Jousting with the Lancet: Pajamas Media Interviews Professor Gilbert Burnham
Tom W:…ignoring those residences produces a bias
It would if the density of residences (normalized to average) times the density of deaths (normalized) is bigger than one. A main drag would more likely have stores, market places, police stations, targets. But as you say, people are not going to set off IED’s next to their own houses. If I was running the insurgency, I would look for operatives who live in the neighborhood, who know the place and can case the targets, and can slink off back home, a couple of blocks away.
It could be miles away.
True, but if they conveniently destroyed that information, there’s no way of checking it. A residential street that’s miles long? Doesn’t sound typical to me.
Bottom line: You just have to see the actual data. There’s plenty of reason for skepticism.
