Sample size at 99% confidence
The 99% confidence level is the choice for high-stakes decisions: regulatory studies, health, hard-to-reverse investments. The trade-off is direct: with z = 2.576, the required sample climbs about 73% above the 95% standard. For ±5%, you need 664 respondents; for ±1%, over 16,500.
Before paying that premium, ask the reverse question: does the error you're guarding against cost more than ~280 extra respondents? If your study feeds an exploration or a directional signal, 95% is plenty. The calculator below is preset to 99% to size your exact case.