Free · 20 questions · 9 min · no sign-up

Statistics test

Statistics is full of ideas that are easy to state and easy to get backwards: what a p-value means, when an average misleads, why a poll of a thousand people works for a whole country.

Question 1 of 20

Towns that sell more ice cream have more drownings. What’s the most likely explanation?

Scored in your browser. Your answers aren’t stored.

About this test

Every wrong answer here is a named mistake. You’ll see why after each question, and at the end you get a course that starts from the ones that caught you.

Questions people ask

What does a p-value actually mean?

How surprising your data would be if there were no real effect. A small p-value means the data would be unusual under that assumption. It isn’t the chance the result is wrong.

Is this good practice for a stats exam?

It covers ideas that come up on most intro statistics courses and exams, like AP Statistics and first-year university stats, but it tests understanding rather than calculation.

Why doesn’t a bigger country need a bigger poll?

A random sample’s precision depends on its size, not the size of the population it’s drawn from, the same way a spoonful of well-stirred soup tells you its saltiness whatever the size of the pot.

How many questions are in the statistics quiz, and how long does it take?

Twenty multiple-choice questions, about nine minutes. After each one you see whether you were right and why, and at the end you get a score, a breakdown by topic and the named mistakes that caught you.

What does the margin of error in a poll mean?

How far the poll’s figure could plausibly be from the true one because only a sample was asked. A poll of 1,000 people has a margin of about 3 points on each share, so a lead of 4 points between two parties is within the noise.

Read more

More tests