(February 9, 2014 at 1:35 pm)Jacob(smooth) Wrote: Lucia de berk for one. I have a whole talk on that.
1 in 2![]()
Bayeristic probability. Great fun.
Yeah, here's another one: http://en.wikipedia.org/wiki/Sally_Clark...l_evidence
Close, but the answer is actually 1/3 (you may have misread that the background rate was 0.5%, not 1%)
Best way to think about it without doing the maths is that if you tested 200 people, 1 of which is actually a drug taker, the most likely outcome would be:
2 false positives.
1 true positive.
Therefore 1/3 positives are true ones.