Controlling the false discovery rate: a practical and powerful approach to multiple testing

By Yoav Benjamini, Y. Hochberg. 1995

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Evaluation & Benchmarks

Citations in the corpus

BLOOM: A 176B-Parameter Open-Access Multilingual Language Model (2022, methodology): “Here, we use the non-parametric test Mann-Whitney U (Mann and Whitney, 1947).”
Tight Lower Bounds for Differentially Private Selection (2017, neutral): “The accuracy condition in Theorem 2 is closely related to the false discovery rate, which is a widely used statistical criterion introduced in the influential work of Benjamini and Hochberg [23].”
RAPPOR: Randomized Aggregatable Privacy-Preserving Ordinal Response (2014, methodology): “Alternatively, controlling the False Discovery Rate (FDR) at level α using the Benjamini-Hochberg procedure [3], for example, could be used.”
Algorithmic stability for adaptive data analysis (2015, methodology): “blem persists despite decades of research by statisticians on methods for preventing false discovery, such as the widely used Bonferroni Correction [Bon36, Dun61] and the Benjamini-Hochberg Procedure [BH95]. False discovery is often attributed to misuse of statistics. An alternative explanation is t”
Interpreting Models via Single Tree Approximation (2016, methodology): “In practice, we implement a scheme [Benjamini and Hochberg, 1995] to adaptively discard splits that perform far worse than the current best.”
Mushroom body output neurons encode valence and guide memory-based action selection in Drosophila (2014, methodology): “A Benjamini-Hochberg False discovery rate (FDR) test (Benjamini and Hochberg, 1995) was performed on the p-values for each Wilcoxon comparison.”
Preserving Statistical Validity in Adaptive Data Analysis (2014, background): “It is well-known that when performing multiple tests on the same data one cannot use significance levels of individual tests and instead it is necessary to control measures such as the false discovery rate [9].”
Learning Explainable Models Using Attribution Priors (2019, methodology): “…the neural network with the tissue-speciﬁc graph attribution prior captures signiﬁcantly more biologically-relevant pathways (increased number of signiﬁcant pathways after FDR correction using the Benjamini-Hochberg procedure) than a neural network without attribution priors (See Figure 2) [2].”
Interactive fingerprinting codes and the hardness of preventing false discovery (2014, methodology): “For decades statisticians have been devising methods for preventing false discovery, such as the “Bonferroni correction” Bonferroni (1936); Dunn (1961) and the widely used and highly influential method of Benjamini and Hochberg (1995) for controlling the “false discovery rate.”
False Discoveries in Mutual Fund Performance: Measuring Luck in Estimated Alphas. (2005, unknown): “(5) To select λ∗, we use the simple data-driven approach suggested by Storey (2002) and explained in detail in Appendix A. Substituting the estimate bπ0 in Equations (2), (3), and replacing E(S+γ ) with the observed proportion of significant funds in the right tail, bS+γ , we can easily estimate E(F”
A comparative study of synchrony measures for the early diagnosis of Alzheimer's disease based on EEG (2010, unknown): “different parameter settings; the Benjamini–Hochberg method (Benjamini and Hochberg, 1995) is then more accurate.”
Multiple Hypothesis Testing by Clustering Treatment Effects (2007, methodology): “…dependencies (van der Laan, Dudoit, and Pollard 2004; Storey, Taylor, and Siegmund 2004), and useful methods are available that from a list of p values produce an error rate–controlled short list of rejected null hypotheses (Storey 2003; Benjamini and Yekutieli 2001; Benjamini and Hochberg 1995).”
FAIRE (Formaldehyde-Assisted Isolation of Regulatory Elements) isolates active regulatory elements from human chromatin. (2007, methodology): “…by FAIRE has a very strong negative correlation with nucleosome occupancy (Hogan et al. 2006), as measured by comparison with nucleosome ChIP– chip experiments (Bernstein et al. 2004; Lee et al. 2004) and high-resolution mapping of nucleosomes with micrococcal nuclease digestion (Yuan et al. 2005).”