Adaptive sampling design for the Italian social sample surveys: an application on the population census

• Published in: Metron (Rome) Vol. 82; no. 1; pp. 19 - 35
• Main Authors: De Vitiis, Claudia, Falorsi, Stefano, Guandalini, Alessio, Inglese, Francesca, Righi, Paolo, Terribili, Marco D.
• Format: Journal Article
• Language: English
• Published: Milan Springer Milan 01.04.2024; Springer Nature B.V
• Subjects: Adaptive sampling; Bias; Context; Data collection; Households; Mathematics and Statistics; Population (statistical); Sampling designs; Statistical Theory and Methods; Statistics; Response propensity; Population subgroups; Responsive-adaptive designs; Optimization
• ISSN: 0026-1424; 2281-695X
• DOI: 10.1007/s40300-023-00262-3

The aim of the paper is to show how a new survey design based on Responsive-Adaptive Survey Designs (RADs) can be adapted to the context of large-scale social surveys carried out by the Italian National Institute of Statistics (Istat) and particularly the Italian Population Census (PC). The Italian PC is the largest and most expensive social sample survey carried out yearly by Istat. In order to reduce the overall cost of the survey, data collection already adopts a sequential mixed mode design (CAWI/CAPI). The adoption of Adaptive Surveys Design (ASD) scheme can represent the solution for a further reduction of the survey costs, by minimizing the expected number of CAPI interviews and optimizing them in order to compensate for the under-coverage of some types of households and individuals that can affect the bias and the variability of the estimates. In the paper, the ASD is developed and applied for planning the design of the Italian PC. Furthermore, in a simulation context based on real data, the performances of the proposed design are compared with non-adaptive designs for assessing the cost reduction and the trade-off between bias and variance in estimating some target parameters.