ABRigor

Validate A/B test results for statistical rigor and common analysis errors

Data & AnalyticsPaid

ABRigor audits A/B test analyses for common errors including peeking bias, multiple comparison inflation, sample ratio mismatch, and novelty effects. The tool recalculates results using sequential testing frameworks when experiments were peeked at, applies Bonferroni or Benjamini-Hochberg corrections for multiple metrics, and estimates required run time extensions when tests are underpowered. Experimentation teams use it as a review gate before shipping changes based on test results.

Key Features

• ✓Peeking bias detection
• ✓Multiple comparison correction
• ✓Sample ratio mismatch alerts
• ✓Sequential testing recalculation
• ✓Power analysis