How we research
One model runs across the four centres. It is written here as the institution intends to work, and we publish it as current practice only to the extent that it is actually in place.
Agree the problem with the researcher, the practitioner, the operator, and the people affected by the work. Identify the decision that the result should inform. A question that informs no decision is unlikely to be worth the disruption.
Record existing practice, relevant conditions, access rights, resources, and the evidence that will be needed for comparison. This is the step most often skipped, and its absence is what makes later claims unfalsifiable.
Choose methods, responsibilities, consent and data arrangements, comparison conditions, and feasible measures of success. Proportionate means adequate to the question, neither more elaborate nor more casual than it warrants.
Run the activity within the actual setting. Record changes and interruptions as they happen, and assess both participant experience and operational burden — the second is what determines whether anything continues.
Distinguish observations from causal claims. Report limitations, unexpected effects, and findings that do not support the original expectation.
Agree ownership, maintenance, skills, and the conditions for transfer. The research question stays within the institutional agenda even when a particular project changes shape.
The institution aims to do more than provide accommodation. Its intended contribution is to take part in formulating the question, organise co-design, operate agreed trials, steward local records, support dissemination through practice, and assess what should continue.
Each of those responsibilities needs a named person and a deliverable. Where scientific leadership is held by a university or a specialist institute, that is the normal and correct arrangement.
Every study should identify nine things, and a study that cannot is not ready to begin:
| Element | What it establishes |
|---|---|
| Question | The decision the result is meant to inform |
| Setting | Where the work happened and under what conditions |
| People responsible | Who led, who participated, who holds the data |
| Method | What was done, in enough detail to be repeated |
| Baseline | The documented starting point for any comparison |
| Intervention | What changed, when, and at what cost |
| Results | What was observed, separated from what is inferred |
| Limitations | What the study cannot show, and why |
| Reuse conditions | What others may do with the material, and on what terms |
Different disciplines need different standards of evidence, and applying one set of outcome indicators across all of them produces false comparisons. A source edition, a garment trial, a wellbeing feasibility study, and a water-retention experiment are each judged by the standards of their own field.
Where cultural knowledge is involved, we document its sources, the relevant rights holders and the permissions granted for the proposed use. Any limits and withdrawal arrangements are agreed explicitly.
Personal data, community rights, and commercially or culturally sensitive information are protected. Participants are told what is collected, why, how long it is held, and how to withdraw. Outputs are shared where sharing is appropriate and where those involved have agreed to it.
Ethical review, research accreditation and clinical governance belong with the university, institute or healthcare partner leading a study. That is the normal and correct arrangement for work of this kind, and it is how a place-based collaboration is properly constituted.
What the Abbey brings alongside it is the setting, the practitioners, the operational record, and continuity of attention across years rather than project cycles.