The Royal Society Science Policy Centre concluded: “[…] data that underpin a journal article should be made concurrently available in an accessible database” (2012). This principle is shared by funding bodies worldwide and is supported by a growing number of major publishers (Nature, 2016; PLOS, 2016; The Royal Society, 2016). Many disciplines have subject specific data repositories that align with open data initiatives, such as Dryad and PANGAEA.
However, there is a gap in this space which can be partly filled by established institutional repository (IR) services, which offer reliable and robust solutions for publication-related datasets.
The Australian National Data Service (ANDS) and the University of Queensland (UQ) Library are collaborating on a project to explore using the IR (UQ eSpace) to store, describe, and share data underpinning UQ publications.
Although pockets of researchers within UQ have well established data sharing practices, a data sharing culture has not yet been institutionalised. This project will allow us to capture the research data and provide us with a context for promoting data sharing practices with researchers.
The aim of the project is to source 500 publication related datasets and describe them in UQ eSpace. The project has two phases: a pilot that will allow us to create processes and gather feedback, and a larger rollout.
Our initial contact list of researchers included those who recently published in a journal on the Nature Index list or in a PLOS publication, all of which have a data sharing policy. In collaboration with Client Service Librarians, we use tailored and flexible approaches to contact the different researchers and groups. We have offered to create IR records for existing data, even if they are stored elsewhere (e.g. in Figshare). We have used this opportunity to advertise the IR as an attractive alternative for future data sharing activities.
The project, which will continue to run until mid-2017, has had positive results. By providing a tangible service to meet an identified need we have established UQ eSpace as a useful tool for meeting publisher data sharing requirements. Taking a personalised approach by contacting researchers and groups directly, we were able to gather iterative feedback on our processes and systems, which allowed us to make crucial improvements along the way.
In conversations with researchers, we have learned about their data sharing practices and pain points, which will inform future Library data management services.
So far, we have discovered that by approaching data sharing in this targeted way we have seen positive outcomes—a greater number of discoverable datasets in the IR and an improved data sharing culture.
The data sharing landscape is still evolving and there are considerable issues for researchers and institutions to overcome. However, at UQ we have had positive outcomes with this approach to data sharing. It is too soon for the full impact on the University’s data sharing culture to be measured, but through this process we will continue to improve it one publication at a time.
Presentation - Available now.
Paper - Available now.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.