Seminar: "Inference with nonprobability samples and survey data integration"

AUEB STATISTICS SEMINAR SERIES 2025-2026

Presenter: Camilla Salvatore, Assistant Professor, Department of Methodology and Statistics, Faculty of Social and Behavioural Sciences, Utrecht University

Room: Troias Amphitheater

ABSTRACT

The growing availability of different data sources, including traditional probability-based surveys, online nonprobability surveys, and digital trace data, offers new opportunities for official statistics and academic research. By integrating multiple sources, researchers can address novel research questions and improve inference for small domains and hard-to-reach populations. However, successful data integration requires careful attention to the varying data quality dimensions of each source, as well as the biases that may arise when combining them. There is no single methodology that suits all cases, so researchers must tailor their approaches to the specific context and available data sources. In this work, we present several case studies that illustrate the use of different data types and integration strategies, highlighting both their potential and their methodological challenges.