Hasty Generalization

Examples

On a trip, a traveler has two unpleasant run-ins.

Owen: “Two people from that city were so rude to me this week. People there just have no manners.”

Two encounters, out of a city of hundreds of thousands, become “people there.” Owen has generalized from a sample far too small — and almost certainly unrepresentative — to say anything reliable about an entire population.

The same jump happens constantly in app reviews after an update:

Review: “New update crashed on my phone once. This app is trash now, avoid it.”

One crash on one device becomes a blanket verdict on the app for everyone. Thousands of other users might have had no issues at all.

Why the reasoning fails

A generalization is only as reliable as the sample it’s built on. Hasty generalization draws a sweeping conclusion — about a group of people, a product, a whole category — from too few data points, or from data points that aren’t representative of the wider group. Two rude individuals could be an unlucky coincidence, not a trait of an entire city; one crash could be a quirk of one device configuration, not the product as a whole. The reasoning doesn’t fail because small samples are useless — they can be a legitimate starting hint — it fails when a small, possibly unrepresentative sample gets treated as if it proves something about the whole population.

How to respond

• Ask about sample size: “How many people is that based on? Just the two you met?”
• Ask whether the sample is representative: “Is there a reason those two would be typical, or could that be bad luck?”
• Point to the wider picture if it’s available: “The app has a 4.5-star average across 50,000 reviews — does one crash change that?”
• Don’t demand a huge sample for everything. A strong pattern noticed after a handful of consistent experiences can be a reasonable first impression; the issue is stating it as a settled conclusion rather than a hunch worth checking further.