Nonparametric Estimation of Cumulative Incidence Functions for Competing Risks Data with Missing Cause of Failure
Discussion Papers on Business and Economics, University of Southern Denmark, 21/2013
26 Pages Posted: 21 Dec 2013
Date Written: December 19, 2013
In this paper, we develop a fully nonparametric approach for the estimation of the cumulative incidence function with Missing At Random right-censored competing risks data. We obtain results on the pointwise asymptotic normality as well as the uniform convergence rate of the proposed nonparametric estimator. A simulation study that serves two purposes is provided. First, it illustrates in details how to implement our proposed nonparametric estimator. Secondly, it facilitates a comparison of the nonparametric estimator to a parametric counterpart based on the estimator of Lu and Liang (2008). The simulation results are generally very encouraging.
Keywords: cumulative incidence function, inverse probability weighting, kernel estimation, local linear estimation, martingale central limit theorem
JEL Classification: C14, C41
Suggested Citation: Suggested Citation