Estimation and Test of Conditional Kendall's Tau under Double-Truncation Data and Bivariate Left-Truncation Data

• Main Authors: Zih-Jyun Li, 李姿君
• Other Authors: Jin-Jian Hsieh
• Format: Others
• Language: en_US
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/43797582535502813953

碩士 === 國立中正大學 === 數理統計研究所 === 99 === In this paper, we extend the paper by Hsieh and Huang (2010), which considered semi-competing risks data and left-truncation data. We focus on estimation and test of the coordinatewise conditional Kendall's tau under double-truncation data and bivariate left-truncation data. We apply the Inverse Probability Censoring Weighted (IPCW) technique to construct an estimator of coordinatewise conditional Kendall's tau, Tc, and provide the Wald's type test statistic to test H0 : Tc = T0, where the elements of T0 are between (-1, 1). Furthermore, we provide the large sample properties for our suggested estimator and examine our estimator and test statistic via simulation studies. When X and T are quasi-independent, it implies Tc = 0. Thus, H0 : Tc = 0 is a proxy for H0. H0 : X and T are quasi-independent. We compare our test statistic with Martin's test statistic for quasi-independence test via simulations.