Research data management
Posted by knockels on 29 November, 2011
Yesterday I attended a meeting of the University Science and Technology Librarians’ Group in Cambridge, on the subject of Data Management.
Research data management is an essential skill for researchers. What data do they delete? What do they back up and how? Who will be able to access the data when they have left the University? What counts as “data”? How can they store and structure the data? A horror story was told in which someone left computer equipment containing five years of PhD data in the pub, with no backup…
Two papers looked at projects in the area of data management training. Incremental produced training material available at www.lib.cam.ac.uk/dataman. DaMSSI used the Sconul 7 Pillars and Vitae Researcher Development Framework to develop a skills framework for training. There is more information at www.tinyurl.com/DaMSSI-DCC and www.tinyurl.com/DaMSSI-RIN. DaMSSI also produced career profiles, to show the relevance of data management skills to a range of scientific careers.
Yvonne Nobis of Cambridge University described interesting research undertaken at Cambridge. There are projects at Cambridge where people from different areas work together – the Pathgrid project, for example, uses image analysis software developed by astronomers to analyse pathology slides. This came about because of a conversation over lunch. Are there any formal mechanisms for helping this happen? The research aimed to discover whether potentially useful software developed for one purpose was shared with or discovered by people in other disciplines. It looked at physics and at bioinformatics. Interviews conducted for the research indicated that some researchers were involved in writing code from scratch because they did not know that someone had already written what they needed. At the same time, it uncovered reasons why researchers did not share code they had written – these included not wanting requests for support, errors being uncovered, and problems getting recognition for your work if it was useful outside of your field. The argument was well made that as research data these days is often derived from other data, using software, that software is a research output that should be made available alongside the data itself.
What is the role of the library in data management and in curating software? It was suggested that libraries are seen as the people who deal with information, and so it falls within our remit. It was further suggested that we might help by running training and producing training materials, by facilitating a workflow to make software discoverable, and by devising ontologies (the JISC-SWOP project has done this, in fact).
I can see the argument here, although I can also see that the subject falls within others’ remit as well. I wonder what is already on offer at Leicester from IT Services, Research Support Office, from College doctoral training programmes and the Skills for the Professional Researcher training, and from departments. That would be interesting to know, and perhaps help us to decide if we do have a role, in coordinating existing provision or offering new training.
On another matter I have offered to write a short piece for the Health Libraries Group Newsletter to publicise USTLG to health and medical librarians who find themselves looking after science. One colleague present yesterday was a medical librarian who had added science to her portfolio.