Barcamp Open Science 2022: Connecting and Strengthening the Communities!

Open Science activities take place at many institutions at the level of individuals or working groups, without there being any exchange between them.

In this session we discussed the question of what means can be used to build a community of those interested in Open Science: What basic requirements are needed? What best practice examples are there? Ideas can be found, for example, in this “Open Science Community Starter Kit”.

Die The Four Sages of Developing an Open Science Community from the „Open Science Community Starter Kit “ (CC BY NC SA 4.0)

There is a perception among many that there is a gap between the existing information offered by central institutions such as libraries and research services and the actual implementer community. These central bodies can take on a coordinating role to promote existing activities and network participating groups. It is important to respect the specialisation within the Open Science community. Grassroots initiatives often form in their field due to specific needs in the professional community.

Key persons such as data stewards, who are in direct contact with researchers, can establish contacts for stronger networking among Open Science actors. The communication of Open Science principles should not be too abstract. Incentives and the demonstration of concrete advantages can increase the motivation to use Open Science practices.

Conclusion: If a central institution from the research ecosystem wants to establish an Open Science community, it would do well to focus, for example, on promoting existing grassroots initiatives and to offer concrete, directly applicable Open Science tools.

In this session all 22 participants introduced themselves and presented a successful (or not so successful!) case study from their institution.

Opportunities for Open Science

A wide variety of examples of improving awareness or rewarding Open Research practices were shared: Several universities have policies in place on Research Data or Open Access. These can be used to refer researchers to and are especially helpful when combined with personal success stories. Some universities offer (small) grants to support Open Science practices (Nanyang Technological University Singapore, University of Mannheim, German). Several universities offer trainings to improve Open Science practices, or support staff who can help.

Offering recommendations or tools that facilitate researchers to open up their workflows are welcome. Bottom-up communities or grassroots initiatives are important drivers for change.

Conferences, such as the Scholarship Values Summit, or blogs could be a way to increase awareness about Open Science (ZBW Blog on Open Science). You can also share your institute’s progress on Open Science practices via a dashboard, an example is the Charité Dashboard on Responsible Research.

Challenges for Open Science

On the other hand, some challenges were also mentioned: For example, Open Science is not prioritised as the current research evaluation system is still very focused on traditional research impact metrics. It can also be difficult to enthuse researchers to attend events. It works better to meet them where they are.

Not everyone is aware of all the different aspects of Open Science (sometimes it is equated with Open Access) and it can also be quite overwhelming. It may be helpful to use different terms such as research integrity or sustainable science to engage people more successfully with Open Science practices. More training is also needed.

There is no one-size-fits-all solution! If new tools are offered to researchers, they should ideally be robust and simplify existing workflows without causing additional problems.

Conclusion: Our main conclusions from the session were that we have a lot of experts and successful case studies to learn from. It is also important to have enthusiastic people who can push for progress in the departments and institutes!

There were ten participants in this session from institutions in South Africa, Germany, Spain, Luxembourg and Canada.

Four key points that arose:

1. One of the first things that came up in dialogue was that Open Science is a very large umbrella that contains a LOT of pieces/separate things within it. Because there are so many moving parts in this giant ecosystem, it is hard to get started in offering supports, and some areas get a lot less attention than others. Open Access and Open Data seem to consistently be flagged first as the areas that generate a lot of attention/support while Open Software and even Citizen Science received a lot less attention from libraries.

2. Come to us versus go to them: Another point of conversation was whether or not the researchers are coming to us (as a library) to get support for their own Open Science endeavours. It was consistently noted that they are not generally thinking about the library when they are thinking, e.g., about research data or Open Access publishing. The library is not on their radar as a natural place to find this type of support/help until they have experienced it for themselves and realise the library might offer supports in these areas.

From this starting point, the conversation morphed to focus on the educational aspect of what libraries offer – i.e. making information available. But it was flagged that we often have a bubble where the information is located that is not often browsed. So the community is a key player in getting the conversation started, particularly as part of everyday research life. This way, the library can be better integrated into the regular flow of research activities when information/help is needed.

3. The value of face-to-face engagement: People discussed the need to identify and work with the “cheerleaders” to get an active word-of-mouth network going to educate more university staff and students about Open Science (rather than relying on Libguides and webpages to do so more passively). Libraries could be more proactive and work more closely with the scientific community to co-create Open Science related products. Provision of information is something we do well, but we often spend less time on personal interactions and more on providing things digitally. Some of the attendees felt this might be detrimental to really understanding the needs of our faculty. More time and energy should be spend on understanding the specific needs of scientists and shaping the scientific communication system rather than reacting to whatever comes our way.

4. The role of libraries as a connecting element: The library is uniquely placed to see across subject disciplines and serve in the role of connector. In this way, it can help facilitate collaborations/build partnerships across other units of the organisation and assist in enabling the exchange of knowledge between people. It was suggested that libraries should be more outgoing in what they (can) do and get more involved in the dialogue with researchers. One point that was debated is the need for the library to acknowledge that it is not and cannot really be a neutral space – certainly not if Open Science is to be encouraged rather than just supported.

Whether journal article, book chapter, data set or sample – these results of science and research must be made openly accessible in an increasingly digital scientific landscape, and at the same time made unambiguously and permanently findable. This should support the exchange of information within science from “closed” to “open” science and promote the transfer of findings to society.

Persistent identifiers (PIDs) play a central role here. They ensure that scientific resources can be cited and referenced. Once assigned, the PID always remains the same, even if the name or URL of an information object changes.

The participants in the spontaneous barcamp session all agreed on this central importance of PIDs for the digital science landscape. All of them were familiar with the principle of PIDs and have contact points in their daily work, especially with DOIs and ORCID iDs (Open Researcher and Contributor iD). In addition to the enormous potential of PIDs, however, the participants also saw challenges in their use and establishment. It became clear that there are still technical as well as ethical and data protection issues to consider.

There was consensus that these questions must be accompanied by a broad education on PIDs, their purpose and how they work; among the scientific staff of research institutions as well as among researchers. Websites tailored to the topic from ORCID DE (German) or Forschungsdaten.org (German) offer a good introduction.

Translating scholarly works opens science for more contributors (who do important work, but are not proficient in writing English), avoids double work and it opens the fruits of science to larger communities. Translated scientific articles open science to science enthusiasts, activists, advisors, trainers, consultants, architects, doctors, journalists, planners, administrators, technicians and scientists. Such a lower barrier to participating in science is especially important on topics such as climate change, environment, agriculture and health.

In this session we discussed why translations are important, tools that could help making and finding translations and foreign language works. An interesting thought was that currently blogs are important for finding foreign scientific articles, which illustrates how much harder it is to find such works and suggests allies to work with. The difficulty of finding foreign works emphasises the importance of at least translating titles and abstracts. Search engines that include automatically translated keywords can also help discovery.

The slides of the session “Translating scholarly articles opens science” can be found here.

In this session we discussed approaches and tools to collaborate on scientific data openly. The starting point of the discussion was the assessment that publishing scientific data openly is already quite well supported and works smoothly thanks to platforms like Zenodo. In contrast, open pre-publication collaboration is difficult because the available platforms impose restrictions, either on the size of the datasets or on the research area supported. Self-hosting a data collaboration platform like gin – Modern Research Data Management for Neuroscience is one solution, but usually not feasible for individual researchers or working groups.

We also talked briefly about experiences with open pre-publication collaboration. Experiences are limited so far, but fruitful collaboration can establish when the datasets in question are useful to a broader group of scientists and contribution is easy and quick. Furthermore, adapting data workflows so that intermediate results and workflows are openly accessible also has benefits for reproducibility and data organisation in general.