You want the statistics done properly

By the end you will know what a distribution is doing, when a test applies, and how a perfectly correct calculation can still hand you the wrong answer.

8 steps · about 12 minutes · free, no account, read in any order you like

Start reading

1. 1

   Statistics in practice · 1 min read · 1 of 8

   What is statistics, and the gap people fall into

   Description, inference, and the gap between them, which is where most confident mistakes live.

2. 2

   Data & decisions · 2 min read · 2 of 8

   Data exploration: the first hour decides the rest

   What to do in the first hour with a dataset, before any test is chosen.

3. 3

   Statistics in practice · 2 min read · 3 of 8

   The normal distribution, and where it does not belong

   The distribution everyone assumes, and the places it does not belong.

4. 4

   Statistics in practice · 2 min read · 4 of 8

   The beta distribution: honest talk about rates

   The honest way to talk about a rate when the sample is small.

5. 5

   Statistics in practice · 1 min read · 5 of 8

   The chi-squared test, and the fine print nobody reads

   What the test measures, and the fine print nobody reads.

6. 6

   Statistics in practice · 1 min read · 6 of 8

   Why the chi-squared test is right-tailed

   Why the test looks one way only, and what a suspiciously perfect result means.

7. 7

   Statistics in practice · 1 min read · 7 of 8

   Chi-squared worked example, and where it goes wrong

   The same test done step by step, with every place the arithmetic can go quietly wrong.

8. 8

   Data & decisions · 2 min read · 8 of 8

   Decision points: where expected value gets it wrong

   Finally, what the numbers are for: choosing, and the trap inside expected value.

A track is only an order. Every piece in it stands on its own, and nothing here is behind a sign up.