Sample investigation · Research and education
A research team needs a defensible number. A ministry and a trade body publish different totals, and the press picks whichever it met first.
Illustrative An invented example. No real client, event or organisation, and the figures are samples. Sample record 7A-D-000655, SAMPLE · 03 SEP 2026 · 11:00 local.
Reports ingested
over six days
Independent origins
ministry and trade body
Primary documents
both tables linked
Gap between totals
explained in part by method
01 Headline
Two sources give two totals. The gap is partly method: one counts planted area, the other counts delivered grain.
23Reports ingested
The headline gives a direction and no number. The team starts by asking which number each source means.
Dispatch
SAMPLE · 03 SEP 2026 · 11:00 local
Two sources give two totals. The gap is partly method: one counts planted area, the other counts delivered grain.
The question
Which figure can the study cite, and what has to be stated alongside it?
02 Sources
An origin is the first to report something. A repeat only passes it on. Counting origins shows how many separate voices there really are.
3Origins among 5 sources shown
Both totals come from primary tables, and the press mostly repeats one or the other. The ledger shows the specialist weekly as the only item that sets both side by side.
Source ledger
| Source | Role | Score |
|---|---|---|
| Agriculture ministry table | Primary | 82 |
| Grain trade body table | Primary | 76 |
| Agency wire, quotes ministry | Repeat | 74 |
| Business daily, quotes trade body | Repeat | 66 |
| Specialist weekly, compares both | Origin | 80 |
3 of 5 shown report on their own or are the document itself
03 Six-axis audit
The audit asks who, what, when, where, why and how. Each question gets its own confidence from 0 to 100. A weak answer is marked contested instead of being averaged away.
How is the axis that explains the gap: area survey against delivery records. What and where are marked contested because the size and the regional split depend on that method.
Six-axis audit
Who
Ministry and trade body, each for its own total.
What
A fall in harvest. The size depends on the method.
When
The current season, to the end of August.
Where
National. Regional splits differ between sources.
Why
Dry spell, per both. Weighting of causes differs.
How
Area survey against delivery records.
04 Impact
The impact matrix does not forecast. It lists where the claim, if it holds, would touch each of nine areas, and how strongly.
Environmental and supply chain readings are highest. The matrix is not a forecast. It shows the pillars where a fall of this size would be felt first, and which findings rest on the contested axes.
Impact matrix
05 Dossier
Season harvest: two totals and the method behind each. The file is fixed as it stood on 03 SEP 2026 · 13:30 local. A later correction would be a new entry.
The team a with both tables, the method note and the two totals. The paper can cite one figure and state the other, with the record attached.
Dossier
Sealed 03 SEP 2026 · 13:30 local. Later changes become new entries.
The study cites the delivery-based total and states the survey-based total with its source.
What happened next. The paper cites one figure, names the other and links the , so a reviewer can check both.
Next
Open the live demo, or ask for access and we will walk through a case from your desk.