Institutions state values and then reward something else — usually output counts.
Researchers respond to the rewards, not the statements, which is entirely rational. That gap is where research culture is actually determined.
Where the gap shows
- Stating that quality matters while assessing on counts.
- Valuing collaboration while assessing individuals.
- Valuing mentoring while counting only publications.
- Valuing openness while rewarding venue prestige.
- People follow the assessment, and they are right to.
What count-based pressure produces
- Work divided into smaller units.
- Reluctance to attempt anything slow or risky.
- Demand for services that produce publications — this connection is direct and documented.
- Neglect of unmeasured contributions until nobody does them.
- These are predictable responses, not moral failures.
What actually changes culture
- Changing what is assessed — everything else is secondary.
- Allocating time for the work you say you value.
- Senior people modelling the behaviour, including admitting error.
- Making it safe to report problems.
- Recognising contribution breadth in promotion decisions.
Integrity is mostly structural
- Training in good practice, provided before it is needed.
- Clear procedures for raising concerns.
- Protection for those who raise them in good faith.
- Realistic expectations that do not create pressure to cut corners.
- Time to do work properly.
- Most integrity failures are pressure responses, not character defects.
For emerging systems specifically
- Imported assessment criteria often fit badly and produce the wrong incentives.
- Rapid expansion without supervision capacity creates pressure.
- Building culture is slower than building output and matters more.
- Reputational damage from integrity failures is collective and hard to reverse.
What individuals can do
- Model the practice, particularly if senior.
- Give credit accurately.
- Report your own errors openly.
- Support people who raise concerns.
- Culture is the accumulation of what people actually do.
One thing worth remembering
If you want to know an institution's research culture, look at what it assesses, not at what it states.
The assessment criteria are the values in practice — and changing them is the only intervention that reliably changes behaviour.
Câu hỏi thường gặp
How is research culture actually set?
By what gets rewarded rather than what gets stated — researchers respond to the assessment, which is entirely rational.
What does count-based pressure produce?
Work divided into smaller units, reluctance to attempt slow or risky work, demand for services producing publications, and neglect of unmeasured contributions.
What actually changes culture?
Changing what is assessed, allocating time for work you say you value, senior people modelling behaviour including admitting error, and recognising contribution breadth.
Why is integrity mostly structural?
Because most integrity failures are pressure responses rather than character defects — requiring training, clear procedures, protection for reporters, and realistic expectations.
How can you tell an institution's real culture?
By what it assesses rather than what it states — the assessment criteria are the values in practice.
