Empirical likelihood regression analysis for right censored data

Linear models are useful alternatives to the Cox (1972) proportional hazards model for analyzing censored regression data. This article develops empirical likelihood methods for linear regression analysis of right censored data. An adjusted empirical likelihood is constructed for the vector of regression coefficients using a synthetic data approach. The adjusted empirical likelihood is shown to have a central chi-squared limiting distribution, which enables one to make inference using standard chi-square tables. We also derive an adjusted empirical likelihood method for linear combinations of the regression coefficients. In addition, we discuss how to incorporate auxiliary information. A small simulation study is carried out to highlight the performance of the adjusted empirical likelihood methods compared with the traditional normal approximation method. It shows that the empirical likelihood confidence intervals tend to have more accurate coverage probabilities than the normal theroy intervals. An illustration is given using the Stanford Heart Transplant data.