Exposure-response relationship for a dichotomized response when the continuous underlying variable is not measured.