kind of thinking appeals to economists:
Betsy's Page evaluates the Type I error committed by a Marine in Fallujah.
That use of the statistical concept of Type I error
neatly encapsulates the whole question of this shooting. It depends on what the null hypothesis is. Is the null that people lying motionless on the battlefield are dead? Or is it that they are alive? Type I error is the probability of a false alarm. Type II errors -- the likelihood of accepting a false null hypothesis, can be thought of as a failed alarm. On a battlefield, failed alarms can be deadly. You could switch the null and assume the motionless are still alive, but since the power of the test
is pretty low, you would have to move very slowly through a battlefield. What is the correct balance? Obviously, opinions will differ, but there are standards used, as Wretchard