Through my work as Open Research Training Lead at The University of Sheffield, I talk to many researchers about open research practices. These practices can improve the quality of research by making the research process more transparent, the final outputs more accessible and/or the study results more reproducible. Many of these practices also focus on making the research environment more equitable and collaborative. Some examples of open research practices are study pre-registration, open study materials, open data and open-access publishing. From the range of open research practices, open-access publishing and open data are the most well-known. This is primarily through institutions, journals and funders increasingly mandating these activities. While open-access publishing is most well-established and generally accepted, open data is a topic that is often approached with a bit more trepidation. This post will outline some commonly faced challenges when making data open and offer some potential solutions and useful resources to address these challenges.
While many researchers understand the benefits of open data and are keen to implement it into their research process, many are understandably anxious. They could be worried about sharing data incorrectly or that others could find mistakes in their work. Further, the size or sensitive nature of some data makes it more complicated or not possible to be shared openly. This often leaves researchers feeling frustrated and concerned that they may be unable to comply with funder, journal or institutional policies. In these cases, I recommend that researchers aim to make their data as FAIR (findable, accessible, interoperable and reusable) as possible and that this doesn’t necessarily require the data to be open. The opposite is also true; data can be open, but not FAIR. For a deeper discussion of this (and how research data management also relates), I recommend reading “Three camps, one destination: the intersections of research data management, FAIR and Open” by Higman, Bangert & Jones (2019).
Some of the key barriers to implementing open and/or FAIR data are as follows:
- Lack of training and incentives are often cited as a barrier when implementing open and/or FAIR data. Even when researchers are motivated and able to share their data, it requires learning new skills and this takes precious time. These new skills include preparing data or other research materials to be shared (for example, anonymisation) and learning how to use new tools like data repositories or the Open Science Framework. Further, there are systemic pressures on researchers that don’t incentivise learning these skills, for example, they may not be recognised in institutional hiring and promotion criteria.
- Where training is available, it is often generic or introductory-level. Standards and norms in different disciplines are often very different and the process for making different data types open and/or FAIR is similarly varied. Each data type brings its own challenges and more training resources are required that address how to make specific data types open and/or FAIR. This training would help bridge the gap between the high level benefits of open research with more concrete, directly applicable guidance.
- Researchers often report lacking enough support to implement open data. Best practices in research data management, open data and FAIR data are evolving rapidly and it can be challenging for researchers to stay up to date. As the complexity of the research process and demands on researchers increase, more support is required. The importance of having expert staff like data stewards on hand to assist is being recognised more.
New resources are being developed to address these challenges:
- New training resources and initiatives to address incentives. The UK Reproducibility Network (UKRN) is aiming to provide these with its Open Research Programme. This is a national 5-year project aiming to increase open research capacity through a train-the-trainer model. The programme links together training providers (such as the Center for Open Science) and over 20 UK institutions to source and train new open research trainers. Another arm of the programme is working with 43 institutions across the UK to better understand and address reward and recognition.
- New training materials using specific data types. ELIXIR-UK is aiming to provide this through the creation of RDMbites. RDMbites are bitesize videos on a wide range of research data management topics that are publicly available through ELIXIR-UK’s Youtube channel.
- More sources of support. ELIXIR-UK is also aiming to address this with its FAIR data stewardship fellowship. Fellows (like me!) create training resources like workshops, RDMbites and blog posts to support researchers. Research tools can also be sources of support, for example, Data Stewardship Wizard is a tool for data management planning that guides researchers through the process of managing their data while also providing advice and tips to effectively promote FAIR data management.
The research landscape is changing rapidly with increasing demands for improved research data management, FAIR and/or open data, leaving researchers with a lot to learn. However, while many researchers are motivated to learn the skills required to implement FAIR and/or open data, more training and support is needed. Programmes such as the UKRN’s Open Research Programme and ELIXIR-UK’s FAIR data stewardship are developing new resources to address this need.
Related resources
- ELIXIR-UK website https://elixiruknode.org/
- ELIXIR-UK Fellowship https://fellowship.elixiruknode.org/
- RDMkit https://rdmkit.elixir-europe.org/
- FAIR Cookbook https://faircookbook.elixir-europe.org/
- FAIRsharing https://fairsharing.org
- Data Stewardship Wizard https://ds-wizard.org/
- Higman, Bangert & Jones (2019) https://doi.org/10.1629/uksg.468
- UKRN Open Research Programme https://www.ukrn.org/open-research-programme/
