Nonparametric estimation of the bivariate survival function for one modified form of doubly censored data (Q2430236)

Bivariate estimation of survival data is studied. Three nonparametric estimators of bivariate survivor functions are developed when both components are subject to left and right censoring. The first two are generalizations of \textit{D.M. Dabrowska} [J. Multivariate Anal. 29, No. 2, 308--325 (1989; Zbl 0667.62025); Ann. Stat. 16, No. 4, 1475--1489 (1988; Zbl 0653.62071)] and \textit{G. Campbell} and \textit{A. Foeldes} [Nonparametric statistical inference, Budapest 1980, Vol. I, Colloq. Math. Soc. Janos Bolyai 32, 103--121 (1982; Zbl 0516.62041)], and the third is an inverse-probability-weighted (IPW) estimator. The consistency of the proposed estimators is established. Simulation results show that all proposed estimators perform reasonably well in finite samples.