Instability of Inverse Probability Weighting Methods and A Remedy for Non-ignorable Missing Data
报告人：Jing Qin（National Institute of Allergy and Infectious Diseases）
地点：Zoom Meeting（838 4123 7567）
Inverse probability weighting (IPW) methods are commonly used to handle non-ignorable missing data problems under a logistic model assumption for the missingness probability. However, solving IPW equations numerically may involve non-convergence problems when the sample size is moderate and the missingness probability is high. Moreover, those equations often have multiple roots; identifying the best one is challenging. Therefore, IPW methods may have low efficiency or even produce biased results. We identify the pitfall in these methods pathologically: they involve the estimation of a moment generating function, which are notoriously unstable in general. As a remedy, we model the outcome distribution given the covariates of the completely observed individuals semiparametrically. After forming an induced logistic regression model for the missingness status of the outcome and covariate, we develop a maximum conditional likelihood method to estimate the underlying parameters. The proposed method circumvents the estimation of a moment generating function and hence overcomes the instability of IPW methods. Our theoretical and simulation results show that the proposed method outperforms existing competitors greatly. We conclude that one has to be cautious to use the existing statistical methods in the non-ignorable missing data inference if only a parametric logistic propensity score model is used but the underlying outcome regression model is left arbitrary.
Jing Qin is a mathematical statistician at Biostatistics Research Branch, National Institute of Allergy and Infectious Diseases, NIH. He obtained MS degree from East China Normal University in 1986 and PhD of Statistics from the University of Waterloo in 1992. He was elected as an American Statistical Society Fellow in 2006. He used to be the joint editor of International Journal of Biostatistics. He has made influential contributions to biased sampling and missing data, empirical likelihood, causal inference, genetic mixture models and infectious diseases.
Meeting ID：838 4123 7567