Start here. Tests your ability to summarize data correctly and reason about probability — the two skills that underpin every statistical analysis. Includes traps around mean vs median, skewness, and conditional probability.

Master the landscape of probability distributions and understand why the Central Limit Theorem is the backbone of inferential statistics. Tests distribution selection intuition, PDF vs PMF vs CDF, and the real meaning of standard error.

The most interview-tested cluster in statistics. Covers null/alternative hypothesis, Type I & II errors, the real meaning of p-values, p-hacking, and the most misunderstood concept in statistics — confidence intervals. Includes traps that fool experienced engineers.

Two of the most practically important topics for data scientists. Tests Bayesian updating, prior/posterior intuition, credible vs confidence intervals, and the full A/B testing lifecycle — including the peeking problem, MDE, and multiple comparisons.

Covers two critical areas ML engineers are grilled on — reading regression coefficients correctly, Simpson's paradox, correlation vs causation traps, and the information theory that powers loss functions, feature selection, and model evaluation.

The most dangerous cluster — where smart engineers make expensive mistakes. Tests stationarity, ARIMA, temporal data leakage, and a catalog of statistical pitfalls: survivorship bias, ecological fallacy, multiple testing, and base rate neglect.

Simulates an entry-level data science interview. Broad coverage of fundamentals with one deliberate trap per question. Tests whether you can reason about data — not just recall definitions. Focus: descriptive stats, probability, distributions, CLT, and p-values.

Second easy-level interview simulation. Shifts focus to inference, Bayesian basics, A/B testing fundamentals, and statistical pitfall identification. Tests practical intuition — can you spot what's wrong with a described analysis before running it?

Applied reasoning interview simulation. Presents real scenarios where you must combine concepts across topics — the kind of questions asked at Google, Meta, and Stripe data science rounds. Tests inference, experimental design, regression interpretation, and time series gotchas.

Second medium interview simulation. Focuses on Bayesian updating, CLT limits, information theory in ML contexts, and diagnosing biased studies. Scenario-based questions that require you to reason about what could go wrong — not just what should happen.

Senior-level interview simulation — first session. Questions from this test appear in staff-level interviews at top-tier AI companies. Covers edge cases, non-intuitive statistical behavior, causal inference traps, and scenarios where the obvious answer is wrong.

Senior-level interview simulation — second session. Focuses on the applied side: A/B testing edge cases, regression pitfalls, information theory in production ML, time series modeling failures, and the most dangerous statistical biases encountered in real data pipelines.

Staff-engineer and AI architect level assessment. Vol. 1 covers the statistical foundations — distribution theory, CLT edge cases, inference machinery, and Bayesian causal reasoning. Every question is a known interview-killer at FAANG companies. Requires multi-step reasoning and a deep mental model of how statistics actually works under the hood.

Staff-engineer and AI architect level assessment. Vol. 2 goes deep on the applied and production side — A/B test failure modes, regression causality, information theory in neural networks, time series modeling at scale, and the most sophisticated statistical pitfalls that cause production ML systems to silently fail. This is as hard as it gets.