05898204

j

05898204 Wang, Dongliang Hutson, Alan D. Miecznikowski, Jeffrey C. L-moment estimation for parametric survival models given censored data. Stat. Methodol. 7, No. 6, 655-667 (2010). 2010 Elsevier, Amsterdam EN bootstrap order statistics

doi:10.1016/j.stamet.2010.07.002

Summary: An estimator for the $r$\,th L-moment, $\lambda _r$, given right censored data is proposed, of the form of $\lambda_r = \sum^n_{j=1} T_{(j)} u_{j(r)}$, where the $T(j)$'s are the ordered censored or failure times. We present the L-moments for several common survival distributions. Under certain regularity conditions, it is shown that $\hat {\lambda_r}$ converges in probability to $\lambda _r$, and has an asymptotic normal distribution. We further develop exact bootstrap estimators of the mean and variance of $\hat {\lambda_r}$. The procedure is illustrated by an application to head-and-neck cancer survival data. Monte Carlo simulation studies show that our L-moment estimator may be used to characterize distributions and provides an alternative approach to parameter estimation.