Simulation study under a Semi-parametric Model for censored gap times

Abstract

We consider a comparison between the Kaplan-Meier and the semiparametric estimators for a censorship models. The observations are assumed to be generated under a semi-parametric random censorship, this mean that a random censorship model where de conditional expectation of censoring indicator given the observations belongs to a parametric family. The semi-parametric estimator of the survival function was de ned in de U~na-Alvarez and Amorim (2011). An asymptotic representation of a general empirical integral as a sum of independent and identically distributed (i.i.d.) random variables under the proposed model was obtained in Amorim (2012). The performance of the corresponding asymptotic con dence intervals (a.c.i.) relative to that of a nonparametric method, de U~na-Alvarez and Meira-Machado (2008), is investigated through simulations Dikta et al. (2005).