The A/B Test “Winner” Might Be Lying to You
中文摘要
A/B测试的“获胜”结果可能具有误导性，文中分析了统计显著性、窥视等因素如何导致错误结论。
English Summary
A/B test "winners" can be misleading; the article explains how statistical errors like peeking and p-values lead to false conclusions.
原文节选
Statistical significance, sample size, confidence intervals, p-values, MDE, power, guardrail metrics, peeking, and business impact can… Continue reading on Medium »