How it works

This calculator runs a pooled two-proportion z-test on your variant data. It pools the conversions from both variants to estimate a shared rate under the null hypothesis, computes the z-score for the observed difference, and converts it to a p-value using the standard normal CDF (via the Abramowitz–Stegun erf approximation, accurate to about 1.5×10⁻⁷). The confidence interval for the difference uses the unpooled standard error. For one-tailed tests, the p-value is reported in the direction of the observed effect. All math runs in your browser — no data leaves your device.