ABSTRACT
Censoring occurs when analysts are unable to observe the response variable of interest. For example, in a clinical trial monitoring the remission of cancer patients, some patients may still be in remission at the trial's closure, so that for these patients, they know only that their true period of remission was longer than the duration of the trial. Finkelstein describes sacrifice experiments in experimental animals. Right-censoring occurred when some hostages had not been released by the time of the news report, or had died from other 'causes.' Of interest is the time from injection of a carcinogen until the onset of tumors. An attempt to measure this time is made by sacrificing the animal, and observing at this time of death if tumors are present – in which case the required onset time is smaller than the censored time actually observed and the observation is left-censored – if tumors are absent, time to tumor onset is right-censored.
