Correlation: describe the relationship without adding a cause

When an article says two things are linked, try to identify the two measurements and the wording used to connect them. Then check whether a retelling quietly replaces “associated with” with “makes,” “prevents,” or “leads to.” You can notice that wording change without calculating a correlation yourself.

The book stall and the longer queue

In an invented neighborhood report, event days with more boxes of secondhand books also have longer entry queues. A message in the family chat says, “Book boxes cause queues.”

Jules writes the measurements on separate lines: boxes brought to the stall, and people waiting at the entrance. Their mother wonders whether popular event days prompted organizers to bring more books. Their brother asks whether the days with a guest performer had both extra stock and more visitors.

Those are possible explanations to investigate, not facts the family has established. Jules rewrites the message: “The report describes a relationship between book-box counts and queue length. I have not checked an explanation for it.” The family has not measured any real events; this is a fictional exercise in describing a claim.

Keep a headline close to its measurements

For a real source, you could ask:

• What exactly are the two quantities or characteristics?
• What people, objects, places, or time periods were measured?
• Does the article report a numerical relationship, or merely use “linked” informally?
• What explanation do the authors offer, and how cautiously do they state it?
• What additional causal language appears in the headline or social post?

Try writing the narrowest sentence you can support from the passage you have read. Do not add an explanation simply to make the sentence sound complete. “These were measured together; the reason is unclear to me” can be an honest stopping point.

Related terms point to different questions

Correlation keeps your attention on the relationship being described. Causation asks whether one thing makes a difference to another. Confounding asks about a possible shared cause that complicates the comparison. Observational study asks how investigators collected information about an existing situation.

In the book-stall example, the family can use all four questions without claiming to have answered all four. Nor do they need to say the reported relationship is meaningless just because they have not established its explanation.

Do not turn a pattern into a personal verdict

A parent reading about a group can pause before telling a child, “That explains you.” A young person may want to know what was actually measured before discussing whether it resembles their experience. A partner can ask for the original sentence while declining to debate a whole topic that evening.

For this page's everyday task, keep the discussion on the wording and source. You do not need to monitor relatives, collect personal data, or change household rules to practice reading a relationship accurately. No numerical threshold or method for evaluating a person's behavior is supplied here.

Source

[1] Australian Bureau of Statistics. Correlation and causation. Released February 2, 2023; accessed September 5, 2026. “The difference between correlation and causation,” correlation paragraph, supports [1]. Only the narrow terminology distinction is used.

Related guides

• Causation
• Confounding
• Observational study
• Primary source