Unverified50% confidenceFactExact time
在5%显著性水平下测试100个随机因子，期望会有5个看起来显著（多重假设检验问题）
1
Sources
50%
Confidence
Long-term
Relevance
7/21/2026
First Seen
Sources
Related Claims
UnverifiedWhen a team tests 1,000 combinations of strategy parameters, even if all combinations have zero predictive power, roughly 50 combinations will appear to perform excellently in backtesting purely due to randomness (the multiple comparisons problem).79% similarUnverifiedWhen testing LLM citation behavior, the same prompt should be run at least 20–50 times, ideally expanded to more than 100 times.72% similarUnverified置换检验通过大量随机置换（通常1000到10000次）构建零假设下的经验分布来估计p值，无需参数假设但计算成本相对较高71% similarUnverified对100万个SNP采用0.05显著性阈值进行假设检验预期将产生5万个假阳性71% similarUnverifiedA 95% confidence level means that if the same methodology were used to conduct 100 surveys, approximately 95 of them would produce results falling within the margin of error.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/577436API
curl https://kongchang.com/api/v1/knowledge/claims/577436MCP
get_claim(id=577436)