Improving the usefulness of research data with better paradata

Abstract Considerable investments have been made in Europe and worldwide for developing research data infrastructures. Instead of a general lack of data about data, it has become apparent that a pivotal factor that drastically constrains data use is the absence of contextual knowledge about how data was created and how it has been curated and used. This applies especially to many branches of social science and humanities research, where data is highly heterogeneous, both by its kind (e.g. being qualitative, quantitative, naturalistic, purposefully created) and origins (e.g. being historical/contemporary, from different contexts and geographical places). The problem is that there may be enough metadata (data about data) but there is too little paradata (data on the processes of its creation, curation and use). The aim of this position paper is to draw attention 1) to the need for a better and more systematic understanding and documentation of the contexts of creation, curation and use of research data to make it useful and usable for researchers and other potential users in the future, and 2) to specific obstacles that make the capturing of this particular type of metadata, known as paradata, especially difficult. Failing to understand what information about the creation, curation and use of research data is needed and how to capture enough of that information risks that the currently collected vast amounts of research data become useless in the future.