The short version of our rule is this. A reader should be able to check the record under the argument.
What we treat as stronger evidence
We prefer primary sources when they exist and can answer the question well. Those include treaty texts, statutes, court rulings, and government designations. They also include official filings, original polling releases, and reports from the group making the claim.
When primary sources are not enough, we lean on solid secondary work. That means CRS reports, official diplomatic history, peer-reviewed work, and research groups that show their method.
News has a place too. It is strong for dated events and breaking stories. It should not be asked to carry more weight than it can bear.
Our evidence hierarchy
- Tier A: official filings, treaty texts, statutes, court rulings, government databases, and original reports
- Tier B: serious secondary work, CRS, FRUS, the State Department’s Office of the Historian, and similar research groups
- Tier C: trusted news for dated events, investigations, and updates when higher-tier work is missing
What a good research page lets you see
- Where the record ends
- Where we draw an inference
- Where the page makes a broader argument
- What is still in dispute
Numbers should be dated. Polling claims should name field dates, sample design, and wording. High-risk claims should lean on Tier A or Tier B when we can.
Outside standards we take seriously
The AAPOR Transparency Initiative is useful for polling work. It treats disclosure as part of credibility, not a side note.
The SPJ Code of Ethics keeps accuracy, attribution, and accountability in view. Those standards do not write our conclusions. They help keep us honest about how we get to them.