Activities CatalogStatLens
Each activity adds a guided, step-by-step instruction panel to a StatLens tool —
predict-then-reveal questions, live feedback, and a conclusion in context. Open one below, or add
?activity=<file>.json to any tool URL. Every activity supports a student
Discovery mode (gated questions) and an instructor Presentation mode
(add &mode=present). Authors: see the
Activity Authoring Guide.
Exploring & describing data
Comparing Charts for Categorical Data
Bar chart vs. pie chart vs. table — which display reveals categorical patterns most clearly?
Comparing Charts for Quantitative Data
Dotplots, histograms, and boxplots side by side — what each one shows and what it hides.
Comparing Distributions Across Groups
Read side-by-side distributions to compare a quantitative variable across groups.
Outliers and Robust Statistics
Watch a single outlier yank the mean while the median barely moves.
Two-Way Tables & Conditional Proportions
Build a two-way table and read conditional proportions to judge whether two variables are associated.
Describing Associations in Scatterplots
Direction, form, and strength — describe the association between two quantitative variables.
Sampling distributions & bootstrap
Building a Sampling Distribution of x̄
Draw repeated samples and watch the sampling distribution of the mean take shape.
Building a Sampling Distribution of p̂
The same idea for proportions — see how sample size sharpens the sampling distribution.
Exploring Bootstrap Resampling
Resample your data with replacement to see where a sample statistic varies.
How Confidence Level Affects CI Width
90% vs. 95% vs. 99% — the confidence-versus-precision trade-off, made visible.
Medical Consultant: Bootstrap CI for a Proportion
Build a bootstrap confidence interval for a single proportion in a real context.
Confidence intervals & hypothesis testing
What Does “95% Confident” Really Mean?
Tackles the top CI misconception head-on: 95% is a property of the procedure, not a chance for one interval.
Randomization Test, Step by Step
Shuffle the group labels, build a null distribution, and read a p-value by construction.
Type I & Type II Errors
Run many studies to see false alarms, missed effects, power, and the α trade-off.
Do Green Jelly Beans Cause Acne?
Read the xkcd comic, then run 20 tests on pure noise and watch a false “discovery” appear — the multiple-comparisons trap.