As one expects, this Malcolm Gladwell piece on child molesters is well-written and interesting and informed by the literature.
What I think is missing is any notion that Type II errors have to be balanced against Type I errors. Gladwell is all about Type I errors - failing to conclude that individuals are child molesters when they actually are - and not at all about Type II errors - falsely concluding that individuals are child molesters when in fact they are not. Both types of error are very costly in this context.
It's really quite easy.
1 year ago