CASE-CONTROL STUDIES WITH CONTAMINATED CONTROLS

This paper considers inference about a parametric binary choice model when the data consist of two distinct samples. The first is a random s ample from the people who made choice 1, say, with all relevant covari ates completely observed. The second is a random sample from the whole population with only the covariates observed. This is called a contam inated sampling scheme. An example might be where we have a random sam ple of female labor force participants and their covariate values and a second random sample of working age women, with covariates, whose pa rticipant status is unknown. We consider the cases in which the fracti on of the population making choice 1 is known and that in which it is not. For both cases we give semiparametrically efficient procedures fo r estimating the choice model parameters.