Unrepresentative sample

Unrepresentative sample is when you ask the wrong people (or too few, or only one type) and then say "everyone" or "most people" think or do that. Your sample doesn't reflect the group you're generalising about, so the conclusion is unreliable. It's a way that hasty generalization happens in practice: the sample is not only small but biased.

Examples

• We asked 100 people at the vegan festival if they eat meat. Nobody said yes. So hardly anyone eats meat.

• We polled only our Twitter followers, and 80% support the change, so the public supports it.

• Everyone at the shareholders' meeting voted yes, so the shareholders support the merger.

• I asked my family and they all liked it, so it's a hit.

• Every student in the honours class passed, so the exam is easy.