(a) Both. "Neither" is the complement of "A or B", so P(A∪B)=1−0.2=0.8. The union is made of the mutually exclusive parts "exactly one" and "both", so
P(A∩B)=0.8−0.5=0.3
P(B) and B but not A. From P(A∪B)=P(A)+P(B)−P(A∩B):
0.8=0.45+P(B)−0.3⟹P(B)=0.65
P(B but not A)=P(B)−P(A∩B)=0.65−0.3=0.35
Check. P(A but not B)=0.45−0.3=0.15, and 0.15+0.35=0.5 is indeed the probability of exactly one.
(b) Range of P(A). The first two facts alone still give P(A∪B)=0.8 and P(A∩B)=0.3, as in (a). Since A∩B⊆A⊆A∪B,
0.3≤P(A)≤0.8
Both ends can occur. P(A)=0.3 when "A but not B" has probability 0 and all of the 0.5 for "exactly one" is "B but not A". P(A)=0.8 when all of that 0.5 is "A but not B". So the least value of P(A) is 0.3 and the greatest is 0.8.