Correcting for misclassification error in gross flows using double sampling: moment-based inference vs. likelihood-based inference

Gross flows are discrete longitudinal data that are defined as transition counts, between a finite number of states, from one point in time to another. We discuss the analysis of gross flows in the presence of misclassification error via double sampling methods. Traditionally, adjusted for misclassification error estimates are obtained using a moment-based estimator. We propose a likelihood-based approach that works by simultaneously modeling the true transition process and the misclassification error process within the context of a missing data problem. Monte-Carlo simulation results indicate that the maximumlikelihood estimator is more efficient than the moment-based estimator.