The first phase of research culture reform has been about visibility. For Alys Kay, the next challenge is about ensuring that evidence is not just documented, but actually leads to changes in decisions

Across UK higher education, research culture has become highly measurable. Universities now hold more data than ever on staff experience, career progression, inclusion, workload and collaboration. Research culture is monitored through staff surveys, Athena Swan submissions, REF People, Culture and Environment narratives, internal culture programmes, and workforce analytics.
Yet a question is emerging: why does so much insight about research culture fail to translate into visible, day-to-day change in how research environments actually operate? The sector is not short of research culture evidence. It is short of ways to use it.
By translation, I do not mean simplifying insight into slogans. I mean the practical work of turning fragmented evidence into decision rules, prompts and accountabilities that shape what leaders actually do: how workload is set, how authorship is handled, how people are recruited and promoted, which teams are resourced, and what trade-offs are treated as acceptable.
At the moment, universities are generating insight faster than they can use it.
For most academic leaders, culture evidence arrives as a request for attention in an already overloaded decision environment. Unless insight is translated into the language, timing and risk frameworks leaders already use, it is likely to sit alongside decision-making rather than shaping it. For many leaders, culture evidence currently arrives as an additional reporting responsibility, rather than as something that makes core decisions easier to take with confidence.
Research culture evidence sits across human resources, equality, diversity and inclusion, researcher development, research finance and research culture programmes. Each area generates useful insight, but each operates with different evidence logics, reporting cycles and definitions of success.
In practice, this often produces a familiar institutional experience. The same issue shows up in multiple places at once: staff survey comments about workload, equality submissions highlighting progression bottlenecks, exit interview themes about burnout, culture programme feedback about psychological safety. The insight is not missing. It is simply not held together.
Leaders, meanwhile, receive multiple dashboards and narratives that are not aligned to the decisions they are about to make. Culture evidence arrives as commentary alongside performance evidence, not as an input that changes the decision itself. It is very hard to act on “everything, everywhere, all at once” – particularly when accountability is split across committees and functions.
This is the structural mechanism by which culture becomes reportable without becoming changeable. Fragmentation means no single picture of what is happening. No single picture means weak ownership and unclear responsibility. Weak ownership means insight does not enter decision cycles in time. When insight arrives late, it is logged, discussed and often agreed with, but it does not change the decision that has already been made.
As research environment expectations evolve, the focus is increasingly shifting from demonstrating that culture has been measured to demonstrating how culture evidence shapes incentives, behaviours and ultimately research outcomes.
One risk in the current debate is treating research culture as something that sits alongside research activity. Culture is produced through research activity, via everyday decisions.
Research culture is enacted through decisions about workload protection, recruitment, progression, authorship, line management, resource allocation and role recognition. These decisions are made in leadership environments shaped by funding pressure, performance metrics, student demand and institutional strategy, and they often happen far away from formal culture governance structures.
If culture evidence is not translated into the decision frameworks leaders use, culture work becomes a parallel conversation: something an institution can evidence without necessarily changing how it behaves.
It is also why the sector’s recurring focus on better metrics only takes us so far. In many institutions, the barrier is not measurement sophistication. It is that insight is not reaching the points where power is exercised.
This is often where institutions feel the gap most clearly: they know something is wrong, but cannot yet see which decisions are producing it.
As culture evidence becomes more distributed, a different capability comes into view: the ability to integrate, interpret and embed insight across systems.
This is not mainly a technology problem. It is an institutional design problem: who has the remit, legitimacy and skill to convene evidence across functions and convert it into something leaders can act on without it being dismissed as “soft”, “nice to have” or “someone else’s job”.
Qualitative insight is particularly important here. Qualitative evidence is often treated as a way to add voice to a report. Its more important function is diagnostic and design-led: it reveals how decisions are actually experienced, where incentives are misaligned, and which rules and routines produce predictable harms. It helps determine not only what is happening, but what should be measured and governed because it is driving behaviour.
In other words, qualitative insight is not simply about describing culture. It is how institutions locate the decision points that produce culture.
This shift reflects a broader move across research systems towards understanding how evidence, incentives and decision structures interact to shape behaviour across teams, disciplines and career stages.
The first phase of research culture reform has been about visibility: establishing that culture matters, creating accountability, generating evidence and building narratives. That work has value. It has made previously invisible experiences harder to ignore.
The next phase is about decision maturity: ensuring culture evidence changes decisions, not just documents them.
A practical way to think about this is as a translation stack: integrate insight across evidence systems, interpret what that combined picture means for risk, equity, sustainability and performance, and embed it into decision routines so insight arrives before decisions, not after.
Done well, translation is how institutions turn culture evidence into better research performance and more sustainable research careers. If institutions cannot make this shift, research culture will remain highly reportable and only intermittently changeable. Universities will get better at evidencing activity, while staff experience continued misalignment between what is said and what is rewarded. Leaders will experience initiative fatigue. Staff will experience consultation fatigue. And culture work will start to look, to many, like theatre.
The sector is not short of evidence. The challenge now is ensuring that evidence shapes the decisions that shape research itself.

Andy Mycock | Comment | 16/11/26

Matt Offord | Comment | 13/11/26

Hugh Jones | Comment | 13/11/26

Shahenda Shehata | Comment | 12/11/26

Debbie McVitty | Comment | 11/11/26

Carolyne Smith | Comment | 11/11/26