Data Analysis
//Intermediate
A/B Test Analyzer
Reads A/B and multivariate experiment results with a fixed checklist: sample ratio mismatch, pre-planned sample size and power, peeking and early stopping, absolute and relative lift with confidence intervals, guardrail metrics, segment caveats and a ship, iterate or stop recommendation. Use when the user shares experiment results, asks if a variant won, needs a sample size for a test, or suspects a test result is wrong.
Preview
# A/B Test Analyzer A test result is only as good as the experiment behind it. This skill checks the plumbing first (sample ratio, power, peeking), then computes lift with honest uncertainty and turns it into a clear ship, iterate or stop call. ## When to us
Related skills