Quote:
Originally Posted by Sombra
Any particular reason why it is nor credible? Any failure in the method used you could find? Sample number to low etc.?
Up to now I couldnt find any serious, scientific debunking of the statistical analysis done here or even an attempt of other organisations to do such an analysis. ( Iraq body count does only count the certified deaths reported in the media AFAIK and even they are reporting about 45000 civilian deaths through violence.
So how do you estimate the number of people who died because of: Violence, missing medicine, food , water etc. because of the instability caused by this war?
|
Sombra,
The general methodology is fine. Because of the low number of clusters, their CI is huge (i.e. the results are very imprecise), but it is still significant. So, this is a point of criticism, but not a fatal flaw. However, when you start benchmarking their results, this is where it begins to unravel. Their estimation of pre-war mortality rates is significantly lower than benchmark results that are out there. Furthermore, when you benchmark their first study, you find that it doesn't benchmark close to other results that are out there. Finally, I find their benchmarking attempts both misleading and completely lacking IMO.
Shek