How Can Asking 1,000 People Reveal What Millions Think?
How Can Asking 1,000 People Reveal What Millions Think? — Without checking everything, a "well-chosen part" can reveal the whole. But bias one way and it's completely wrong. Statistics teaches that "how fairly you choose" decides the answer.
The news says "6 of 10 citizens approve." But they didn't ask all 50 million — just 1,000. How can 1,000 reveal what 50 million think? Can that really be right?
An old cook used the same wisdom. To check the seasoning of a huge pot of soup, you don't drink it all — you stir well and taste one spoon. That spoon represents the whole pot. Old merchants checked grain quality by pulling out one handful. They knew from experience: "mix the whole well, and a part speaks for the whole."
Statisticians turned this wisdom into precise numbers. The key is choosing "evenly, without bias." This is called a sample. Ask only people from one city, or only the elderly, and it skews wrong. So you mix region, age, and gender evenly and pick at random. Amazingly, a well-chosen 1,000 gives nearly the same result as asking all 50 million! There's a small "margin of error," which is why you see notes like "±3%." Ask few but choose well, and you can know the whole — that's the magic of statistics.
- Election polls and exit polls
- TV ratings (estimating a nation from a sample of homes)
- Factory quality control (sampling instead of every item)
- Health checkups (one drop of blood reveals your whole body's state)
抽 (chu) shows a hand (扌) drawing one out from many — the 抽 of "extraction, drawing lots." Fairly drawing a sample from the whole, in statistics, is exactly the math of 抽.
Meet this hanja in Cheonjamun →Without checking everything, a "well-chosen part" can reveal the whole. But bias one way and it's completely wrong. Statistics teaches that "how fairly you choose" decides the answer.
