The central limit theorem under random censorship (Q1896262): Difference between revisions

The author's objective is to extend the central limit theorem (CLT) for a sample \(X_1, \dots, X_n\), under a general transformation \(\varphi\), to the random censorship model. If \(\widehat{F}_n\) denotes the Kaplan- Meier estimate of the underlying distribution function \(F\), the main result provides a representation of integrals \(\int \varphi dF_n\) as a sum of i.i.d. random variables plus a remainder term of order \(o_P (n^{-1/2})\). The CLT for the above Kaplan-Meier integrals is an immediate consequence of this representation. The results are proved under minimal moment assumptions reducing to the existence of a second moment in the non- censored case.