Multiple Hypothesis Testing & False Discovery

This simulator runs 224 independent experiments at once — a bank of simulated scientific studies, each with its own random sample. Some experiments secretly test a real effect; others test nothing at all. Every experiment produces a p-value from its own noisy data, and a pillar rises to mark how strong that evidence looks. Tune the sample size, the true effect size, and the mix of real-vs-null hypotheses, then switch between no correction, Bonferroni, and Benjamini-Hochberg to watch statistical power and the false discovery rate trade off in real time — the same forces that drive underpowered research toward false "discoveries" in the real world.