Censoring and Truncation Mechanisms

Censoring and truncation are the special types of characteristics of time to event data. A censored observation arises when the value of the random variable of interest is not known exactly, that is, only partial information about the value is known. In the case of truncation, some of the subjects may be dropped from the study due to the implementation of some conditions such that their presence or existence cannot be known. In other words, the truncated subjects are subjects to screening by some conditions as an integral part of the study. This chapter presents the maximum likelihood estimation method for analyzing the censored and truncated data.