rl'he data for this problem come from a smoking cession study. Smokers were given hypnosis therapy to help them quit smoking. We call the subjects every morning to determine if they have had a cigarette. For n. subjects, they answer the phone every day until they report a relapse. Let y 1.. . . . , y\" be the relapse times for these subjects. For 771. subjects. they stop answering the phone and we never know their relapse time. For these censored subjects. let yn+1. . . . ,yn+m be the number of days until they stopped answering the phone. We model the times until relapse as exponential with mean A, ptif far}: and GUI" 1"[y]. The likelihood for subject 1". is flail *5 0. (c) In terms of n and the true value of A, when is A2 preferred over