How we work
Research that others can examine
Our research starts with a question and a plan for investigating it.
The approach depends on the task. Documentary research requires careful source assessment and context. Comparative work requires clear definitions. Statistical analysis requires an appropriate baseline and tests of the assumptions behind it.
Across these approaches, the aim is the same: to give readers enough information to understand how a result was reached and where its limits lie.
Sources and records
We record where material comes from, when it was accessed and which version was used. Where conditions allow, a copy or reference is preserved so later changes can be understood.
Sources vary in quality, purpose and completeness. Their limitations matter to the conclusions that can be drawn from them.
From information to analysis
Turning records into comparable data involves choices. Definitions, classifications, exclusions and ambiguous cases need to be documented.
For an index, this means explaining what qualifies for comparison and how it is assessed. For a statistical model, it means stating the baseline and examining whether a pattern persists under other reasonable assumptions.
For documentary research, it means showing how the sources support the account and where they leave questions open.
Interpreting a result
A repeated pattern may be worth examining. A measurable difference may justify further research. Neither, on its own, establishes cause or intent.
We distinguish recorded information, analytical results, possible explanations and allegations of wrongdoing. Formal legal or disciplinary findings belong to the bodies responsible for making them.
An analysis that finds no clear signal also has limits. Its result depends on what was measured and on the information available.
Publication and review
Before a research output is released, the sources, method, limitations and relevant disclosures must be documented. The publication process includes a right of reply where relevant, legal review and a recorded release decision.
These procedures are being finalised ahead of the first research publication. Describing the programme on this website does not mean that an individual output has passed those checks.
Access and correction
We intend to publish methods and supporting material as openly as conditions allow. Some source material may be subject to licensing, privacy or safety restrictions. Any such limits should be explained alongside the work.
Material errors will be corrected visibly. Changes to methods and results will be recorded so readers can distinguish versions and understand what changed.