Although the EVS teams in Tilburg and Cologne were able to move their work to home office relatively well, the challenges associated with the Coronavirus outbreak led to a loss of efficiency in the first few weeks.
It seems very difficult to meet the planned deadline for the EVS 2017 Full data release, which is very complex in terms of data and documentation. Therefore, the final release is postponed until early October.
At the same time, however, to make the current data available for research, a third (slim) pre-release will be available on 27 May 2020.
This EVS pre-release will include data from North Macedonia, Montenegro and Bosnia Herzegovina. In addition and due to their specific samples, individual datasets for Greece and the Swedish minority in Finland will be published in the context of EVS 2017 (30 + 5 countries). Finally, it will contain all new variables planned for the full release.
The new dataset structure (Main data file and Add-on Matrix data file) will already be applied. That is, we provide the ‘average user’ with a ready-to-use main mixed mode/matrix design data file (full cases only), where no selection of cases nor merging datasets is necessary, and with adequate sample sizes in each country. And provide the ‘expert user’ with an additional matrix design data file containing information and cases on the matrix ‘experiment’ (no follow-up, follow-up refusal, break-off cases), that could be useful for methodological research.
The Full release, in early October, will additionally include data from Portugal and (depending on the impact of coronavirus on fieldwork and on data quality) Luxembourg, Ukraine, and Latvia. Furthermore, there will be individual datasets for Turkey, Romanian-Hungary, and Malta.
It will include the following comprehensively documented datasets
Main data file
Add-on: Matrix data file
Add-on: Additional country-specific variables
Add-on: Sensitive data
Full release: 35 countries in 3rd pre-release + 7 new countries (probably not all possible until early Oct)
IDS Full release: 37 integrated and 5 individual datasets (probably not all possible until early Oct)