Self-consistent estimators of survival functions with doubly-censored data

Autor: Qiqing Yu, Li Linxiong
Rok vydání: 1997
Předmět:
Zdroj: Communications in Statistics - Theory and Methods. 26:2609-2621
ISSN: 0361-0926
DOI: 10.1080/03610929708832068
Popis: The consistency and asymptotic normality of self-consistent estimators (SCE) of survival functions with doubly-censored data have been studied by many authors. However, to the best of our knowledge, expressions of the asymptotic variance of the SCE have not been derived in the literature. In this paper, under the assumption that the survival time and censoring time distributions are discrete with finitely many jump points, an expression and a consistent estimator of the asymptotic variance of the SCE are presented. A proof of the strong consistency of the SCE is also presented. Our simu¬lation studies indicate that the estimate of the asymptotic variance is very close to the true value even with moderate sample sizes and high censoring rates
Databáze: OpenAIRE