A/B Testing Statistical Experiment Designer

Act as a Chief Experimentation Officer and Applied Statistician.

Design a rigorous A/B testing plan for this product experiment:
- Hypothesis: [e.g. Changing checkout button from green to blue increases checkout completion]
- Primary Metric: [Conversion rate %, average order value, retention rate]
- Baseline Metric Value: [Current baseline, e.g. 3.2% conversion rate]
- Minimum Detectable Effect (MDE): [Relative or absolute improvement expected, e.g. +10% relative increase]
- Daily Traffic: [Unique visitors/users entering experiment daily]

Generate the experiment protocol:
1. Sample Size & Duration Calculation:
   - Set Statistical Power ($eta = 0.80$) and Significance Level ($lpha = 0.05$).
   - Total sample size needed per variant.
   - Required run time in full weeks to control for day-of-week seasonality.
2. Randomization & SRM Check:
   - Sample Ratio Mismatch (SRM) chi-square test protocol to verify allocation integrity.
3. Statistical Test Selection:
   - Specify whether to use Two-sample Z-test for proportions, Welch's t-test, or Mann-Whitney U test.
4. Peeking Problem & Stopping Rules: Guidelines to prevent early-stopping bias.
5. Guardrail / Secondary Metrics: Key health indicators that must not regress (e.g. page latency, return rates).