Incentivising sustainable and collaborative research
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Title of Series | ||
| Number of Parts | 13 | |
| Author | ||
| License | CC Attribution 4.0 International: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
| Identifiers | 10.5446/55219 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
00:00
Computer animation
00:11
Meeting/Interview
00:23
Meeting/InterviewDiagramProgram flowchart
02:59
Computer animation
07:24
Meeting/Interview
08:46
Meeting/Interview
09:10
Computer animation
11:18
Computer animation
12:57
Meeting/Interview
13:40
Computer animation
Transcript: English(auto-generated)
00:12
Hi, I'm Rachel Ainsworth and I'm the Research Software Community Manager at the Software Sustainability Institute, the UK's national facility promoting the advancement of software
00:22
and research. Thank you to the organizers for inviting me to share my thoughts on incentivizing sustainable and collaborative research. We are experiencing a strong movement towards reproducible and open research, which has presented new challenges and opportunities with regard to ensuring the rigor and integrity of data-intensive science.
00:41
Embedding openness and reproducibility within every stage of the research workflow is not trivial, from the initial research questions, planning and design, to the collection, processing and analysis of research data, to the publication, accessibility, and preservation of research outputs to enable reuse. It is also not yet fully supported or incentivized within modern research culture.
01:04
Studies show that the evaluation of researchers still emphasizes traditional criteria, such as the number of publications, authorship order and publication, and journal impact factor, over more collaborative criteria, such as data sharing, open access publication, and registration of studies.
01:22
This emphasis on traditional criteria and evaluation and promotion is highly focused on the publication outputs of the individual, instead of the process by which the research is conducted. It does not incentivize reproducible research or open and collaborative science. But modern research requires collaboration between individual scientists, research teams,
01:45
data service providers, research infrastructures, and institutions. It is increasingly recognized that data-intensive science requires not only technical skills in experimental design, software development, and data management, but also people-focused skills, such as communication and team working.
02:03
In many fields, there is also a need for ethical and legal expertise, particularly when sensitive data is being used. Science needs both digitally skilled researchers and a variety of professional research support staff in order to ensure rigor and integrity along the entire research workflow.
02:21
The Organization for Economic Cooperation and Development identified key action areas that need to be addressed in order to build and maintain this digital workforce capacity for science. They include supporting the development of communities for new professional roles, learners, and trainers, and changing academic evaluation and reward systems in order to attract and
02:42
retain diverse digitally skilled staff. In this talk, I will highlight some of the progress in these areas to build systemic change towards reproducible, sustainable, and collaborative research. I will begin by highlighting some of the professional research support roles that are emerging. The first is the data steward.
03:02
Data stewardship is the responsible planning and executing of all actions on digital data before, during, and after a research project, with the aim of optimizing usability, reusability, and reproducibility of the resulting data. It is crucial for good scientific practice and research excellence, as it enables long-term
03:21
and sustainable care across multiple research cycles. There has been a lot of effort, particularly in the Netherlands, to professionalize the data steward function. However, there is currently a lack of consensus around the responsibilities, tasks, required knowledge, skills, and abilities. And this hampers adequate data steward capacity and complicates efficient data management
03:43
and handling in the various scientific disciplines. It also makes recognition, reward, training, and progression difficult for the data steward, which is an essential part of the research process. This report, Towards Fair Data Steward as Profession for the Life Sciences, by Shelton
04:01
et al., works towards a common job description and agreement on the required competencies and identifies that data stewards at the interfaces of policy, research, and infrastructure ensure that research carried out at institutions and in projects produces fair data.
04:21
That is, ensuring that the research data are findable, accessible, interoperable, and usable along the data life cycle. Furthermore, data stewards will be essential to help researchers interact with emerging national and international data infrastructures, such as the European Open Science Cloud, which aim to provide a trusted virtual environment, enabling open and seamless services
04:44
for data storage, management, analysis, sharing, reuse across disciplines. The second research support role that I want to highlight is the research software engineer. Software is vital to research and data intensive science, which increasingly needs to adopt
05:00
advanced computational processes to manage and understand research data. This article on the four pillars of research software engineering illustrates that good research software can make the difference between valid, sustainable, and reproducible outputs and short-lived potentially unreliable or erroneous outputs.
05:20
The past few years have seen the emergence and rapid growth of the concept of research software engineering, which combines professional software engineering expertise with an intricate understanding of research. Discussions initiated by the Software Sustainability Institute in 2012 identified that this combination of skills was extremely valuable, but there was no formal place
05:40
in the academic system at the time for the people who write, maintain, and manage research software. This meant that there was no easy way to recognize their contributions, reward them, or represent their Ds. Since then, the Institute has campaigned to raise awareness of the role of the research software engineer and to build a community around it, with the aim that if we could
06:01
support the rise of expert software developers in academia, there would be a growth of reliable, well-engineered research software, which provides reproducible results. Research software engineers are now increasingly seen as critically important members of research teams and institutions to facilitate this. Although more funders are recognizing the importance of providing funding for research
06:23
software development, this is normally focused on the development of new or novel software for a particular research project and not on its long-term sustainability or maintenance. However, research software often continues to be of use for years, if not decades, requiring maintenance to realize its full value.
06:42
I love the quote from this blog post, which states, hardware changes, languages and tools develop, bugs are found, needs evolve, and unless the software is updated to reflect this, it will become increasingly full of potholes, like an unmaintained road. Furthermore, as research teams seek to work in a truly open manner, using open source
07:03
software can become an issue to reproducibility. This is because these tools rely on their communities to maintain them, which is not always sustainable, or they get bought out and merged into commercial offerings. So as a researcher, you can work entirely openly and still discover that in a few years' time, your work is not reproducible because the software has not been
07:22
maintained or it no longer exists. So in addition to fully professionalizing and supporting career paths for research software engineers, software maintenance must also be incentivized and funded to facilitate reproducible research. This initiative to provide funding to adapt or maintain existing software used by
07:41
researchers by UK research and innovation is a very welcome start. The third research support role that I want to highlight is the community engagement manager. Science is inherently a community-based endeavor. The generation, validation, and dissemination of knowledge requires a network of diverse roles and a range of community configurations to meet specific needs.
08:02
Communities in science are also key to the ongoing cultural shifts in how science is carried out, from reproducibility and open science to diversity, equity, and inclusion efforts. Communities provide places to try, iterate, and adopt new norms, and they are the best places to cultivate and nurture a culture of sustainable and collaborative
08:21
research. Scientific community management is an emerging role found in a range of different contexts, including communities of practice, professional associations, infrastructure organizations, and research collaborations. The essential function of a community manager is to align community member goals and create and maintain pathways by which members can engage with one
08:44
another and or the community's projects. The Center for Scientific Collaboration and Community Engagement is working to professionalize and institutionalize the role of the community engagement manager within science. As a community manager at the Software Sustainability Institute, I built community through activities such as the institute's fellowship program,
09:03
which provides funding for researchers who want to improve how research software is used in their domain, in their domains, an area of work. I also managed the institute's premier annual unconference event, the Collaborations Workshop, which brings together all stakeholders within the research software community to explore best practice and shape the future of
09:21
research software. It was through discussions at Collaborations Workshop 2012 where the research software engineering movement began. There have also been ongoing discussions through Collaborations Workshop and the fellowship network on how we can ensure the long-term sustainability and maintenance of research software, which resulted in the blog post I showed a few slides back.
09:42
Building community is important for promoting and empowering new and collaborative ways of working. So now I will highlight some initiatives which are not only successfully providing resources and training to improve practices within data intensive science, but are also inspiring examples of open and collaborative projects. The first is the Turing Way Project.
10:02
We know that reproducible research is necessary to ensure that scientific work can be trusted, and we've seen that it means understanding data management and software development. Both are skills that are not widely taught or expected of academic researchers and data scientists. The Turing Way is an open source collaborative project that aims to
10:21
help researchers, software engineers, project leaders, and funding teams understand their roles and responsibilities of reproducibility in data science. It involves and supports its members of diverse skills and backgrounds to ensure that data science is accessible and useful for everyone. Its moonshot goal is to make reproducible research too easy not to
10:41
do. It comprises a handbook with guides for reproducible research, project design, communication, collaboration, and ethical research. The project and community has grown significantly in the past two years as a result of absolutely brilliant community management. There are currently over 170 sub chapters across the five guides and more than 275 direct contributors on GitHub where
11:05
the resources are openly developed. Contributors include researchers, research software engineers, data stewards, librarians, community managers, and even you. Everyone is invited to get involved and there are already thousands of users. In order to ensure that community members are able to
11:21
participate irrespective of their previous experience of working with the open source or data science community, the project provides the resources, guidance, templates, training, and pathways to stay involved in the community. All these practices are recorded within the community handbook so that they can be adopted by other open source communities.
11:40
A similar project is the RDM kit. We have talked about how the increase in volume, complexity, and speed of data creation has required researchers to increasingly rely on computational support. The FAIR principles place specific emphasis on enhancing the ability of machines to automatically find and use data as well as supporting its reuse by other researchers, which
12:03
facilitates knowledge discovery and improves research transparency. However, researchers need relevant tools and guidance to better manage their data and make FAIR a reality. The Elixir Research Data Management Kit, or RDM kit for short, is another open source community project to develop an online guide containing good data management practices
12:23
applicable to research projects from the beginning to the end. It is also openly developed on GitHub and provides multiple pathways to contribute. Both the Turing Way and RDM kit projects solve skills gaps and demonstrate collaborative ways of working by actively nurturing their respective communities and valuing
12:42
and acknowledging the diverse range of contributions to the projects. These emerging roles and ways of working highlight how unfit current measures of success in research are, which are still highly focused on the publication outputs of the individual researcher. It means that there are a myriad of contributors to the research process who are undervalued and unrewarded
13:03
within the current evaluation system and have difficulty progressing even though they are crucial for research data, software, and project management. Recognizing the value of each skill set and providing progression opportunities is critical. However, the long-term career pathways needed for these new professional research support roles are only
13:21
emerging very slowly. Systemic solutions such as promoting collaborative ways of working and professionalizing these alternative but essential roles to support them, such as research software engineers, data stewards, and community managers, can lead to more sustainable, reproducible, and efficient research in a much healthier research culture.
Recommendations
Series of 18 media