Draw samples from a population and watch a distribution of statistics build up. Two separate controls: sample size n changes which distribution you get (a different, narrower one), while drawing more samples just fills in the same distribution.
The true population value is hidden — estimate it from the samples, then reveal it.
Population \(\mu\):—Population \(\sigma\):—
Samples
Populationwhere samples come from
One samplethe latest draw of n = 30 values
Sampling distributionone dot per sample's \(\bar{x}\)
Frozen for comparison —
Current —
Samples drawn:0SD of \(\bar{x}\)'s:—\(\sigma/\!\sqrt{n}\) (theory):—
The center of the sampling distribution sits at \(\mu\) for every \(n\) — increasing \(n\) does not move the center, it only shrinks the spread.
Choose a population shape and click a button to draw samples.