The authors have developed a method for estimating the distribution of an unobservable random variable from data that are subject to considerable measurement error and that arise from a mixture of two populations, one having a single-valued distribution and the other having a continuous unimodal distribution. The method requires that at least two positive intakes be recorded for a subset of the subjects in order to estimate the variance components for the measurement error model.
Iowa State University
Nusser, Sarah M.; Fuller, Wayne A.; and Guenther, Patricia M., "Estimating Usual Dietary Intake Distributions: Adjusting for Measurement Error and Nonnormality in 24-Hour Food Intake Data" (1995). CARD Staff Reports. 32.