Think with data
Reason clearly about uncertainty, inference, models, causality, and Bayesian evidence
The statistical thinking track: five lessons on simulating your way to intuition, from sampling variation to Bayesian updating.
Track description
This track teaches statistical reasoning by simulation first, formula second: you’ll see sampling variation, hypothesis testing, regression uncertainty, causal bias, and Bayesian updating happen in front of you before – or instead of – proving them algebraically. Several lessons run live, interactive R in your browser via WebAssembly, so there’s nothing to install to follow along.
Who this is for
Analysts and researchers who use statistics regularly but want firmer intuition for why the methods work, not just how to call them.
Prerequisites
- Minimum: comfort with R and the tidyverse pipe (
|>), as taught in the Build with R track’s first two lessons. - Helpful: having fit a linear model before (even in another tool) makes lesson 3 land faster.
- No prior statistics coursework assumed – concepts are built from simulation, not formulas, though formulas are shown alongside.
Lessons
1
Embrace the noise
Simulate sampling variation to see the law of large numbers and the central limit theorem in action.
2
There is only one test
Run a hypothesis test by simulating a null world, first by hand, then with infer.
3
All models are wrong
Fit, tidy and bootstrap a regression line, and learn where it breaks.
4
Draw your assumptions before drawing your conclusions
Diagram causal assumptions with DAGs and see confounders and colliders bias an estimate.
5
Think Bayes
Update a probability distribution one observation at a time, live in the browser.
Estimated total time: ~4.75 hours across 5 lessons (durations are rounded active-work estimates, not automated reading time).