Eight ways universities can foster an open research culture now

In 2021, my colleagues Parveen Yaqoob and Robert Darby outlined eight ways to make university research cultures more open. Since the original article was published, the challenge of open research has moved from recognition towards implementation. 

Universities need to make transparent, collaborative practices part of everyday research. The Research Excellence Framework (REF) 2029 reinforces this direction. Its Strategy, People and Research Environment guidance recognises inclusive, supportive and collaborative research cultures. 

But institutions should build these conditions into everyday work, well beyond an assessment deadline, and make it business as usual. Sharing methods, data and software enables scrutiny, and a culture that encourages questions, acknowledges uncertainty and supports correction makes that scrutiny useful.

These eight updated actions connect the two.

1. Join a reproducibility network and contribute

The UK Reproducibility Network connects grassroots communities, institutional leaders and research organisations. Its 2024–25 annual report records local network leads at about 80 UK institutions, and over 40 as institutional members, connecting senior management.

• Open research by principle, not by checklist
• A beginner’s guide to open science
• Campus Talks: What is open access?

Make participation useful: give representatives formal time to contribute, adapt shared training and bring researchers’ concerns into institutional discussions. Share unsuccessful approaches alongside achievements, so others can learn from them.

Membership should create an ongoing exchange between local experience and collective expertise. Identify which institutional decisions exchange can inform.

2. Give a working group authority to deliver

A dedicated working group gives open research a visible face and coordinates implementation. It needs senior sponsorship, a clear remit and access to people who can approve resources or remove obstacles.

Include researchers across disciplines and career stages alongside library staff, research software engineers, technicians, research integrity specialists and colleagues responsible for staff development and promotion. Students should have a voice, too.

Give the group practical questions to answer. Where do researchers encounter obstacles? Who can remove them? Which commitments lack resources? Members need recognised time to undertake this work, and the group should report what has changed as a result.

3. Create a plan that survives changes of personnel

People leave and roles change. An institutionally owned action plan preserves commitments and helps successors understand decisions.

Set a few priorities with named owners, resources and review dates. Connect them to school plans and existing services. Document handovers.

Measure whether activities help: attendance shows who reached a workshop; follow-up focus groups reveal whether they applied the learning. Ask what prevented change and adjust support accordingly. Review the plan regularly, including actions that should stop, and objectives that need to be adapted to evolving contexts.

4. Build research integrity and responsible AI use into everyday support

A statement of commitment can clarify expectations, and provide a solid anchor. Those expectations become credible when staff and students can obtain practical help throughout the life cycle of their research project. 

Generative artificial intelligence makes this especially pressing. UKCORI’s April 2026 update reports uneven institutional readiness to support AI use in research. The UK Research Integrity Office’s “Embracing AI with integrity” offers a starting resource.

Use disciplinary examples to practise checking sources, validating outputs, handling sensitive material and documenting consequential AI use. Include supervisors and experienced staff alongside students.

Provide trusted routes for discussing mistakes and concerns. UKCORI is developing recommendations for governance beyond April 2027; institutions can contribute while improving their own arrangements now.

5. Celebrate practices that others can learn from

Recognition need not involve selecting a winner. Invite colleagues to demonstrate a useful workflow, explain a difficult sharing decision or describe how they corrected a problem.

Use short showcases and case studies that explain the time, support and compromises involved. Include templates or materials others can adapt.

Credit technical and professional services contributions alongside academic work. Make room for modest improvements and unfinished learning, so participation feels achievable for colleagues beginning to change their practice.

6. Give local communities the support to act

Champions’ effectiveness depends on institutional structure and local backing.

At my university, our first approach struggled because schools often saw champions as representatives of a university initiative and provided limited support. Many were early career researchers being asked to influence colleagues without sufficient authority.

Smaller institutions may benefit from one community; larger ones may need several connected groups. The Open Science Community Starter Kit and the Open Science Learning Gate can help kickstart initiatives.

Secure school leadership support, workload recognition and access to relevant meetings. Share responsibility across career stages.

7. Invest in software expertise and infrastructure

AI-generated code increases the importance of checking whether an analysis is reliable. Provide access to research software engineers and training in testing, version control and documentation. The Society of Research Software Engineering supports this community and its development.

Involve specialists when projects are planned. Budget for maintenance and preservation, and connect software support with data stewardship and repositories.

Where sharing is restricted, help researchers explain why and provide appropriate descriptions or access arrangements. Make support easy to find across disciplines.

8. Make open practices count in career decisions

Recruitment, promotion, assessment and reward must recognise open research if it is to become routine.

Ask people to explain contributions to transparency, reuse and collaboration, including software maintenance, data curation and support for colleagues. UKRN’s recognition and reward toolkit provides practical implementation guidance.

Train panels to judge contributions fairly across disciplines, roles and career stages. Avoid replacing publication counts with simplistic counts of shared outputs.

Review actual decisions to establish whether revised criteria influence outcomes. The next test for universities is whether someone practising open research can identify the time, support and recognition that makes it sustainable.

Etienne Roesch is professor of applied statistics and cognitive science at the University of Reading and chair of the supervisor board at the UK Reproducibility Network. ChatGPT was used to brainstorm and collate ideas.

If you would like advice and insight from academics and university staff delivered direct to your inbox each week, sign up for the Campus newsletter.