A Note on Nonparametric Estimation of the Distribution Function from Interval-Censored and Truncated Observations

Abstract

SUMMARY This paper is concerned with the nonparametric estimation of the distribution function F from interval-censored and truncated data as discussed by Turnbull. It is observed that Turnbull's characterization of F̂, the maximum likelihood estimator of F, is not applicable in general when the data are interval censored and truncated. It is shown, however, that Turnbull's characterization can be easily modified to obtain the appropriate characterization of F̂.