Higher-order accuracy of multiscale-double bootstrap for testing regions

DOI10.1016/J.JMVA.2014.05.007MaRDI QIDQ2252895zbMATH OpenOpenAlexFDO

Authors Hidetoshi Shimodaira

Publication date 24 July 2014

Published in Journal of Multivariate Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1312.6348

zbMATH Keywords

approximately unbiased testsbias correctionscaling-lawmean curvatureiterated bootstrapfourth-order accuracybootstrap resampling

Mathematics Subject Classification ID

Nonparametric hypothesis testing (62G10) Nonparametric statistical resampling methods (62G09) Hypothesis testing in multivariate analysis (62H15)

Abstract: We consider hypothesis testing for the null hypothesis being represented as an arbitrary-shaped region in the parameter space. We compute an approximate p-value by counting how many times the null hypothesis holds in bootstrap replicates. This frequency, known as bootstrap probability, is widely used in evolutionary biology, but often reported as biased in the literature. Based on the asymptotic theory of bootstrap confidence intervals, there have been some new attempts for adjusting the bias via bootstrap probability without direct access to the parameter value. One such an attempt is the double bootstrap which adjusts the bias by bootstrapping the bootstrap probability. Another new attempt is the multiscale bootstrap which is similar to the m-out-of-n bootstrap but very unusually extrapolating the bootstrap probability to $m=-n$. In this paper, we employ these two attempts at the same time, and call the new procedure as multiscale-double bootstrap. By focusing on the multivariate normal model, we investigate higher-order asymptotics up to fourth-order accuracy. Geometry of the region plays important roles in the asymptotic theory. It was known in the literature that the curvature of the boundary surface of the region determines the bias of bootstrap probability. We found out that the curvature-of-curvature determines the remaining bias of double bootstrap. The multiscale bootstrap removes these biases. The multiscale-double bootstrap is fourth order accurate with coverage probability erring only $O(n^{-2})$, and it is robust against computational error of parameter estimation used for generating bootstrap replicates from the null distribution.

Recommendations

• Approximately unbiased tests of regions using multistep-multiscale bootstrap resampling
• Testing regions with nonsmooth boundaries via multiscale bootstrap
• Asymptotic theories of multiscale bootstrap
• A bias correction and acceleration approach for the problem of regions
• Frequentist and Bayesian measures of confidence via multiscale bootstrap for testing three regions

Cites work

• scientific article; zbMATH DE number 708500 (Why is no real title available?)
• scientific article; zbMATH DE number 1057566 (Why is no real title available?)
• A Weighted Bootstrap Approach to Bootstrap Iteration
• Approximately unbiased tests of regions using multistep-multiscale bootstrap resampling
• Asymptotic iterated bootstrap confidence intervals
• Better Bootstrap Confidence Intervals
• Bootstrap Recycling: A Monte Carlo Alternative to the Nested Bootstrap
• Bootstrap confidence intervals for a class of parametric problems
• Bootstrap confidence intervals. With comments and a rejoinder by the authors
• Bootstrap confidence levels for phylogenetic trees
• Bootstrapping phylogenies: Large deviations and dispersion effects
• Calibrating Confidence Coefficients
• Importance sampling and the nested bootstrap
• Large sample confidence regions based on subsamples under minimal assumptions
• Local sufficiency
• More accurate confidence intervals in exponential families
• Notions of Limiting P Values Based on Data Depth and Bootstrap
• On Bootstrap Iteration for Coverage Correction in Confidence Intervals
• On the bootstrap and confidence intervals
• Prepivoting to reduce level error of confidence sets
• Simulation-Extrapolation Estimation in Parametric Measurement Error Models
• Simultaneous confidence intervals for all distances from the best
• Testing regions with nonsmooth boundaries via multiscale bootstrap
• The bootstrap and Edgeworth expansion
• The problem of regions

Cited in

(6)

• Frequentist and Bayesian measures of confidence via multiscale bootstrap for testing three regions
• A bias correction and acceleration approach for the problem of regions
• Second-order accuracy of depth-based bootstrap confidence regions
• Approximately unbiased tests of regions using multistep-multiscale bootstrap resampling
• Testing regions with nonsmooth boundaries via multiscale bootstrap
• Mathematical proof of the third order accuracy of the speedy double bootstrap method

Describes a project that uses

Uses Software

• CONSEL
• bootstrap
• pvclust

This page was built for publication: Higher-order accuracy of multiscale-double bootstrap for testing regions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2252895)