A0937
Title: Statistical data integration using a prediction approach
Authors: Estelle Medous - University of Toulouse 1 (France) [presenting]
Anne Ruiz-Gazen - Toulouse School of Economics (France)
Camelia Goga - Universite de Bourgogne (France)
Jean-Francois Beaumont - Statistics Canada (Canada)
Alain Dessertaine - La Poste (France)
Pauline Puech - La Poste (France)
Abstract: In a finite population setting, it is possible to improve the efficiency of estimators based on a probability sample by using non-probability big data sources. However, the target variable may not be observable in the big data sources, while the auxiliary information present in these sources may not be measured in the probability sample. In such a situation, new estimators can be proposed with a prediction approach. These estimators are either design-based, model-based, or cosmetic. Their properties in terms of bias and efficiency are studied using some theoretical and simulation results. The interest of the new proposals is illustrated in the context of the French postal service, where the objective is to estimate the monthly postal traffic through a survey of the mailmen rounds while taking advantage of the database containing information on the automatically processed postal mail.
