Rate Change Significance

Analyze whether a binary rate differs significantly and meaningfully between two periods, versions, cohorts, groups, or the same users before and after. Use for retention, conversion, repurchase, churn, renewal, payment and other 0/1-rate comparisons when the user asks whether an uplift/drop is real, whether a campaign or version improved the rate, whether the sample supports a conclusion, whether overlapping users require paired analysis, or whether Z-test, Fisher exact, or McNemar is appropriate. Includes cohort selection, sample-quality checks, overlap-aware test routing, reproducible statistics and business interpretation. Not for continuous metrics, multi-period forecasting, or auditing how the source metric was calculated.

thinkingaiagenticengine Updated 23 repo stars

File contents

thinkingaiagenticengine/scenario-skills commit 6498088e4b

Frequently asked questions

npx get-skill add thinkingaiagenticengine/rate-change-significance