Corroboration

Also known as: triangulation, independent confirmation, cross-checking, 交叉印证.

Confirming a claim with evidence independent of the original source. Reports that trace back to the same origin count as one.

Synthetic dataExamples on this page use synthetic data created for illustration.

Definition

Corroboration is confirming a claim with evidence that is independent of the original source. Two reports that trace back to the same origin count as one, however many outlets repeat them; real confirmation comes from different routes to the same fact.

In plain terms

The test is independence. If ten articles all rely on one press release, you have one source repeated ten times. If one article quotes a company filing and another quotes a regulator’s database, you have two routes to the fact, and agreement between them means much more.

In practice, trace each claim back to where it started (see Data Lineage for the same idea applied to numbers), count the independent origins, note their incentives, and record what disagrees. Corroboration raises confidence; it does not prove a claim.

Example

Five outlets report that a company will cut 500 jobs. Tracing each back:

OutletCites
1News agency A
2News agency A
3News agency A
4News agency A
5The company’s own filing F

Five reports, two origins. And if agency A’s story was itself based on filing F, there is really one origin: the filing. Counting outlets would have given five confirmations where there is one.

Common misunderstandings

  • “Five reports are five confirmations.” They are five confirmations only if their origins are independent.
  • “Corroborated means true.” It means better supported; independent sources can still share the same blind spot.
  • “Sources that agree with each other corroborate each other.” Not if they share an interest or an origin.

References

  1. Craig Silverman (ed.). Verification Handbook . European Journalism Centre, 2014.
  2. Bill Kovach, Tom Rosenstiel. The Elements of Journalism . Crown, 2014. ISBN 978-0804136785.