11.2 Selecting Research Data for Reuse: Challenges During the Implementation of Support for Data Archiving in Research Libraries

This paper offers a new approach for improvement and cooperation on data collection management for research libraries. As a consequence of a scientist’s interest or curiosity, every object or digital asset can become input for research and consequently be called scientific data. Research libraries are adopting a variety of new responsibilities to help researchers with their data collections, giving support during planning, gathering, managing, sharing and archiving scientific data. This paper will focus on the libraries’ task of archiving and disseminating research data, offering a reflection on the challenges that research libraries in the Netherlands face when implementing support for data archiving and diffusion.

Dutch libraries have basically two main roles when supporting their institutions with data archiving. First, libraries can be responsible for the institutional data archive. The institutional archive has a different function than a data repository. According to the Netherlands Code of Conduct for Scientific Practice, scientific results such as data should be made available to peers for verification on accuracy and truthfulness. Each research institution has responsibility for archiving all the scientific data that has been used as evidence for a theoretical framework, and which has resulted in a publication for at least ten years. Libraries can manage the local archive and give access to the data for control purposes when needed.

A second important role on data appraisal for libraries is opening up data for reuse. Knowledge of collection management applies to all libraries, but selection criteria for research data differ considerably from acquisition criteria for publications. The selection of data is not based on the question whether the object fits in the library collection, but whether the object is relevant for others. General selection criteria on data selection are available in different studies on this subject. In this paper a brief overview is presented. The remaining challenge for libraries is to transform those general criteria to suit the needs of data reusability for a specific discipline. Each research library has its own collection management profile with core elements that are common to all scientific practitioners. For the selection and publication of data, every library will implement its own collection model in accordance with its research fields, cost models and (inter)national appointments for specific collections.

Ana van Meegen is Data Librarian at the Vrije University Amsterdam and a member of the Special Interest Group (SIG) Research Data at the Netherlands. Since 2012 she has been working on developing data management support at her own University and in co-operation with other research libraries and different stakeholders.