Education

In the final REF guidance, the devil is not in the detail

In the final REF guidance, the devil is not in the detail


One of the ambitions for Research Excellence Framework (REF) 2029 was to better recognise the people, cultures, and environments that make excellent research possible.

The proposal in June 2023 to include a new People, Culture, and Environment (PCE) element promised a significant step forward in how research excellence would be assessed, bringing recognition for the influence of employment practices, institutional leadership, research culture, and support for the wider research community. Following several years of development and testing, PCE was renamed and reshaped as Strategy, People, and Research Environment (SPRE) in December 2025. With the final SPRE guidance and panel criteria now published, we can see where those ambitions have landed.

The guidance and criteria retain more of the PCE work than we anticipated – institutions are consistently asked not simply to describe initiatives, but to explain how they work, who benefits, what evidence has been gathered, and what has been learned. But we have questions about whether these expectations remain equally visible in the quality descriptors that will ultimately determine assessment outcomes.

This matters because institutions optimise for what they believe assessors will reward, and anyone involved in REF will know that the criteria and quality descriptors quickly become shorthand for deciding which activities deserve investment, attention, and ultimately inclusion in a submission.

What’s survived?

The PCE Indicators Project, led by Technopolis and Vitae with sector partners, explored how research culture could be assessed in a meaningful and practical way. It considered responsible research, employment conditions, leadership, career development, inclusion, recognition and the wider range of people contributing to research. The resulting framework informed the PCE pilot involving 40 institutions and 115 unit-level submissions.

The pilot demonstrated that research culture can be meaningfully assessed, while identifying the importance of contextualisation, flexibility in evidence, and recognition of differing institutional circumstances. It emphasised looking beyond the existence of policies to the journey travelled, valuing downstream adoption, benefits, and evaluation.

The SPRE guidance retains more of this thinking than many expected. The templates embody a clear focus on how institutions operate rather than simply what initiatives were introduced. Institutions and units must include indicators of effectiveness or learning, explain how evidence has informed decisions, and provide balanced reflections on both success and limited progress.

The guidance also retains a broad understanding of the research community, including researchers, technicians and research-enabling staff, and explicitly recognises the importance of institutional context and resource constraints.

The learning cycle is present, but less explicit

Effectiveness, learning, and reflection appear repeatedly throughout the SPRE guidance, but the critical factor is whether they remain sufficiently visible in the criteria and quality descriptors that distinguish excellent environments from merely good ones.

The PCE pilot explored the progression from implementation, to impact, to reflection, and improvement. Participants argued that the highest levels of performance should reflect not only evidence that a policy existed and had been implemented, but also evidence that institutions had learned from experience and adapted their approaches.

That logic is much less explicit in the final quality descriptors than was recommended in the pilot report. While reflection has become somewhat more explicit in the assessment criteria, the quality descriptors are less clear about how evidence of implementation, impact, and learning distinguishes one level from another. Instead, excellence is defined through an approach that is “conducive to producing research of world-leading quality”, the exact wording that was considered problematic in the pilot.

This creates a potential disconnect. The templates emphasise the journey travelled, institutional learning, and evidence of effectiveness, but the quality descriptors don’t mirror this emphasis. Whether panels reconnect these components in practice may become one of the defining questions of SPRE assessment.

Whose learning counts?

Where this connection is retained, and statements are reflective, a further important question emerges: whose reflection are we reading?

An institution may regard an initiative as successful because it was implemented widely or achieved its intended objectives. Those affected by it may have a different perspective. Researchers might report limited time to engage with development opportunities. Technicians may experience implementation differently from academic staff. Professional services colleagues may identify unintended consequences. These diverging perspectives do not undermine institutional accounts. They strengthen them.

We hope institutions will draw on established evidence, including CEDARS (now the Research Culture Survey) to ground these accounts in the experiences of their research communities. The real test is what they do with that evidence: whether it changes their understanding of the problem and informs subsequent decisions.

Strategy and context

The final SPRE guidance places considerable emphasis on institutional strategy for research, engagement and impact. There are good reasons for this. Decisions about investment, staffing, infrastructure, and organisational priorities shape research culture and should be subject to assessment.

However, the PCE pilot revealed differing views on whether broad research strategies help or hinder assessment of research culture. Some argued that embedding culture within broader research strategies demonstrates institutional commitment. Others worried that culture-specific actions and learning could become obscured within larger strategic narratives.

SPRE appears to land firmly in the first camp. The challenge for institutions will be showing how evidence about people, culture, and environment influenced strategic decisions, rather than simply describing the strategy itself.

The guidance is stronger when it comes to context. Panels are explicitly asked to consider institutional size, type and available resources. In a period characterised by financial pressures, restructuring, and workforce challenges, this recognition is particularly important.

The devil may be in the quality descriptors

The final SPRE guidance on how to approach the statement asks institutions to explain what changed, how they know, what they learned, and how that learning informed future action. Yet the quality descriptors are less explicit about the role this evidence should play in distinguishing good from excellent environments.

Institutions will decide what is worth evidencing. If they conclude that reflection and learning are the path to higher ratings, they will feel emboldened to share honest accounts of limited progress, unintended consequences and course corrections. If they conclude that the quality descriptors primarily reward ambitious initiatives and compelling narratives, evidence of learning may be deprioritised.

The disconnect between template and descriptors therefore presents a risk that institutions see little reason to give much weight to learning cycles in their submissions.

For us, identifying problems, involving those affected, making informed decisions and evaluating the results remain hallmarks of a positive research culture – as well as of a REF SPRE exercise that is truly meaningful for institutions. We are delighted that the detailed guidance largely reflects that view. But whether the assessment process ultimately rewards it may depend less on the templates than on how panels interpret and apply the quality profiles.

In REF 2029, the devil may not be in the detail after all. It may be in the quality descriptors.



Source link