← All questions
Stats & probability
Distributions → inference → experiments → regression → causal
Why this sequence
Build the foundations (distributions and MLE), master frequentist vocabulary (CI, p-value, power), then walk through experiment design, A/B pitfalls, and regression-based causal reasoning. The final third integrates everything on Ads-scale production problems.
38 questions ·
0 mastered ·
0 in review
Start here
…
- 01 NewBinomial Distribution — Probability of Fewer Than 2 Girls Out of N Children
- 02 NewHow Would You Prove a Sequence Follows a Uniform Distribution?
- 03 NewPoisson vs. Binomial — Assumptions, When to Use Each, and the Poisson Approximation
- 04 NewDerive the Conditional Distribution Y | X for a Bivariate Normal
- 05 NewDerive the MLE for Normal(μ, σ²) from Scratch
- 06 NewTruncated Normal — Estimate μ and σ² When You Only Observe X > 3
- 07 NewCensored Normal — Estimate μ, σ² When X > 10 Is Right-Censored
- 08 NewExplain CI, P-value, and Alpha — How Would You Explain Them to a PM?
- 09 NewTesting a Population Mean — What Does p-value = x% Actually Mean?
- 10 NewWhy Can't We Reverse the Null Hypothesis to "Prove" Something?
- 11 NewHow Does a Data Scientist Decide the P-Value Cutoff?
- 12 NewHow Do You Choose Between 99% and 95% Confidence Level?
- 13 NewYour SE is 0.1, You Want SE = 0.01 — What Do You Do? What If Sample Size Can't Change?
- 14 NewGiven a Sample X₁, …, Xₙ ~ X, Estimate P(X > 10) and Construct a 95% CI
- 15 NewHow Does Bootstrap Work? Can It Be Used for Variance Reduction?
- 16 NewCompute a P-Value from the Definition via Monte Carlo Simulation — No Formulas
- 17 NewDrug A (Placebo) vs B — Between-Subjects vs Within-Subjects Design
- 18 NewDrug Trial — Parallel vs Crossover Design (Detailed Follow-ups)
- 19 NewDesigning an Experiment for an Android App Update (Self-Selection Bias)
- 20 NewExperiment Where Users Must Sign an Agreement — Selection Bias and How to Fix It
- 21 NewCRM Algorithm A/B Test — Identifying Wrong CI Calculation & Proper Experiment Design
- 22 New30 A/B Tests, One Significant at p = 0.04 — Should You Ship?
- 23 NewNovelty and Primacy Effects — How to Detect Them and What to Do
- 24 NewThe Peeking Problem — Why Early Stopping Inflates False Positives, and How to Fix It
- 25 NewNetwork Effects and Spillover — Why User-Level A/B Breaks and How to Fix It
- 26 NewIncrementality Testing — Measuring What Ads Actually Cause, Not Just What They're Credited For
- 27 NewExplain Linear Regression to a Non-Technical Person
- 28 NewWhat Happens If You Duplicate Data in Linear Regression?
- 29 NewThe Four Assumptions of Linear Regression — What Breaks When Each Is Violated, and How to Diagnose
- 30 NewWhy Is the Test on a Regression Coefficient a t-Distribution, Not Normal?
- 31 NewDoes OLS Give the Same Predictions If You Rotate the Features?
- 32 NewLasso vs. Ridge — When to Use Which, and Why Lasso Creates Sparsity
- 33 NewRelationship Between Product Sales and Review Data
- 34 NewDoes YouTube Ad Playback Drive Sales? Spot the Problems in This Simple Regression
- 35 NewDoes Listening to the YouTube Music "Commute" Playlist Make People Drive Faster?
- 36 NewA Campaign Launched in One Country — Did It Actually Cause the Metric Lift?
- 37 NewGlobal Sales Went Up — But Could Regional Analysis Show It Went Down?
- 38 NewBad Search Algorithm A/B Test — Three-Part Analysis