Premium
Empirical Likelihood for Censored Linear Regression
Author(s) -
Qin Gengsheng,
Jing BingYi
Publication year - 2001
Publication title -
scandinavian journal of statistics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.359
H-Index - 65
eISSN - 1467-9469
pISSN - 0303-6898
DOI - 10.1111/1467-9469.00261
Subject(s) - mathematics , empirical likelihood , censoring (clinical trials) , statistics , likelihood principle , restricted maximum likelihood , linear regression , likelihood function , regression analysis , likelihood ratio test , empirical distribution function , econometrics , maximum likelihood , quasi maximum likelihood , confidence interval
In this paper we investigate the empirical likelihood method in a linear regression model when the observations are subject to random censoring. An empirical likelihood ratio for the slope parameter vector is defined and it is shown that its limiting distribution is a weighted sum of independent chi‐square distributions. This reduces to the empirical likelihood to the linear regression model first studied by Owen (1991) if there is no censoring present. Some simulation studies are presented to compare the empirical likelihood method with the normal approximation based method proposed in Lai et al. (1995). It was found that the empirical likelihood method performs much better than the normal approximation method.