if you want to be a skeptic and give a margin of error of 10% those numbers are still huge, way more than what our government is telling us.

 

[Mr Stabby]

Did they ask 1500 people how many people they thought had died, I still don't get how polling 1500 people's opinion show a death toll

 

[Raziaar]

lol, your "mo"? .."method of operations"? "modus operandi" ..heh just find it funny for some odd reason

Modus Operandi.

I'm not interested in disputing the numbers because I no longer support this war.

 

[No Limit]

Did they ask 1500 people how many people they thought had died, I still don't get how polling 1500 people's opinion show a death toll

No, they asked a household how many people died in that particluar household as a result of the war. They collected a fairly large sample of households, 1400 in this case, and once they had the stats on that they were able to apply that trend across the board and get the numbers that they did.

 

[HunterSeeker]

No, they asked a household how many people died in that particluar household as a result of the war. They collected a fairly large sample of households, 1400 in this case, and once they had the stats on that they were able to apply that trend across the board and get the numbers that they did.

The part that is most unreliable is the definition of the "becouse of the war" of those that they asked. I still think this is reliable enough to be within the error margian, although at the lower end.

 

[CptStern]

No, they asked a household how many people died in that particluar household as a result of the war. They collected a fairly large sample of households, 1400 in this case, and once they had the stats on that they were able to apply that trend across the board and get the numbers that they did.

I'd also like to add that they sampled 17 of 19 regions in iraq ..the other 2 their iraqi handlers couldnt/wouldnt get access to

 

[No Limit]

I'd also like to add that they sampled 17 of 19 regions in iraq ..the other 2 their iraqi handlers couldnt/wouldnt get access to

Do you know if they did the Kurdish north also? Can't read through the entire thing right now, swamped at work.

 

[CptStern]

I cant remember where I read that ..could be in the aforementioned links ..dont have time to sift through it all

 

[venturon]

I'd also like to add that they sampled 17 of 19 regions in iraq ..the other 2 their iraqi handlers couldnt/wouldnt get access to

Yeah I found it in the ORB explanation:

•The methodology uses multi-stage random probability sampling and covers fifteen of the eighteen governorates within Iraq. For security reasons Karbala and Al Anbar were not included. Irbil was excluded as the authorities refused our field team a permit.

It's good that that was cleared up, I was a little dubious until I saw the spread of the survey. That's pretty horrible.

 

[Solaris]

The sample size is more than enough.

 

[K e r b e r o s]

The household survey studies are the most methodologically secure, reliable estimates of civilian death toll; they also usually give the most appalling results.

I disagree with you're endorsing it. The flaw with a house hold survey is that people can lie -- that alone, if its powerful enough, can trick people.

I still don't get how polling 1500 people's opinion show a death toll

Exactly. :thumbs:

 

[Solaris]

Why would your average person lie

Ring ring
Average Iraq: Hello
Man:Hello, i'm from ... has anyone in your familly died recently?
AI: Yes my brother, *hangs up

AI: LOLOLOLOLOL

 

[K e r b e r o s]

Why would your average person lie

Simple -- to benefit they're cause. Bush lied -- Bin Laden lied. People lied.

Or, do you believe everyone on this Earth is just morally and ethically sound?