Session 9: Trusted partners in research data support
Friday 8 July – 9:00 – 10:30
Chair: Hardy Schwamm, James Hardiman Library, National University of Ireland, Galway, Ireland
9.1 Building a research data support service from an experiment: the case of the University of Strasbourg, Stéphanie Cheviron, Adeline Rege, University of Strasbourg, France
When University of Strasbourg started to develop research data management services in 2014, the Library Services (SBU) conducted a quantitative survey and interviews to find out the needs and practices of researchers. These studies revealed that researchers in the humanities and social sciences needed tools to visualise their data and make them available online.
We therefore set up in 2015 an experimental support service for researchers jointly with the IT department (DNum). We chose 4 pilot projects in the humanities and social sciences as use cases to develop web applications to exploit, describe and disseminate data.
We aimed to :
Determine what were the most common requests;
Provide the human and technical resources necessary to address them;
Define an adapted service offer;
Size the research data management team within the library and develop the appropriate skills.
From 2015 to 2021, the data librarian was “embedded” in the 4 research teams. The library helped researchers write functional specifications for the applications. We proposed a data model to structure the data. As a result, several databases were developed, including an application that became the backbone of several other projects.
We discovered during this experiment that working with researchers is multifaceted. Not all projects can be supported because some are not sufficiently technically defined. Researchers also have to be available. Therefore, lack of maturity and schedule conflicts can be detrimental to the success of a project. In addition, participating in a software development project is a long-term commitment because of new features and updates. The accumulation of projects supervision is time consuming.
Finally, the library has often gone much further in its activities than originally thought because we have in fact taken on a role of coordination and interface between the various stakeholders that no one had previously assumed. This is especially true with the DNum because a research project is different from an IT development project, and most of the researchers had never worked with IT specialists, while the DNum had never worked with researchers. However, the requests are diverse and can involve other university services.
The need to structure the service in a more robust way became apparent. The first step was to put an end to this experiment by defining a comprehensive and common service offer with the DNum, the Research Data Helpdesk. We now provide data management plans support, advice on repositories, and guidance on metadata standards. The library team strengthened with an additional FTE. The DNum created a service dedicated to research data.
The second step was to structure our organization at the institutional level and with our partners, by grouping all data support services within the university in a « data workshop ». At the end of 2021, the Ministry of Higher Education, Research and Innovation launched an initiative encouraging research institutions to create and certify “data workshops”. This one-stop shop will provide support covering the entire data lifecycle and will bring together all the skills needed to guide, help and train researchers.
9.2 Data as a new research publication type : What could be the role of research libraries as service providers? Mari Elisa Kuusniemi, Helsinki University Library, Finland, Susanna Nykyri, Tampere University Library, Finland
It would be beneficial for the research and researcher that the data would be recognised as a concrete research output, which also meritates its contributors. Research libraries could take an important role in the process in which a data set will be curated and polished to a data publication.
Research libraries already provide support for research data management especially in the form of guides and training. If data support in research organisations is focused solely in supporting data management, it leaves the possibilities and benefits of FAIR data and also the quality assurance of the research process halfway.
In this paper we analyse what is already commonly included in current research library service models, what is not, and what could be the extended role of the research libraries in data publishing. We see the library as a relevant, potential instance in taking and carrying the responsibility of the entirety, but in broad and well-formulated co-operation with other stakeholders.
As background we illustrate research data as academic publication type, which includes also the needed curation and peer-review process.
Data publishing service requires enhancing the skills of library staff
The new kind of research data publishing service challenges e.g. current skills, service and business models. It also concretises the needed broad multiprofessional co-operation. Libraries are already competent in knowledge management and ensuring the long-term accessibility to information and developing open access models for publishing. However, enhancing skills in the area of data quality, documentation and metadata is needed to cover the most important aspects of research data curation, which challenges traditional librarianship.
We have already data publishing services
The data publishing channels play a critical role in enabling FAIR principles. Many libraries provide a data repository or provide recommendations to use certain data repositories maintained by third parties (commercial or governmental). We consider the minimum level of supporting data publishing achieved, if a repository provides persistent identifiers and sufficient metadata for data discovery. Further needed, more advanced services are provided, when data is curated or peer reviewed during the submission process. Several libraries develop curation processes e.g. when they create criterias to select data sets for long-term preservation. However, how often do we see this as data publishing? Do we see these services as data publishing services? It is also quite rare that a library is a publisher of a data journal.
As a result we suggest concrete aims and possible service models for the research libraries as data publishers. The benefits and weaknesses of different solutions are illustrated, comparing institutional repository, long-term preservation and archiving data, as well as establishing data journals. Different kinds of solutions are mirrored also comparing the organisational vs. international (e.g. within EOSC) possibilities.
About methodologies for user-centered design of research data services , Karin Cecilia Rydving, University of Bergen, Norway
In September 2020, the University of Bergen (UiB) adopted a policy for Open Science. The University of Bergen Library has been assigned responsibility for following up open access to research data, in collaboration with the Research and Innovation Department and the IT Department. A cross-departmental working group has been set up with a mandate to plan and implement the following main themes based on the follow-up points in UiB’s policy:
Establishment of a joint service for guidance and training within making available and archiving research data at UiB.
Tasks related to digital solutions for making research data available and archiving.
In order to approach the work with research data, the working group has received support from UiB’s Service Development Project in collaboration with AFF, which is part of the Norwegian School of Management’s research environment. The UiB Service Development Project has a user-friendly and exploratory approach to the tasks that are to be solved and will challenge working methods and practice, and contribute to future-oriented and innovative services to ensure good support for the university’s primary tasks.
The contribution presents methodologies from UiB Service Development that the working group has used and the experience of these. Furthermore, the contribution reflects on how the methodologies develop the library’s competence and role at its own institution. Examples of methodologies are Start Smart and Google Design sprint.
Start Smart has been developed by researchers at the Norwegian School of Management and is a structured and time-limited method for starting up in teams and groups. During a facilitated workshop, the participants work actively to clarify ambitions and goals, competence and resources, roles and functions as well as the form of work for the group as a whole.
The working group for research data has also carried out a seminar and used elements from the Google Design Sprint. For example insight interviews with researchers, developing personas, stakeholder analysis and Cover Story Vision. The sprint methodology’s various roles as participants, experts, decision-makers facilitate the inclusion and anchoring of the process.
The working group’s experience is that the methodologies create unity and enthusiasm both within the group and across UiB. The methods help to highlight and complement each other’s expertise. The procedure strengthens user involvement and provides new insights. The methodologies also require the active participation of the participants in the various roles. The need for cooperation between different departments to develop good services within the institution is successfully highlighted.
The working group notes that the methodologies used have effectively managed to create consensus on a draft for an action plan for the next three years, that is linked to specific tasks with roles and responsibilities. We have established a user-centered focus and will continue on this way. The methodologies contribute to achieving good quality in a short time in the work with service development within Open Science.