Sampling bias
When the people or things you measured aren't a fair picture of the whole group, so your results lean one way.
In plain English
A sample is only useful if everyone in the group had a fair chance of being in it. If some kinds of people were much more likely to be picked, the answer is skewed, and a bigger sample from the same place won't fix it.
Uncle Gary explains
Uncle Gary tastes the chilli from the top of the pot and declares it mild. All the chillies sank to the bottom. Aunt Linda's first spoonful is still discussed at family gatherings.
Why it matters for your money
Survey results, online reviews and trading backtests all suffer from it. Always ask who had a chance to be counted, and who didn't.
Where we tested it
- Lesson 1: What Is Statistics Actually For?
- Why You Always Pick the Slowest Queue
- Lesson 6: Why Polls Only Ask About 1,000 People
- Are Employee Surveys Really Anonymous? The One That Needed Your Login
- "Everyone Else Is Allowed"
Related terms
All glossary terms. Educational content, not financial advice.