Jousting with the Lancet: Pajamas Media Interviews Professor Gilbert Burnham
Dale said:
Go to any American City, near a major US Army or US Marine Corps Base, and ask 100 individuals if they “lost” anyone in the Iraq war!
Probaby come up with 5 families, per City/Military Base, that have had some dear one killed over there in the War, out of the approximately 3000 war dead we’ve had there so far!
Wow, 5, out of 100, equals 5%; since the US just reached 300 Million last week, as was all over the news, 5% of 300 MILLION is FIFTEEN MILLION AMERICAN WAR DEAD!
THAT is what Lancet did in Iraq!
The claim is presumably based on the so-called ‘main street bias’ claim:
The study suffers from “main street bias” by only surveying houses that are located on cross streets next to main roads or on the main road itself. However many Iraqi households do not satisfy this strict criterion and had no chance of being surveyed.
Unfortunately this is not an accurate description of the methodology. Burnham et. al. (the 2006 Lancet article) described describe the household selection thusly:
The third stage consisted of random selection of a main street within the administrative unit from a list of all main streets. A residential street was then randomly selected from a list of residential streets crossing the main street. On the residential street, houses were numbered and a start household was randomly selected.
In other words main streets were EXCLUDED leading to “residential street bias”.
All when the streets define a regular grid then EVERY residential street will cut a main street and choosing a main street at random merely fixes the orientation of the residential street (e-w or n-s for example). The houses are then chosen randomly over the length of the residential stree…so there will be no ‘main street bias’. There is however ‘residential street bias’ which would presumably lead to an UNDERESTIMATE.
When the streets do not define a regular grid then there will some bias. Exactly how much and if it is sufficient to overcome ‘residential street bias’ is unclear as it depends on the details of the street layout and the violence differential between main streets and residential streets.
