All Times ET
Program is Subject to Change
Statistical Data-Production in a Digitized Age: Potentials for Statistics, Research and Society (308299)*Stefan Bender, Deutsche Bundesbank
Digitization affects our daily life more and more. A decade ago, data were scarce, nowadays, data are everywhere. It is also affecting the production of official statistics or data handling in research. Digitization opens new possibilities, but can we use these data (like big data) as input for official statistics? In this lecture, first we will present new data sources and methods that are available nowadays, next to surveys and registers. This would include e.g. scraping the internet, the use of satellite images, or sensor data. Next, the presentation will shed light on different aspects of the so-called data life cycle or value chain of micro data production. We will discuss the advantages and disadvantages of bringing new data sources in official statistics and show how digitization can improve data quality by – for example - using machine learning algorithms. There is a strong need to link these different data sources also for data quality reasons and to document these kind of data in a new way (annodata schema) that the FAIR principles are fulfilled. The presentation will close with first steps how to measure the impact of official statistics and research in society.