FullCalculator

A/B Test Statistical Significance

A/B Test Significance

Formula

Z = (Rate_B - Rate_A) / sqrt(P_pooled * (1 - P_pooled) * (1/n_A + 1/n_B))

Frequently Asked Questions

What is statistical significance in A/B testing?
Statistical significance means the difference in results between your control and variation is unlikely due to chance. A p-value below 0.05 (95% confidence) is the standard threshold for declaring significance.
How long should I run an A/B test?
Run your test until you have enough sample size for statistical significance, typically at least 1-2 weeks. Ending tests too early can lead to false positives. Aim for at least 100 conversions per variation.

You may also need