A semi-parametric empirical likelihood approach for conditional estimating equations under endogenous selection

The estimation and inference for conditional estimating equations models with endogenous selection, are considered. approach takes into account possible selection which may lead to a bias. It can be used wide range of statistical not covered by the model-based sampling theory. Endogeneity either part or within covariates. is particularly well suited unknown heteroscedasticity, uncontrolled confounders measurement errors. will necessary model relationship between covariates instrumental variables, offers major advantages over two-stage least-squares. proposed has advantage being based on fixed number constraints determined size parameter.