The core of the EuVisions data project is composed of three manually coded datasets – Politics/Events (PE), Institutions/Formal acts (IF), and Ideas/Think tanks (IT) – and a number of event-centred social media datasets. Collection protocols and codebooks for each of our manually coded datasets have been elaborated and refined over several months by the EuVisions team in constant dialogue with the broader REScEU research group. Procedures for social media data, on the other hand, are formulated on a case-by-case basis in accordance to the specific question being investigated. Software for the collection and classification of social media data is provided by our London-based partner, the Centre for the Analysis of Social Media (CASM).
The formulation of our methodology has been guided all along by the principles of coherence, transparency and reliability. In setting up variables and procedures we have tried to remain as close as possible to established best practices in relevant fields such as event data, qualitative content analysis and natural language processing. Time and again, however, we have had to break new ground due to the novelty of our project as regards both the type and range of data collected.
Below are links to the latest versions of the complete methodology and related documents for our manually coded datasets. As it is usually the case, these documents remain subject to constant scrutiny and will be reviewed periodically to ensure their continued solidity, particularly—though not only—in case of future expansions of the project. As we do so, we welcome any feedback and requests of further information.