Part (iii)
4 marks
Approach
and are mutually exclusive, so , and partition the sample space into three disjoint pieces. Splitting across this partition, . This turns the two given pieces of data — and — together with independence, into two equations relating to , the one remaining unknown. The extreme values of then come from the extreme values of allowed by the constraints of the problem (each probability lying in ).
We are given , and . Write and , both unknown for now, and let , so .
Since , the events , and partition :
Using independence, . Writing ,
Also, , and this equals :
Substituting into the equation for :
Therefore
a decreasing function of , so the maximum of occurs at the smallest possible , and the minimum at the largest possible .
We now bound . Since , . Since is a probability, , i.e. . (Checking the remaining probabilities — , , — against these two bounds shows none of them cuts the range further, and throughout, so and staying not mutually exclusive is automatic.)
At : .
At : .
| To check your answer: at , , and , a valid probability; at , and also. ✓ |
the maximum possible value of is , and the minimum possible value is .