Weighted semiparametric estimation in regression analysis with missing covariate data

A method for the estimation of the regression coefficients in an assumed mean function was developed by using different estimates for selection probabilities. The nonparametric smoothing of the selection probabilities was a useful tool for large sampling sizes with limited data on probability density and mean function. Furthermore, the proposed weighted semiparametric estimation in regression analysis with missing covariates was designed mainly for continuous data.