Stijn Michielse, a visiting PhD student at the University of Cambridge (Department of Developmental Psychiatry) from Maastricht University, Netherlands talks about his visit to The Data Dialogue: Time to Share and what he took away.
My background is in neurobiological sciences working on a project involving loads of confidential data such as MRI scans, DNA material and medical records, so with great interest I took part in The Data Dialogue in order to learn more about proper data management, documentation and sharing from different viewpoints. Indeed it is Time to Share!
Two very relevant topics were raised during the conference:
- The use of Digital Object Identifiers (DOI’s) came up as a way to make use of each piece of research, linking them together. This concept can be applied to link a presentation with data, papers and personal OrchID identification so that data has a life outside of PDF or PPT via DOI. Data sharing initiatives such as Figshare, Repositive.io, GitHub and NeuroSynth (some of which were mentioned during the conference) are very useful and help enable sharing.
- Transparency and reproducibility are big items in the scientific research community, where proper studies/projects should be making replication possible. Sharing code and data actually allows for a duplication of findings and an independent researcher can now have a new look at research data, improving upon the used methods. Replication has added value over duplication and can be a good way to teach students on how to do research when working with statistics and real data.
Although it can be a bit frustrating, it might be a first step towards a publication for students. A failed replication is still a learning point and indirectly advances science. The panel discussion was also great with some practical question on how to put this all into action. In short one might say: Sharing = caring (about research).