SCIE1001 Module 3 Quiz Practice Questions (Sem 2 2023)
Question 1:
Imagine that your city of exactly 1 million inhabitants has elected a mayor who
is pro-surveillance. Keen to reduce the crime rate of the city, he announces the
installation of surveillance cameras with automatic facial recognition software. If
a scanned individual is a known terrorist, the scan will detect them. However, it
has a 1% false positive rate.
The mayor decides to put the facial recognition to test and scans the entire
population of the city. Suppose there are 100 true terrorists in the city. How
many individuals will do we expect will be identified by the software?
a) 10,000
b) 1,000
c) 9,900
d) 900
e) 10,099
Question 2:
The morning after an epic party in a large office, a man was found dead with
stab wounds on his back. His arch nemesis at work was suspected of the murder,
and fingerprint analysis was positive. Using just the information from the
fingerprint analysis, if there were a total of 400 guests at this party and the
chances of a false positive is 1 in 100, what are the chances that this man is
guilty? Assume there is one murderer and that their fingerprints were left on the
knife.
a) 0.2
b) 0.01
c) 1.0
d) 0.3
e) 0.5
f) 0
Question 3:
Which of the following may increase the likelihood of false positive results?
Select one or more answers.
a) Random Sampling
b) Novelty Seeking
c) Measures of Uncertainty
d) Testing Multiple Hypotheses
Question 4:
Can we conclude from the following diagram that it is safer to drive while under
the influence of alcohol? Select one or more answers.
a) Yes, because there are fewer people injured while driving under the
influence.
b) No. Only a small percentage of drivers will drive under the influence and
so the data is not a reflection of the true proportion of accidents.
c) Yes. Assuming that all injuries are recorded in this figure, driving under the
influence gives us between 20% to 25% chance of being injured.
d) No. This reflects only the proportion of people injured while driving under
the influence, or not, and has no information about how safe these drivers
are.
e) No. Drivers under the influence are likely dangerous drivers and therefore
the data is reported in a misleading way.