Resampling Statistics

Explore data through resampling-based statistical methods

← Back to Home

2. Describing Data

Explore dot plots, jitter, and beeswarm plots — and learn what to look for when visualizing a distribution.

3. Probability Simulations

Explore how probability distributions emerge from randomness through Poisson and coin flip simulations.

4. Confidence Intervals

Use resampling to build a confidence interval for a study's observed rate, then see what "95% confidence" really means across repeated studies.

5. Normal Distribution

Explore the normal distribution properties through the Quincunx simulation. Then sample directly from a normal distribution and watch the fitted curve converge as draws accumulate.

6. Comparing Two Groups

Compare two groups' T-cell counts with beeswarm plots, KDE curves, and a big-box resampling test.

7. Comparing Three+ Groups

Test three T-cell groups at once with an omnibus resampling test, then find out which pairs differ with post-hoc comparisons.

8. Categorical Data

Test whether categories are related using contingency tables and a shuffle-based NHST — Titanic survival by passenger class, with a "Mt. Fuji" insignificance-band plot, and heart disease by gender.

9. Correlation Explorer

Test the link between CO2 and temperature over 800,000 years with shuffling and bootstrap intervals, then see how an outlier and sample size can change a correlation.

10. Regression Analysis

Fit trends to real climate data — cherry blossom bloom dates, ice breakup and cover, global temperature, US and California wildfires, and atmospheric CO2 — then test each trend with permutation shuffling and quantify it with resampling-based confidence intervals.

11. Statistical Power

Drag a sample-size slider and watch statistical power and the Type II error rate update live, then see the full power curve. Then find how many years of data it takes to detect a real climate trend.

12. Bayesian Updating

Step through a clinical trial patient by patient and watch a posterior belief update in real time, then test it against a null distribution.


Tips for Using These Simulations

These interactive simulations are designed to help you develop intuition about statistical concepts. Try adjusting parameters to see how they affect the results, and compare different scenarios to deepen your understanding.