Multiple corrections: FDR vs. Bonferroni

juliosun

于 2023-07-10 13:57:16 发布

阅读量549

点赞数

文章标签：其他

本文链接：https://blog.csdn.net/colddie/article/details/131637216

版权

文章探讨了在统计分析中，Bonferroni方法和FalseDiscoveryRate(FDR)控制在多次假设检验中的应用。Bonferroni方法严格控制家庭错误率(FWER)，防止至少一次TypeI错误的发生，而FDR则关注在所有测试中错误发现的比例，提供了一种较不严格的控制，更适用于大量假设检验的情况。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

Bonferroni: Testing the same hypothesis many times, or many different ways. (Some tests may be much less powerful than others). If even one test really shows that the null hypothesis is wrong, then it is dead.

FDR: We are testing many hypotheses (in a similar way). We want to know which ones are likely to be true.

FDR = FP / (FP + TP).

FDR-controlling procedures provide less stringent control of Type I errors compared to family-wise error rate (FWER) controlling procedures (such as the Bonferroni correction), which control the probability of at least one Type I error.