A breathalyser is used to check whether a driver’s blood alcohol level is over the legal limit. We define these events:
$A:$ breathalyser shows driver is over the limit;
$B:$ the driver’s blood alcohol really is over the limit.
(a) Suppose $P(A|B) = 0.9$, $P(\bar A|\bar B) = 0.99.$ A driver is stopped at random and breathalysed. If a conviction is made on the basis of the breathalyser alone, determine the error rate (rate of wrongful convictions) $P(\bar B|A) = P$(driver is under the limit given that the breathalyser shows driver over the limit).
Got stuck on this one, any help/tips would be much appreciated. Thanks!
This problem has more than one solution.
Let us denote:
The events mentioned above are disjoint and covering.
We have the following equalities:
If $p$ and $q$ are nonnegative and satisfy $10p+100q=1$ then a solution is:
leading to: $$P(B^{\complement}\mid A)=\frac{q}{q+9p}$$