Abstract
SUMMARY In this paper, we consider nonparametric procedures for assessing goodness of fit when the data may be subject to random censoring. We introduce two classes of large-sample tests, based, respectively, on the maximum and average weighted difference between the specified survival function and the Kaplan-Meier estimate of the true survival function. The procedures in the first class are inverted to obtain simultaneous confidence bands for the survival and cumulative hazard functions. These bands provide simultaneous bands associated with the percentage (P-P), quantile (Q-Q) and hazard plots for censored data. Some key word8: Hazard function; Nonparametric procedure; Probability plot; Shift function; Survival function.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
