Cross-Checking and Triangulation
Imagine three witnesses to a car accident who all independently describe the same car as red. That convergence of independent accounts gives you confidence. Now imagine that one witness says red, a second says they were not paying attention, and a third says blue. The disagreement tells you something important: someone is wrong, their memory is unreliable, or the lighting made the color ambiguous. Using multiple independent sources and looking for convergence is called triangulation — the same principle geographers use to pinpoint a location from multiple reference points.
Triangulation is the practice of confirming a claim by finding at least two or three independent sources that arrived at the same conclusion through their own research or reporting — not by copying each other. When independent investigations converge on the same answer, confidence in that answer rises substantially.
Independence Is the Key Word
A critical requirement for triangulation to work is that sources must be genuinely independent. Three articles that all copy from the same original wire service report are not three independent sources — they are one source wearing three different outfits. This is called citation chaining: source B cites source A, source C cites source B, and a reader finds three apparent sources that are actually all tracing back to the same single original. To find truly independent sources, you need sources that conducted their own reporting or research, not sources that summarized each other.
In practice, checking for independence means asking: did this outlet conduct its own interview, experiment, or investigation? Or did it summarize another outlet's story? News wire services like the Associated Press, Reuters, and Agence France-Presse produce original reporting that many outlets then republish. Finding multiple outlets each citing the same AP story does not give you independent confirmation — it gives you AP's reporting appearing in multiple places. For true independence, find sources that went to primary sources on their own.
The Power of Disagreement
Triangulation is not only valuable when sources agree — it is also valuable when they disagree. If two credible, independent sources describe the same event differently, that disagreement is informative. It signals genuine uncertainty, a contested interpretation, or a factual question that has not been settled. Finding multiple credible sources arguing a scientific question is still open, for instance, is itself useful information that should change your confidence level in any single definitive-sounding answer you previously encountered.
There is no magic number, but a practical standard used by investigative journalists is: a claim is worth reporting when at least two independent sources corroborate it. For claims with high stakes — medical, legal, scientific — more is better. For well-established facts covered by major institutions over many years, one high-credibility primary source may be sufficient. Match your verification effort to the stakes involved.
Match each cross-checking concept to its precise meaning.
Terms
Definitions
Drag terms onto their definitions, or click a term then click a definition to match.
Cross-Checking AI Outputs
Triangulation is especially important when working with AI-generated content, for a subtle reason: asking the same AI system the same question twice, or asking two different AI systems, does not give you independent confirmation. Both systems were trained on overlapping data and share the same underlying biases and knowledge gaps. If both systems hallucinate the same fabricated fact, you have two sources pointing to the same error — which looks like confirmation but is actually common-source contamination. For AI outputs that matter, triangulation must involve non-AI sources: primary documents, peer-reviewed literature, official databases, or original reporting.
You find three news websites all reporting the same surprising statistic. All three articles appear to be citing the same original press release. Does this count as triangulation — independent confirmation from three sources?
Why does asking two different AI chatbots the same question NOT provide independent triangulation the way asking two human journalists does?
Trace the Source Chain
- Step 1: Find a statistic or specific claim in a news article that seems striking or surprising.
- Step 2: Look for the citation or link the article uses. Follow it. Does it go to a primary source (original study, official report, direct interview) or to another article?
- Step 3: If it goes to another article, follow that link too. Keep going until you reach the original source or realize there is no original source cited.
- Step 4: Map out the chain: Article A cites Article B, which cites Report C, which cites... Draw this as a simple diagram.
- Step 5: Evaluate: Does the primary source actually support the claim as stated? Has anything been lost, exaggerated, or distorted along the chain?
- Step 6: Write two sentences about what you found and whether the original claim holds up.