Daniel Davies has an excellent roundup of the Lancet discussion.
I've added an update to my post about Gerard Alexander's attack on the Lancet.
Chris at Mixing Memory takes down another Lancet critique, this one by John Ray.
More like this
This American Life have a show on Lancet 1 and Lancet 2. Much of it is a repeat of their story on Lancet 1, which is well worth listening to if you haven't already. Their comments on Lancet 2 are in the last ten minutes.
I think it is worthwhile to update James Wimberly's comparison of surveys of deaths in Iraq. In the table below death tolls have been extrapolated to give a number of deaths due to the war up to Oct 08.
The results of the Iraq Living Conditions Survey 2004 have been published.
Have you looked at Mike Dunford's criticism on Usenet? What do you think?
Brad DeLong has mirrored DD at his site, and
added a useful comment
regarding the likelihood of overestimating the incidence of a rare event with cluster sampling.
DD added two thoughtful responses in the comments.
It's worth a read.
I'm not impressed by Dunford. Much of it is just flatly wrong, like: "The results section of the paper contained NO data at all regarding when during the conflict deaths occurred." See figure 2 of the paper.
The only reasonable point he makes is that because people might have been displaced to safer areas, more dangerous areas might have been oversampled.