Skip to content
AliceMethod
← All topicsAges 10–17

Statistics

Detective work on data

Where it goes wrong

Traditional teaching

Compute the mean, median, mode, and range of a list. Draw the bar chart. Statistics becomes arithmetic homework wearing a lab coat — and the actual skill, deciding what data can and cannot tell you, is never practiced.

The reteach

The Alice Method

Statistics is detective work: every dataset is evidence, and the question is what conclusions the evidence supports. Start with real questions the child cares about (do our family's pancakes get eaten faster on Saturdays?). Teach the summary statistics as compressions that lose information — the mean of {50, 50} and {0, 100} is the same, and the story isn't. Always ask: what got left out? Who collected this? Compared to what?

Mental model

Compression with loss

An average is a lossy summary: one number standing in for many, chosen to be representative. Every summary throws information away — the skill is knowing what got discarded and whether it mattered. 'The average hides the spread' is the statistician's first reflex.

Transfer

Where this shows up for the rest of their life

This is the defense kit for modern life: news claims, product reviews, medical studies, school dashboards (including MAP reports — parents use exactly this skill in Layer 3). Correlation vs. causation, sample vs. population, outlier vs. trend — children who ask 'compared to what?' grow into adults who can't easily be fooled.

Watch for these

Common misconceptions

The misconception

The average is the typical case.

The repair

One billionaire in a café makes the 'average customer' a millionaire. Skewed data breaks the mean — that's why medians exist, and why 'average' claims deserve suspicion.

The misconception

Correlation implies causation.

The repair

Ice cream sales and drownings rise together (both follow summer). Give kids a stock of silly confounders and they'll spot real ones for life.

Seen, not said

The visual explanation

Two datasets with identical means and wildly different spreads, drawn as dot plots side by side. The mean as a balance point you can see — and see failing to tell the whole story.

Try it

Interactive example

Sampling variability, live: small samples mislead, large samples stabilize.

Probability is long-run behavior, not fortune telling

Ten flips are wild. Three hundred flips settle near 50%. Randomness is messy up close and orderly from far away.

50%flip some coins →