WP8 Overview1

Workpackage 8 (WP8) of ESSnet Big Data focuses on the methodology, IT and quality aspects of the ESSnet Big Data project. Because these aspects can only be addressed after the content workpackages 1 to 7 have laid the necessary foundations, activity in WP8 only started at the beginning of the second stage of the project (SGA-2, running from January 2017 to May 2018).

WP8 is carried out by representatives of seven ESSnet Big Data partners: CBS (Statistics Netherlands) which is leading WP8, Statistics Austria, Statistics Bulgaria, ISTAT (Statistics Italy), GUS (Statistics Poland), Statistics Portugal and Statistics Slovenia.

Getting a firm grip on methodology, IT and quality issues is essential for any attempt to integrate big data in the production of official statistics. Although it is obvious that the content workpackages focusing on specific big data types also need to consider methodology, IT and quality, it was deemed necessary to create a dedicated workpackage starting in the second project stage for two reasons. First of all, in the paradigm shift big data have caused in official statistics, the importance of methodology, IT and quality cannot be overestimated and they are too crucial to be left as a side issue only in the various content workpackages. Secondly, big data methodologies are diverse and dependent on the specific type of data used, but there are still commonalities which need to be identified. And finally, in the multidisciplinary approach required by big data, the methodologists and IT specialists are needed too.

Because big data methodology, in the broader sense, is a relatively new field and because many different and very diverse types of big data exist, it is far from easy to identify the methodological challenges they present, to find common ground among them and to rank them as to priority; this is the first task of WP8. The next steps, after having identified the most pressing methodological problems, is then to devise workable solutions and to test them. This should then result in a set of recommendations and guidelines which can constitute the basis of high-quality official statistics based on big data sources.