Further details on "On semiparametric efficient inference for two-stage outcome-dependent-sampling with a continuous outcome"

Rui Song, University of North Carolina at Chapel Hill
Haibo Zhou, University of North Carolina at Chapel Hill
Michael R. Kosorok, University of North Carolina-Chapel Hill


Two-stage Outcome dependent sampling (ODS) designs have been showed a cost effective way to enhance study efficiency in epidemiology and econometrics studies (Weaver and Zhou, 2005). In this paper, we develop a maximum likelihood estimation for continuous outcome regression models under this two-stage ODS setting. The profile likelihood function is used to give a consistent estimator for the asymptotic variance of the regression coefficients. Simulation studies show that our estimator outperforms existing methods.

Keywords: Maximum likelihood estimation; Missing data; Biased sampling; Outcome dependent; Two-stage; Empirical processes; Profile likelihood.