Do you remember the 2004 Lancet study that blamed the United States for at least a hundred thousand civilian deaths in Iraq? It sparked exhilarated feelings among anti-Americans and Saddam Hussein fans around the world. Finally, they had "scientific proof" that suggested that neo-conservative democratization was a deathly business. Hurrah!

Well, the scientific proof was based on some very unscientific methods. Not that the looniest left-wingers bothered about that, but it made the rest of us question the world's best-regarded medical journal. It simply doesn't matter how many Iraqi doors you knock on, you cannot make any scientific claims based only on estimates from people with nothing to go on except their guesses. That is a bit like asking the local hairdresser for her opinion on global warming and present what she says as meteorological evidence.

The cartoon comes from the Cox & Forkum blog. Read more about the death-toll survey on Michelle Malkin's blog.