Case · Unit 2 of 5 · 6 min

The headline that everyone shared

A headline crosses your feed: "People who drink coffee live longer, major study finds." Eleven people you know have shared it. It matches something you already suspected. You feel the small pull to share it too.

Take it apart

Claim — as written, it says coffee causes longer life. What the study almost certainly measured is that coffee drinkers, on average, died later than non-drinkers.
Evidence — a population of people, observed over time, sorted by a habit they chose themselves.
Reasoning — from "these two things occur together" to "one produces the other".
Assumption — that coffee drinkers and non-drinkers are otherwise alike.

That last assumption is where it collapses. People who avoid coffee often do so because a doctor told them to, or because they are already unwell, or because caffeine disturbs sleep they were already struggling with. The groups differ before the coffee ever enters the picture.

The useful question

Not "is this true?" but "what else would produce exactly this result?" If you can name two other explanations that fit the same evidence, the evidence has not yet picked out the claim.

None of this means the headline is false. Coffee may well be mildly good for you. It means the headline has not earned the confidence that eleven shares implied. Holding it at "plausible, unproven" costs you nothing and keeps you honest.