Comment by shirkey

Let's put this another way: if the data suggest a false positive rate of 90%, does that constitute an effective test?