Code4REF
Quick links to system-specific instructions:
Reporting research software in CRIS (Current Research Information Systems)
Research software is an eligible REF submission. However, the majority of REF submissions are research publications. Indeed, refereeing research software is a difficult task, and reviewers may not have clear guidelines for its evaluation for the REF purposes (something else we need to work on), and many researchers prefer the safer route of submitting papers. As a result, software outputs are not always being recorded in Current Research Information Systems (CRIS) used by universities. The first step here would be to help to research software developers to find a way to record their outputs in CRIS. This may not always be obvious, so on this website we provide general guidelines and step-by-step system-specific instructions for doing this.
Recording research software in CRIS is useful for many purposes, not only for REF, and we encourage developers to report all your software outputs there. A survey of 15 Russell Group universities found that 92% of researchers use research software, 67% find it fundamental to their research, and 56% develop their own software. To get further information, for example, an overview of all research software developed at an institution, in the research group or by an individual developer, one could use CRIS or their public views (e.g. research portals), provided they contain relevant records. This will provide further evidence that software is vital for research, should be treated as research output, and will make a stronger case for the campaign for the recognition and adoption of the RSE role within academia, led by the UK Research Software Engineer Association.
Current Research Information Systems used in UK Universities
UCISA regularly survey UK Higher Education Institutions about the systems they use, including CRIS. The results are available online.
Elsevier’s Pure is the most widely-used CRIS in the UK, followed by Eprints and Elements/Symplectic. If you use one of the systems not yet covered here, please consider contributing to this project via our GitHub repository.
- Pure
- EPrints
- Elements (Symplectic)
- Converis
- IRIS
- Worktribe/RIS
- Vidatum
Process Overview
There are several broad categories of software which should be recorded in a CRIS.
- Open-source software
- Closed source software for which executables can be made available
- Software projects which are unable to make any outputs available, but which can be publicly acknowledged
- Software projects which can not be publicly disclosed
You may wish to create a single record for a piece of software, or multiple records for different versions of that software.
For open-source software, we suggest that the project is archived using GitHub’s integration with Zenodo (or similar), which will provide a DOI.
In our experience, a CRIS typically classifies a software output as a kind of research output, effectively treating it as a special kind of paper. The exact details will depend on the system in use and how it has been configured.
If open source software has been archived elsewhere, you should reference that in the CRIS record; otherwise you can reference the source code repository.
If executables are to be shared, you can upload them to the CRIS or link to them if they are hosted elsewhere.
You should be able to apply access restrictions to the record as appropriate.
Other useful resources
- The Software Sustainability Insitute has published Software Deposit Guidance for Researchers.
- GitHub’s integration with Zenodo provides a powerful and user-friendly mechanism to archive software and get a DOI, including automated tracking of future releases.
About this project
This project started at the hackday at the SSI’s Collaboration Workshop 2019.
The hackday team:
- Louise Brown, University of Nottingham
- Alexander Konovalov, University of St Andrews
- Patrick McCann, University of St Andrews
For a recent update, see Alexander Konovalov’s talk at RSE19: Code4REF: How to learn which software is developed in your institution?