Estimating Kendall's τ when both times are subject to interval censoring

Document Type : Original Scientific Paper

Authors

Department of Statistics‎, Sistan and Baluchestan University, Zahedan, Iran

Abstract

A common challenge in working with longitudinal data is dealing with incomplete data‎. ‎According to the existing studies on the dependence structure of survival times‎, ‎it is a riveting topic for researchers to estimate survival functions and dependence parameters‎, ‎especially in biology and medical research‎. ‎Some researchers have studied the aforementioned subjects with left‎- ‎or right-truncated or censored data‎. ‎When the data involves interval censoring‎, ‎the mentioned issues still need to be solved or modified‎. ‎In this article‎, ‎we propose two alternative approaches to the estimation of a dependence parameter and Kendall's $\tau$‎, ‎given an interesting covariate and interval-censored dataset‎. ‎More precisely‎, ‎these approaches include non-parametric and semi-parametric methods to estimate the copula dependence parameter and Kendall's τ‎, ‎which are evaluated by simulation‎. ‎Finally‎, ‎we apply the mentioned approaches to a real-world dataset and copula's goodness-of-fit tests.

Keywords

Main Subjects

Journal of Statistical Modelling: Theory and Applications
Volume 4, Issue 2
July 2023
Pages 97-110

Files

History

  • Receive Date: 20 February 2024
  • Revise Date: 11 June 2024
  • Accept Date: 16 August 2024

Share

How to cite

Statistics