If a screening test is 90% accurate, and your result comes back positive, what are the chances it is a false positive?

10% i say

You left out how certain I am about the “urgent information”. Realistically, I’m 0% sure. Moreover, what makes someone an attacker or a terrorist or not, really? Mathematically, it’s not the worst-worded story problem I’ve seen, but it seems like you mean to say that you want us to base our answer on the given (but not well explained) fact that there is exactly one “attacker” in the group of 3,000.

Playing along, the bizarre test will say that on average (also not well defined) 10% of the 2,999 innocent people are attackers. The actual attacker will presumably also be detected by the test only 90% of the time. So we can expect about 300 false alarms and probably one correct detection. So for each detection, the chance it’s the right guy is about one in 300, or 0.3%. You gotta educate those security guards better. ;-)

@Zaku

Wins one internet!

Usually a percentage of false positives is not the same as the percentage of false negatives. So if someone just told me that a test was “90% accurate” I’d say the test is suspect. Also, 90% is a pretty round number. How was that arrived at? Hmmmm. Sounds like bullshit to me.