I couldn't find a thread discussing
this recent paper, and was interested in other opinions.
Specifically opinions that don't suck. Such as the opinion of the American right, where
A) It's not true because it's a large number
B) It's not true because you can't extrapolate a survey of a relatively small population to a large population
etc.
or the American left where
A) It's a large number and therefore true
B) It also contained a pie chart
or the American media
A) It doesn't matter because Iraqis aren't people anyway
My opinion, which you may freely claim to suck, is that the data set used seems to be worthwhile and in order, the general logical premise seems to be fine, and the writing and paper seem to be pretty shoddy.
"654,965 iraqis killed by the conflict, 601,027 by violence." Really? Is there any possibility of error in that calculation that you would like to mention before page 12? No? Can't wait to see their election survey where Bush's performance is approved of by 39.8445% of americans. Or possibly, by 42,355,812 registered voters.
Care to comment on the difficulty of getting a proper random sample set in a country with a large tribal population? No?
Care to comment on the fact that your paper finds an order of magnitude more deaths than other reports on the same topic? No? K. Care to at least acknowledge the existence of said? No? K.
But an obvious mistake that would lead them to being off by the order of magnitude 'generally assumed' isn't clear to me. Thoughts?