What you will learn
- Write a precise version of a claim.
- Trace repeated reports to their original evidence.
- Choose a confidence level and an appropriate response.
Reduce the message to a checkable claim
A dramatic headline often mixes a factual assertion, an interpretation, and a recommendation. Separate them before searching. “A company introduced a new hiring test” is different from “the test is fair” and from “everyone should prepare for it.” Specify who, what, where, and when. A claim about one employer cannot establish a rule for an entire occupation. Save the wording and date so that you do not accidentally verify a weaker claim than the one that influenced you. If the claim cannot be stated clearly, that is already a reason to slow down before passing it on or spending money.
Trace the evidence rather than counting repetitions
Ten pages may repeat a single press release. That is one underlying source, not ten independent confirmations. Follow links until you find the original document, data, announcement, or recording when available. Examine what it actually says and what it leaves out. A product maker can reliably tell you what it announced, but its own announcement is not independent proof that the product outperforms competitors. A firsthand witness can describe a particular event but may misunderstand its cause. Match the source to the question, and distinguish expertise, proximity, incentives, and the quality of the evidence instead of relying on a familiar name alone.
Inspect method, scope, and missing context
For a numerical claim, ask who was counted, how they were selected, when the information was collected, and what comparison was made. For a quotation, read enough surrounding material to understand the speaker’s meaning. For an image, check whether its date and location match the claim attached to it. Cornell University Library’s evaluation resources offer further questions for judging sources. Your conclusion should remain proportionate: a small voluntary survey may show what its respondents said, while telling you little about everyone else. Missing details do not automatically make a claim false, but they limit what you can responsibly conclude.
Let confidence guide the next action
Useful categories include well supported, plausible but incomplete, disputed, and not yet verified. These are temporary judgments, not permanent labels for a person or publication. Seek an independent source that could reveal a mistake, especially when the claim matches what you already want to believe. The evidence needed should rise with the consequence of the decision. A low-cost curiosity can wait; an expensive course purchase deserves careful verification of terms and outcomes. When you share uncertain information, preserve its uncertainty. Saying that you have not verified a claim is more accurate than either endorsing it or declaring it false without evidence.
Preserve the difference between an outdated claim and a fabricated one. Information can have been accurate at publication and become unsuitable for today’s decision because circumstances or definitions changed.
Fictional case: Mei checks a training claim
Mei sees three posts saying a certificate guarantees interviews. Each links to a different blog, but all the blogs refer to the training seller’s announcement. She locates the original terms and finds that interview support means access to a group session, not a promise of employer interviews. She also notices that the advertised success stories do not state how many learners enrolled. Mei records what is established, what is missing, and which question she would need answered before paying. She does not accuse every graduate of lying. She rejects the unsupported guarantee while leaving room to evaluate the course’s actual teaching value.
Your practical assignment
- Choose one claim that could affect a decision.
- Write its precise scope and follow it to the earliest available source.
- Record the evidence, incentives, date, and one missing detail.
- Seek an independent check and assign a provisional confidence level.
Further reading
- Cornell University Library: Critical analysis of information sourcesUniversity library guidance for evaluating sources; the certificate example is fictional.