Separate reported facts from commentary

Label what each sentence is doing

Articles often combine several kinds of statements. A report may state that a meeting occurred, quote a participant’s explanation, describe a pattern in available figures, and then interpret what that pattern could mean. Those parts do not have equal status. As you read, label a sentence as a verifiable event or detail, an attributed claim, an interpretation, or a value judgment. Then ask whether the article shows where the factual details came from and makes interpretations recognizable as interpretations. This simple sorting lets you evaluate an argument without rejecting every part of an article that contains an opinion.

Consider a hypothetical story about a town adding a bus route. “The council voted on Tuesday” is a checkable event if a record is available. “A rider said the route made an appointment easier to reach” is an attributed personal account. “The route will transform access for everyone” is a broader prediction that needs more support than one account. A reader can appreciate the rider’s experience while still asking what information is available about frequency, coverage, and who can use the route. Keeping the categories separate makes the question more precise.

Check the evidence at the right scale

The type of support should fit the size of the claim. One example can show that something happened to one person, but it cannot establish how common the experience is. A poll, dataset, or document may answer a different question, depending on how it was collected and what it measures. Look for the date, scope, definitions, and limitations when they are relevant. If those details are not provided, describe the conclusion as uncertain rather than treating a vivid example or a confident phrase as a complete answer.

1. Choose a paragraph and label its sentences as event, attributed claim, interpretation, or opinion.
2. For one factual claim, identify what record, source, or method could check it.
3. Ask whether the evidence describes one case or supports a broader pattern.
4. Rewrite the paragraph’s conclusion in cautious language that matches the evidence you found.