In this Links Lecture, sponsored by the American Statistical Association, Professor of Statistical Science and DUPRI Scholar Jerry Reiter will discuss integration of data from multiple sources, often known as data blending, which is a core activity for many statistical agencies. According to Reiter: "It promises to become increasingly prevalent in the future. However, data blending creates challenges for protecting the privacy and confidentiality of data subjects' identities and sensitive attributes, especially when the agency responsible for the blending seeks to disseminate public-use data products. In this talk, I describe some of the challenges to protecting confidentiality with blended data products. Along the way, I discuss techniques that statistical agencies have traditionally used to reduce disclosure risks in public use data files, as well as newer methods like synthetic data, verification/validation services, and differential privacy. I conclude with a set of big picture questions intended to motivate research on how agencies can facilitate access to confidential, blended data."
The Links Lecture is sponsored by the American Statistical Association to bring more visibility to issues surrounding the advancement of official statistics, particularly issues such as statistical use of administrative records and alternative data sources, record linkage, statistical methods for creating blended estimates, and reproducibility of results.
The Links Lecture Award is presented annually to an individual who has made a significant contribution to advancing official statistics and whose work and vision offer the promise to advance the role of official statistics in addressing critical data needs. Created in 2017, the award honors the significant contributions of three statistical pioneers — Constance Citro, Robert Groves, and Fritz Scheuren.
To listen to this free, virtual lecture, you need to pre-register via this link, after which you will receive a confirmation, invitation email from Zoom.