INSTRUMENTAL VARIABLE ESTIMATION IN BINARY REGRESSION MEASUREMENT ERROR MODELS

We describe two approaches to instrumental variable estimation in bina ry regression measurement error models. The methods entail constructin g approximate mean models for the binary response as a function of the measured predictor, the instrument, and any covariates in the model. Estimates are obtained by exploiting relationships between regression parameters, just as in linear instrumental variable estimation. In the course of deriving the approximate mean models, we obtain an alternat ive characterization of instrumental variable estimation in linear mea surement error models.