3 ms·
1) Those tests results would be highly correlated, so you wouldn't necessarily be much more than 99% accurate still. 2) If the prior probability of someone com
by hidenotslide 8y ago
1) Those tests results would be highly correlated, so you wouldn't necessarily be much more than 99% accurate still.
2) If the prior probability of someone committing a crime on a given night is low, you are in the classical example of false positives doing harm via Bayes rule.