Data analytics in practice - real examples from the ESS


While intuitively understandable, the concept data analytics can mean different things to different stakeholders.

At this daWos session, concrete examples from different ESS members will illustrate what data analytics actually could achieve for a statistical institute. We will then discuss which lessons could be learned from them. 

Questions that could be addressed include

  • How was the use case "delivered": as an ad-hoc prototype, an ad-hoc study, a pilot or an experimental statistical product?
  • Which processes of the statistical workflow were concerned? Was it possible to provide the "full operational stack" (i.e., implementing a totally new product)?
  • Before a "full operational stack" is actually adopted in production, are the implemented data analytics tools/methods (or parts of it) reusable and applicable to other studies or other domains?
  • Altogether, what is actually the impact of data analytics on the role of statistical institutes in the provision of statistical products for policymakers? 
    • Is data analytics "yet another tool" to carry out (and maybe improve) classical statistical operations (e.g., any common process identified in the GSBPM[1]), i.e., at the end, run the business as usual? Which classes of methods/techniques are actually useful?
    • is data analytics disrupting the way statistical institutes currently inform policy with data (e.g, from data collection to decision-making, through data collection and indicator estimation)? How?
  • Beyond development and deployment, what are the main challenges ahead for fulfilling the use cases for data analytics?

[1] https://statswiki.unece.org/display/GSBPM/GSBPM+v5.0.