Learn statistics from zero

Nine steps, in order, from what a variable is to fitting a regression line. Each one is a short read and something you can drag until the idea stops being abstract. No account, no textbook, and no calculus anywhere.

  1. Step 1

    What statistics is actually doing

    Statistics is the science of saying something about a group you cannot fully see, using a smaller group you can. Everything after this is machinery for that one move, and the first thing to get straight is which numbers describe your sample and which describe the population behind it.

    Read: Parameter vs statisticTry: Study design terms

  2. Step 2

    Describing one variable

    Before you can infer anything you have to summarize what is in front of you: where the data sits, how spread out it is, and what shape it makes. The mean and median disagreeing is not a mistake, it is the data telling you it is skewed.

    Read: Mean vs medianTry: Drag points and watch the mean move

  3. Step 3

    Spread, and why standard deviation is the one that matters

    Two data sets can share a mean and have nothing else in common. Standard deviation is the number that separates them, and it is the building block every later formula reaches for, so it is worth computing by hand once.

    Read: Standard deviation by handTry: Check your arithmetic

  4. Step 4

    The normal curve and z-scores

    Once you can standardize a value you can compare anything to anything: a test score against a height against a reaction time. The z-score is the unit conversion that makes those comparable, and the normal table turns it into a percentile.

    Read: How to find a z-scoreTry: Z-score calculator

  5. Step 5

    Probability, just enough of it

    You do not need measure theory. You need to know what independent means, what conditional probability is asking, and why disjoint and independent are different things that students swap constantly.

    Read: Disjoint vs independent eventsTry: Practice conditional probability

  6. Step 6

    Sampling distributions, the idea everything hinges on

    This is the step where statistics clicks or does not. If you took another sample you would get a different mean, and the pattern those means make is predictable. That pattern is what lets one sample say anything about a population.

    Read: The central limit theoremTry: Watch sample means turn normal

  7. Step 7

    Confidence intervals

    An interval is an estimate that admits its own uncertainty. The hard part is not the arithmetic, it is saying what it means without saying the thing that loses the point, which is that the parameter has a 95 percent chance of being in your interval.

    Read: What 95% confidence really meansTry: Build an interval

  8. Step 8

    Hypothesis tests and p-values

    A test asks one question: if nothing were going on, how surprising would this data be? The p-value is the answer, and almost every mistake made with it comes from forgetting that it assumes the null is true.

    Read: What a p-value meansTry: P-value calculator

  9. Step 9

    Relationships between two variables

    Scatterplots, a fitted line, and the two numbers people misread most: r and r-squared. Finish here and you have the whole arc, from describing one column of data to modelling how two of them move together.

    Read: Least-squares regressionTry: Move one point, break the line

Frequently asked questions

Can I learn statistics on my own, without a class?

Yes. Statistics needs almost no prerequisite maths: arithmetic, a bit of algebra, and the ability to read a graph. What it does need is practice at interpreting results in words, which is why every step on this path pairs a short read with something you can manipulate and a worked example to check yourself against.

How long does it take to learn statistics?

Working through the nine steps here takes most people a few weeks at an hour or two a week, which covers the core ideas of an introductory course. A full AP Statistics course runs a school year because it adds depth, more test procedures, and exam technique.

Do I need to be good at maths to learn statistics?

No calculus is required for introductory or AP Statistics. The arithmetic is what a calculator handles. The genuinely hard part is conceptual, holding on to the difference between a sample and a population and saying what a result means in context, and that is a reading and writing skill more than a maths one.

Is this the same as AP Statistics?

This path covers the core ideas that AP Statistics and most introductory college courses share. If you are taking the AP exam, follow the course pages instead: they are organized by the five units of the Fall 2026 framework and cover every topic in the official course description.