Small-scale inference: empirical Bayes and confidence methods for as few as a single comparison

DOI10.1111/INSR.12064MaRDI QIDQ4968612zbMATH OpenOpenAlexWikidataFDO

Authors David R. Bickel

Publication date 16 July 2019

Published in International Statistical Review (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1104.0341

zbMATH Keywords

multiple testingempirical Bayesconstrained maximum likelihood estimationconfidence distributionfoundations of statisticsLindley's paradoxrestricted parameter spacelocal false discovery ratemultiple comparison procedureobserved confidence level

Mathematics Subject Classification ID

Foundations and philosophical topics in statistics (62A01) Empirical decision procedures; empirical Bayes procedures (62C12) Paired and multiple comparisons; multiple testing (62J15)

Abstract: By restricting the possible values of the proportion of null hypotheses that are true, the local false discovery rate (LFDR) can be estimated using as few as one comparison. The proportion of proteins with equivalent abundance was estimated to be about 20% for patient group I and about 90% for group II. The simultaneously-estimated LFDRs give approximately the same inferences as individual-protein confidence levels for group I but are much closer to individual-protein LFDR estimates for group II. Simulations confirm that confidence-based inference or LFDR-based inference performs markedly better for low or high proportions of true null hypotheses, respectively.

Recommendations

• Simple estimators of false discovery rates given as few as one or two \(p\)-values without strong parametric assumptions
• Estimators of the local false discovery rate designed for small numbers of tests
• Empirical Bayes methods for estimation and confidence intervals in high-dimensional problems
• Empirical Bayes interval estimates that are conditionally equal to unadjusted confidence intervals or to default prior credibility intervals
• Confidence distributions applied to propagating uncertainty to inference based on estimating the local false discovery rate: a fiducial continuum from confidence sets to empirical Bayes set estimates as the number of comparisons increases

Cited in

(11)

• Empirical Bayes interval estimates that are conditionally equal to unadjusted confidence intervals or to default prior credibility intervals
• Correcting false discovery rates for their bias toward false positives
• Large-scale inference. Empirical Bayes methods for estimation, testing, and prediction
• Model fusion and multiple testing in the likelihood paradigm: shrinkage and evidence supporting a point null hypothesis
• Simple estimators of false discovery rates given as few as one or two \(p\)-values without strong parametric assumptions
• Confidence distributions applied to propagating uncertainty to inference based on estimating the local false discovery rate: a fiducial continuum from confidence sets to empirical Bayes set estimates as the number of comparisons increases
• Blending Bayesian and frequentist methods according to the precision of prior information with applications to hypothesis testing
• Estimators of the local false discovery rate designed for small numbers of tests
• Empirical Bayes methods for estimation and confidence intervals in high-dimensional problems
• Sharpen statistical significance: evidence thresholds and Bayes factors sharpened into Occam's razor
• Interval estimation, point estimation, and null hypothesis significance testing calibrated by an estimated posterior probability of the null hypothesis

This page was built for publication: Small-scale inference: empirical Bayes and confidence methods for as few as a single comparison

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4968612)