Intuitive Statistics

Statistics you can see before you calculate.

A first course in statistics, built on real questions and real data. Each idea is put to work first. Then we look at why it works, with pictures and simulations rather than derivations.

Everyone in a population (lopsided: most values are small, a few are large)

Averages of random samples, 25 from the population each time

Each dot is the average of one sample. The people are spread far and wide; their averages land close to the dashed line.

How the course works

Start from a question

Ideas come in when a real question needs them.

Use it, then open the hood

A method is first used like a trick that works. Then we look underneath to see why it works.

See it, then name it

Pictures and simulations you can run come first. The vocabulary arrives when there is something to name.

Ask what a result can support

Who was counted, what was measured, what is missing and why, whether a difference matters, whether a causal claim is earned.

The path

  1. Describe descriptive statistics

    What a pile of numbers looks like: center, spread, shape, and how groups compare.

  2. Estimate confidence intervals

    Saying something about everyone from a sample, and how far off that could be.

  3. Test hypothesis tests

    Whether a difference is real or could be chance.

  4. Model regression

    A line through a cloud of points: what it says and what it does not.

  5. Cause causal inference

    When “goes with” means “leads to,” and how we could tell.

Lessons will be posted here as they are written.