Data integration and other challenges: A Bayesian perspective |
|
| Coordinator 1 | Dr Veronica Ballerini (University of Florence) |
Nowadays, survey and official statistics are experiencing a new era, founded on the integration of different data sources and handling complex data. In this context, it is more and more important to enable researchers to integrate prior knowledge with observed data to derive more nuanced and informative conclusions, and to easily quantify uncertainty. In this framework, the Bayesian paradigm provides a versatile framework for handling these new challenges. The session will delve into the role of Bayesian inference and its adaptability to a wide range of survey methodologies, from sampling designs to estimation techniques. By emphasizing the incorporation of prior beliefs and uncertainty quantification, Bayesian approaches offer a holistic perspective on survey analysis, enhancing the robustness and interpretability of statistical findings. Furthermore, the session will showcase cutting-edge applications of Bayesian methods in survey research, including but not limited to small area estimation, missing data imputation, and complex survey designs. Case studies and applications will show how Bayesian techniques can address challenges unique to survey data, such as nonresponse bias, hierarchical structures, and model complexity.
This sessions is thought for hosting contribution from early career statisticians.