A review of inferential methods for the Kaplan-Meier estimator

Stuart Barber & Christopher Jennison.
We survey existing methods of inference using the Kaplan-Meier estimator, focusing on testing Ho:S(t*) = p* for some chosen t* and p*. Performance of these methods is assessed by simulation. The problems these methods encounter are analysed and the reasons for these problems discussed.

Some key words:
Beta distribution, bootstrap sampling, censoring, confidence interval, Greenwood's formula, Kaplan-Meier estimate, likelihood ratio test, survival data.

Back to publications list