Statistics makes more sense when you can see it move.

Sort real research scenarios into observational studies and experiments, then see how the same question leads to different conclusions depending on the design.

Run biased and unbiased samples side by side on the same population and watch how a "good-looking" sample can still be wildly off.

Drag a bin-width slider and watch the same dataset tell completely different stories depending on how you group it.

Feed one dataset into three graph types simultaneously and see which patterns each display reveals -- and hides.

Match datasets to the right graph type and see why the wrong choice doesn't just look bad -- it misleads.

Drag an outlier slider and watch the mean chase it while the median barely flinches. This is why we report median income, not mean income.

Add and remove data points while the box plot rebuilds live. Outlier flags appear with the 1.5×IQR rule explained on the spot.

Compare z-scores across two different distributions and discover why a 90 on one exam can be more impressive than a 95 on another.

Flip coins, roll dice, and draw cards while watching experimental probability converge toward theoretical probability as trials stack up.

Build your own probability distribution from scratch. A visual filling bar enforces the rule that all probabilities must sum to 1.

Click regions of an interactive Venn diagram and watch union, intersection, and complement probabilities calculate live.

Set event probabilities and see the addition and complement formulas fill in with your values, step by step.

Toggle between mutually exclusive and independent events and watch the Venn diagram structure change. Two concepts students constantly confuse, side by side.

Enter your own probability distribution, watch the bar chart build, and see the mean and standard deviation calculate automatically as you type.

Drag boundary markers under a smooth curve and watch the shaded area -- your probability -- change in real time.

Slide μ and σ and watch the bell curve stretch, compress, and shift. Convert any x-value to a z-score instantly.

Toggle between left tail, right tail, and between-two-values modes. The shaded area is your probability, calculated and displayed live.

Draw samples from skewed, uniform, and bimodal populations and watch the sampling distribution of the mean become normal before your eyes.

Enter a confidence interval to extract the point estimate and margin of error, or enter parameters to build one. Formula shown step by step.

Watch 100 confidence intervals animate one at a time. Green captures μ, red misses. The running percentage shows what "95% confident" actually means.

Build a confidence interval on a number line, then reverse the formula to find the sample size. Doubling precision requires quadrupling n.

Walk through an interactive flowchart of the hypothesis testing process. Each stage reveals its explanation when you click it.

Select a claim type and watch the null and alternative hypotheses generate in correct notation, with the rejection region shaded in the right tail(s).

Drag the α slider and watch the rejection region, critical value, and p-value all respond in real time on a single distribution curve.

Read a real research scenario, identify the test type, enter the statistics, and build a conclusion step by step with feedback at each stage.

Set your decision and the true state in a Type I / Type II error matrix. See which error applies and why these are tradeoffs, not mistakes to eliminate.

Calculate the z-value step by step with every formula substitution shown, then see it plotted on the distribution curve.

Run a complete hypothesis test from claim to conclusion. The p-value shades on the curve, compares to α visually, and the conclusion states in both statistical and plain English.

Drag points on a scatter plot and watch the least-squares regression line and equation recalculate with every move.

Move one outlier and watch r jump from weak to strong. Make predictions on the regression line and see r² explained as "percent of variance."

Review real correlations -- ice cream sales vs. drowning, shoe size vs. reading level -- and identify the lurking variable that makes r meaningless.