Random Sample Picker
Paste a population, choose a sample size, and draw with or without replacement, with optional strata. Every draw uses crypto-strong randomness and gets a shareable draw ID. Works offline.
Built by the AbraCalc team
How to play
- Paste your population, one item per line, optionally adding ", group" for strata.
- Set your sample size and toggle replacement and stratified sampling as needed.
- Tap Draw sample to see the chosen items highlighted against the full population.
- Tap Draw again to pull a fresh sample instantly and compare results.
Random Sample Picker is a hands-on statistics demo for teaching how sampling actually works. Paste a population list, one item per line, choose how many you want to draw, and pick whether the draw happens with or without replacement. Add an optional group after any item (like "Ava, red") to turn on stratified sampling, which allocates the sample proportionally across groups instead of pulling from the whole pool at once. Every draw is generated with the same crypto-strong randomness used throughout AbraCalc, and each result gets a shareable draw ID so a class can refer back to "which draw" produced a given sample. It's built for showing sampling bias, fairness, and the difference between sampling methods live, right in the browser, with no spreadsheet required.
Frequently asked questions
- What's the difference between with and without replacement?
- Without replacement removes each drawn item from further consideration so nothing repeats in one sample; with replacement puts every pick back, so the same item can be drawn more than once.
- How does stratified sampling decide how many to take from each group?
- It allocates the sample size proportionally to each group's share of the population, using a largest-remainder rounding method so the totals always add up to exactly the sample size you asked for.
- Is the draw ID a seed I can replay?
- No, it's just a shareable label for that particular draw. The randomness itself is crypto-strong and not re-seedable, so every draw is genuinely fresh rather than reproducible from an ID.