Methodology
How we collect, assess and report.
The methods and level of assessment depend on the question and agreed scope. We explain where information comes from, what supports a finding and what remains unresolved.
Collection
We obtain information through local-language research, document retrieval, discreet source inquiries and site observations, as the assignment requires. We record the origin, date and relevant context of the material collected.
Access and availability vary by subject and location. We agree the collection requirements and practical limits before work begins.
Assessment
We assess source reliability, access to the subject and the credibility of individual claims separately. Where the agreed scope includes corroboration, we compare accounts and supporting material, examine conflicting evidence and identify unresolved gaps.
A source’s position does not establish that every statement is accurate. We distinguish direct knowledge, secondhand reporting, opinion and analytical interpretation.
Reporting
We provide the requested documents, source reporting or analytical assessment. We distinguish documented facts, attributed accounts and Cascade’s judgments, explaining the supporting evidence and relevant limitations.
We identify what has been corroborated, what remains unverified and which questions the work could not resolve.
What we don’t claim
- Certainty we can’t evidence
- Some questions end at “unverified,” and we say so.
- Total coverage
- Some records are closed. We tell you where the picture stops.
- Guaranteed outcomes
- We assess risks and possible developments. We do not guarantee outcomes.
- Permanence
- Our findings reflect the evidence available at the time of research. Later developments may change the assessment.
- Inference as observation
- Where we provide an assessment, we distinguish what the evidence establishes, what sources report and what we infer.
How we grade evidenceThe reliability and credibility scales, documents and records, and how findings are presented.
Two separate ratings
For sources of reporting, we use the Admiralty scale, a long-standing intelligence standard. The letter rates the source’s reliability. The number rates the credibility of the specific information. A reliable source can still pass on something unconfirmed, so the two are shown separately.
Sources without a track record
Sources with no prior reporting history with Cascade are always rated F — Cannot be judged. Their access to the subject and the corroboration of individual claims are assessed separately.
An F rating reflects the absence of a reporting history with Cascade. It does not indicate poor access or false reporting, and it says nothing about the credibility of the information itself.
Documents and records
Retrieved documents and records are identified by their origin, date and relevant context. Where a deliverable includes a graded source register, documentary and official sources may also receive evidence grades. The assessment is limited to what the document establishes: an official statement records what was said, but does not by itself verify the underlying claim.
| A | Completely reliable |
|---|---|
| B | Usually reliable |
| C | Fairly reliable |
| D | Not usually reliable |
| E | Unreliable |
| F | Cannot be judged |
| 1 | Confirmed |
|---|---|
| 2 | Probably true |
| 3 | Possibly true |
| 4 | Doubtful |
| 5 | Improbable |
| 6 | Cannot be judged |
How to read a grade
- A1
- A completely reliable source, information confirmed.
- C3
- A fairly reliable source, information possibly true.
- F3
- A source with no reporting history with Cascade, information possibly true.
How findings are presented
Where assessment forms part of the agreed deliverable, we explain the evidence supporting our findings and the limits of the conclusions drawn. We distinguish documented facts, attributed source accounts and analytical judgments, identifying what has been corroborated and what remains unverified.
Discuss an assignment.
Tell us the subject, the countries involved and the information needed.