How we document methods in every report
· Dr Helena Marsh
Every report we publish carries a methods appendix. It is not a formality. It is the part of the document that lets a reader decide how much weight a finding can bear.
What the appendix contains
At minimum, each study states its data sources, the period covered, the unit of observation and the sampling approach. Where we run models, we record the specification, the assumptions and the sensitivity checks that would change the result.
- Data provenance — where each series comes from, and how it was cleaned.
- Coverage — the countries, sectors and years in scope, and the known gaps.
- Uncertainty — confidence intervals, or a plain statement of the limits when intervals are not meaningful.
Why we hold to it
Independent analysis is only useful if it can be checked. A number without a method is an assertion. We would rather publish a narrower claim that a reader can verify than a broad one they have to take on trust.
If a finding cannot survive its own methods note, it is not ready to publish.
The appendix also disciplines our own work. Writing down the assumptions before the analysis is finished tends to surface the weak ones early, while there is still time to test them.