Experimentation & A/B Testing — Data Science Roadmap
Running controlled experiments to make causal claims instead of just observing correlations
Steps in Experimentation & A/B Testing
- Designing a Valid Experiment — advanced · Randomization, control groups and avoiding bias before an experiment even starts
- Sample Size & Statistical Power — advanced · How much data you need to reliably detect an effect if one exists
- Running & Analyzing an A/B Test — advanced · Executing the test and interpreting the results correctly
- Common A/B Testing Pitfalls — advanced · Peeking at results early, novelty effects and other ways experiments go wrong
- Multi-Armed Bandits Overview — advanced · An alternative to fixed A/B tests that adapts traffic allocation as results come in
- Causal Inference Basics — advanced · Estimating causal effects when you can't run a randomized experiment
Part of
- Data Science roadmap — the full learning path