Topic hub
Experiments & Causality
How to design A/B tests and experiments that answer the question you asked, and how to separate genuine cause from coincidence in observational data.
A/B Test Significance: How to Read Results
Learn how to judge A/B test significance: p-values, confidence intervals, effect size and common traps, with a worked conversion-rate example you can follow.
Blocking vs Stratification: Block or Sample?
Blocking vs stratification: learn when to block treatments inside an experiment versus stratify a sample, plus how an A/B conversion test applies each method.
Bonferroni Correction: A Practical Guide
The Bonferroni correction lowers your significance threshold to control the family-wise error rate across multiple hypothesis tests for reliable results.
Control Confounders with DAGs
A practical guide to control confounders with DAGs: apply the disjunctive cause rule, run balance and propensity-score checks, and report sensitivity analyses.
Difference in Differences Explained
Practical guidance on difference in differences: the $4,000 worked example, parallel trends checks, and modern estimator choices for staggered treatment timing.
False Discovery Rate: How to Control It
The false discovery rate controls false positives among your significant results. Learn Benjamini-Hochberg, q-values, and how to pick a threshold.
How to Design an A/B Test That Answers Your Question
Learn how to design an A/B test before you run it: frame a hypothesis, pick metrics, set the minimum detectable effect, and size the sample.
Instrumental Variables Explained
Instrumental variables explained: diagnostics, falsification checks, weak-instrument fixes, and a quarter-of-birth worked example for causal inference.
Intention to Treat: ITT vs Per-Protocol
Intention to treat (ITT) means analyzing every randomized participant as assigned. See a worked ITT vs. per-protocol example, a checklist, and sources.
Interrupted Time Series Analysis Guide
A protocol-first guide to interrupted time series: pre-specify your model, then check autocorrelation and seasonality before reporting an effect.
Mediation Analysis: How to Test for It
Mediation analysis measures how much of an effect flows through a mediator: define the paths, bootstrap the indirect effect, and check causal assumptions.
Multiple Comparisons Problem: A Practical Guide
The multiple comparisons problem means 12 independent tests carry a 46% false-positive risk; see when Bonferroni or FDR correction applies.
Propensity Score Matching: A Step-by-Step Guide
Learn propensity score matching step by step, including variable selection, matching methods, caliper choice, and covariate balance checks.
Randomized Controlled Trial (RCT) Explained
See how a randomized controlled trial isolates causal effects in medicine, economics, and business, and what makes results credible today.
Regression Discontinuity Design Explained
Methods-first guide to regression discontinuity design: run McCrary and covariate checks, test bandwidths, and plan for 9 to 17x larger samples.
Selection Bias in Experiments: Causes & Fixes
Selection bias in experiments explained: where it enters randomized studies, with worked examples and a checklist to detect and reduce it.
Sequential Testing Guide: When to Stop Early
Learn sequential testing with group sequential design and alpha spending, plus a six-step prelaunch checklist for adjusted interim A/B tests.
Spillover Effects in Experiments
A measurement-first primer on spillover effects in experiments: identify and estimate them in 5 steps, with real empirical magnitudes and sources.
Synthetic Control Method: A Practitioner Guide
Learn when the synthetic control method fits, how to choose donors, and how placebo tests, robustness checks, and Statohub tools support policy evaluation.