Proposal Description

In recent years, the increasing availability of data from multiple sources has stimulated growing interest in data integration methodologies. In particular, nonprobability samples have emerged as an important source of information owing to their lower cost and the speed and ease with which they can be collected, compared with traditional data sources such as sample surveys.
However, the arbitrary selection mechanisms underlying nonprobability sampling prevent the direct application of standard design-based inference based on probability randomization theory. This has sparked considerable debate in both theoretical and applied literature, particularly when statistics for public decision making have to be produced.
The purpose of this session is to provide a forum for discussing recent methodological advances and future perspectives in data integration, with special emphasis on the integration of probability and nonprobability samples.
This invited paper session is proposed by the ISI Special Interest Group on Data Science.
Pier Luigi Conti and Daniela Marella, Sapienza Università di Roma, Italy, Title “Bridging probability and nonprobability samples: inference and data integration under nonstandard conditions"
Ralf Münnich, University of Trier, Germany Title “Statistical digital twins using multiple and diverse data sources”
Danny Pfeffermann, University of Southampton, United Kingdom