|Open Research, sometimes referred to as Open Science, is the idea that both the research process and outputs should be made publicly available as soon as possible. Much of the discussion around Open Research focuses on Open Access and Open Data but it also includes peer review, protocols, notebooks, citizen science, software and much more. It is widely acknowledged that making research more open will support reproducibility, enable collaboration, increase re-use and make it accessible to wider society.||
Policy support for Open Research
Policies supporting open access to publications began over a decade ago, all major UK research funders now have open access mandates and the University has an open access policy. These have often been followed by policies encouraging or mandating data sharing from funders and the University, and the Concordat on Open Research Data. The Concordat calls for data to be made openly available and was signed by Universities UK, UKRI and the Wellcome Trust amongst others. The Wellcome Trust have led the way in extending their policy to include mandating the sharing of software and materials as well as publications and data. Most recently the University’s Vision has been published which includes a commitment to create an open research culture aligned with the FAIR principles.
These have been supplemented by the Declaration On Research Assessment (DORA), which has been signed by the University of Sheffield, and aims to improve how research outputs are assessed and encourage the responsible use of metrics. This commitment to responsible assessment is reflected in the University’s Academic Career Pathway which emphasises the quality of individual articles rather than the journals they are published in. Researchers are supported in engaging with Open Research by the Library, Research Services, IT Services and the new Research Practice Lead, Tom Stafford.
Practical support for Open Research
Open Research practices vary between disciplines and methodologies, but below we have gathered a range of tools, platforms and practices which support different aspects of open research across disciplines. If you are concerned about what others will be able to do with materials you have shared, or would like to know how you can re-use any materials you have found, then you can contact our copyright team who can provide advice on licensing.
Services such as GitHub, GitLab, BitBucket allow researchers to manage software and share it openly during the development process. At the end of the research project software can be archived in ORDA and publications like the Journal of Open Source Software allow researchers to get credit for their software in an academic journal.
Posting preprints, research articles which have not yet been peer-reviewed, is an easy way to share results quickly. There are many preprint services for different disciplines and these are detailed in our Scholarly Publishing guidance.
Publicly sharing protocols for research experiments can assist others to reproduce your research, speed up progress in a field by improving the methods and allow others to build on the protocols you have developed. Sharing protocols can also be useful when journals have strict word limits in the methodology section, as you can simply link to the detailed protocol shared separately from your article. There are protocol sharing websites, such as protocols.io, where you can usually receive a DOI to make your protocol easily citeable or you can share protocols in methods journals, such as JoVE.
Registered reports involve breaking down peer review into two stages; the first review is of the idea, design and analysis plan of a project prior to the research being conducted. There is then a second stage of peer review after the research has been undertaken, looking at the final report. This process supports reproducibility as registering the analysis plan for a project prior to data collection prevents researchers hypothesising after the results are known ( also known as HARK-ing). Find out more about the journals involved at https://www.cos.io/initiatives/registered-reports.
|Hardware||Open hardware refers to making the specifications and designs for physical hardware openly available and licensed for re-use by others. There are detailed descriptions of best practice in making hardware open in The Turing Way.|
|Notebooks||Electronic Lab Notebooks can make it easier to open up research workflows, and some groups have even started making their notebooks open during the research process.|
|Peer Review||Peer review is not only affecting the outputs which can be shared, it also has implications for peer review, with some arguing for an open peer review process. There are several platforms which provide post-publication peer review, such as F1000, and journals offering partially open peer review, such as eLife which publishes peer review reports after publication. Additionally, Publons allows you to get ‘credit’ for the publications you have reviewed and Hypothes.is provides annotation services, another form of post-publication peer review.|
|Identifiers||As more research is being made openly available it is becoming increasingly important to add identifiers to each research output so that they can be linked and to ensure that researchers get appropriate credit for their work. Journals and repositories (including ORDA) will routinely assign DOIs (Digital Object Identifiers) to articles, datasets and software that they host. Doing this makes it easier to persistently link outputs and track their citations. Additionally it is really important to include your ORCiD in each output so that you can be credited for your research. There are also moves to make it clearer who contributed what to an output in order to improve transparency by using the CRediT taxonomy.|
|Platforms for sharing and discovering other outputs||If you wish to share other outputs or facets of your research then there are a number of platforms which will support this. The University’s research data repository, ORDA, accepts most non-article outputs including reports, workflows and posters. Alternatively there are similar public platforms such as Zenodo and the Open Science Framework which is particularly popular for sharing research throughout a project to improve reproducibility. There are also platforms for specific funders, such as Wellcome Open Research and Gates Open Research, which allow the sharing of a wide range of outputs including protocols, descriptions of results and method articles. For teaching focused texts it is worth considering specific textbook platforms such as the Open Textbook Library.|