COURSE 04L2100% FREE
Verified 2026-08-10

Mean, variance, and sampling

The **mean** is a typical value. **Variance** (and standard deviation) describe spread. **Sampling** means learning about a larger world from a smaller set of examples—with uncertainty.

What it is

The mean is a typical value. Variance (and standard deviation) describe spread. Sampling means learning about a larger world from a smaller set of examples—with uncertainty.

Why it matters

Metrics, A/B tests, and “the model scored 92%” all depend on careful sampling language. Overconfident averages mislead product decisions.

How it works (plain)

Averages hide multimodal messes. Spread tells you whether points cluster or scatter. Small samples wobble; report ranges and limitations, not fake certainty.

Everyday example

Class average test score of 80 with everyone near 80 differs from average 80 with scores from 40 to 100.

Try it

Write five numbers. Compute a rough mean and say whether the set feels tight or spread out.

Myths

⚠️ Myth: The mean always represents “the typical person.”
✓ Reality: Skewed data needs medians and full distributions.
⚠️ Myth: Bigger samples erase bias.
✓ Reality: Biased sampling stays biased at any size.

Sources