A0163
Title: Temporal M-quantile models and robust bias-corrected small area predictors
Authors: Maria Bugallo - Miguel Hernandez University of Elche (Spain) [presenting]
Domingo Morales - University Miguel Hernandez of Elche (Spain)
Nicola Salvati - University of Pisa (Italy)
Francesco Schirripa - University of Pisa (Italy)
Abstract: In small area estimation, it is a smart strategy to rely on data measured over time. In this regard, linear mixed models are unable to properly capture time dependencies when the number of lags is large. Since there are no published studies addressing robust prediction in small areas based on time-dependent data, the M-quantile models are sought to be extended in this field of research. Indeed, the proposed methodology successfully addresses this challenge and offers flexibility to the widely imposed assumption of unit-level independence. Under the new model, robust bias-corrected predictors of small area linear indicators are derived. In addition, the optimal selection of the robustness parameter for bias correction is a theoretical contribution of the current research, exploring its applicability in outlier detection. As for the estimation of the mean squared error, a first-order approximation has been obtained under general conditions, and analytical estimators have been proposed. Several simulation experiments are carried out to investigate the performance of the new predictors and ensuing MSE estimators, as well as the optimal selection of the robustness parameters. Finally, an application to the Spanish Living Conditions Survey data is included to illustrate the usefulness of what has been done.
