Weighted Semiparametric Estimation in Regression Analysis with Missing Covariate Data

摘要: Abstract This article investigates estimation of the regression coefficients in an assumed mean function when covariates on some subjects are missing. We examine performance a Horvitz and Thompson (1952)-type weighted estimator by using different estimates selection probabilities, which may be treated as nuisance parameters (or function). In particular, we investigate properties estimate probabilities estimated kernel smoothers. present large sample theory for new conduct simulation studies comparing proposed to maximum likelihood multiple imputation under various model assumptions missingness mechanisms. addition, provide two real examples that motivate this investigation.