distribution of the largest observed uncensored and the largestobserved survival times

We calculate finite sample and asymptotic distributions for the largestcensored and uncensored survival times, from a sample of survival data generated accordingto an iid censoring model. a key structural result obtained is that, conditionalon the value of the largest uncensored survival time, and knowing the number of censoredobservations exceeding this time, the sample partitions into two independentsubsequences, each subsequence having the distribution of an iid sample of censoredsurvival times, of reduced size, from truncated random variables. we apply extremevalue methods to deriving asymptotic distributions for some statistics.