Home▸Data Science▸Sampling Bias Simulator

Sampling Bias Simulator

Interactive sampling bias simulator: a whole survey population sits in four neighborhoods with different real averages, and you draw random, convenience or stratified samples to see how each method's sample mean drifts from the true population mean.

Data Science3DEasy60 FPS

Good data collection depends on more than measuring accurately — it depends on measuring the right people. This simulator lays out an entire survey population of 400 households across four neighborhoods, each with a different real average income and a different share of the town, then lets you draw a sample using three classic techniques: pure random selection, convenience sampling from whoever is nearest, and stratified sampling that mirrors each neighborhood's true share. Every point in the town is visible at all times; the highlighted subset is whatever the current method actually samples. Compare the resulting sample mean against the true population mean to see, directly, why the same sample size can be either a reliable estimate or a systematically biased one depending purely on how it was collected.

⚙ Under the hood

Interactive 3D survey town of 400 households across four neighborhoods with different real incomes: draw random, convenience or stratified samples and watch the sample mean drift from the true population mean depending purely on how the sample was collected.

Three.jsdata-collectionsampling-biasstatisticssurvey-methodsdata-science

3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install