An Alternative to Efron's Redistribution-of-Mass Construction of the Kaplan—Meier Estimator
Gregg E. Dinse
Abstract
Gregg E. Dinse
Abstract
Kaplan and Meier (1958) derived the nonparametric maximum likelihood estimator of the survival function for the case in which some survival times are right-censored. Efron (1967) proposed a redistribution-of-mass construction of the Kaplan—Meier estimator that emphasized and illustrated the contribution of the censored observations. This article presents an alternative construction that, unlike Efron's method, redistributes the mass initially associated with each censored observation directly to the uncensored observations. The proposed construction avoids distributing a given mass more than once and provides additional insight into the nature of the Kaplan—Meier estimator.
OpenAlex reports 13 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Kaplan and Meier (1958) derived the nonparametric maximum likelihood estimator of the survival function for the case in which some survival times are right-censored. Efron (1967) proposed a redistribution-of-mass construction of the Kaplan—Meier estimator that emphasized and illustrated the contribution of the censored observations. This article presents an alternative construction that, unlike Efron's method, redistributes the mass initially associated with each censored observation directly to the uncensored observations. The proposed construction avoids distributing a given mass more than once and provides additional insight into the nature of the Kaplan—Meier estimator.
Key concepts: Kaplan–Meier estimator, Estimator, Survival function, Survival analysis, Statistics, Mathematics, Econometrics, Nonparametric statistics