Hypothesis testing near singularities and boundaries

DOI10.1214/19-EJS1576MaRDI QIDQ85619zbMATH OpenWikidataFDO

Authors Jonathan D. Mitchell, Elizabeth S. Allman, John A. Rhodes, Jonathan D. Mitchell, Elizabeth S. Allman

Publication date 1 January 2019

Published in Electronic Journal of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1806.08458, https://projecteuclid.org/euclid.ejs/1561687407

zbMATH Keywords

hypothesis testingboundarychi-squaredcoalescentlikelihood ratio statisticphylogenomicssingularity

Mathematics Subject Classification ID

Parametric hypothesis testing (62F03) Applications of statistics to biology and medical sciences; meta analysis (62P10) Asymptotic properties of parametric tests (62F05) Problems related to evolution (92D15) Approximations to statistical distributions (nonasymptotic) (62E17)

Abstract: The likelihood ratio statistic, with its asymptotic $ch{i}^2$ distribution at regular model points, is often used for hypothesis testing. At model singularities and boundaries, however, the asymptotic distribution may not be $ch{i}^2$, as highlighted by recent work of Drton. Indeed, poor behavior of a $ch{i}^2$ for testing near singularities and boundaries is apparent in simulations, and can lead to conservative or anti-conservative tests. Here we develop a new distribution designed for use in hypothesis testing near singularities and boundaries, which asymptotically agrees with that of the likelihood ratio statistic. For two example trinomial models, arising in the context of inference of evolutionary trees, we show the new distributions outperform a $ch{i}^2$.

Recommendations

• Likelihood ratio tests and singularities
• On the asymptotic behaviour of the pseudolikelihood ratio test statistic with boundary problems
• On the asymptotic distribution of likelihood ratio test when parameters lie on the boundary
• A modified likelihood ratio statistic for some nonregular models
• The likelihood ratio test for simple tree order: A useful asymptotic expansion

Cites work

• scientific article; zbMATH DE number 3911472 (Why is no real title available?)
• Asymptotic Properties of Maximum Likelihood Estimators and Likelihood Ratio Tests Under Nonstandard Conditions
• Asymptotic Statistics
• Asymptotic distribution of test statistics in the analysis of moment structures under inequality constraints
• Asymptotic properties of maximum likelihood estimates in the mixed model of the analysis of variance
• Asymptotic size and a problem with subsampling and with the \(m\) out of \(n\) bootstrap
• Bonferroni-based size-correction for nonstandard testing problems
• Coalescent theory. An introduction
• Constrained Statistical Inference
• Hybrid and Size-Corrected Subsampling Methods
• Hypothesis tests for phylogenetic quartets, with applications to coalescent-based species tree inference
• Identifying the rooted species tree from the distribution of unrooted gene trees under the coalescent
• Inconsistency of the Bootstrap when a Parameter is on the Boundary of the Parameter Space
• Incorrect asymptotic size of subsampling procedures based on post-consistent model selection estimators
• Likelihood Inference for a Class of Latent Markov Models under Linear Hypotheses on the Transition Probabilities
• Likelihood ratio tests and singularities
• NIST handbook of mathematical functions
• On the Distribution of the Likelihood Ratio
• P Values Maximized Over a Confidence Set for the Nuisance Parameter
• Pearson's X 2 and the Loglikelihood Ratio Statistic G 2 : A Comparative Review
• Positive Quadrant Dependence and Marginal Modeling in Two-Way Tables With Ordered Margins
• Probability and stochastic processes
• The Large-Sample Distribution of the Likelihood Ratio for Testing Composite Hypotheses
• The probability of topological concordance of gene trees and species trees.

Cited in

(8)

• On the validity of the likelihood ratio and maximum likelihood methods
• Power of the likelihood ratio test for models of DNA base substitution
• MSCquartets
• The tree of blobs of a species network: identifiability under the coalescent
• Hypothesis Testing in Integral Geometry
• A note on likelihood ratio tests for models with latent variables
• Identifiability of species network topologies from genomic sequences using the logDet distance
• Likelihood ratio tests and singularities

Describes a project that uses

Uses Software

• DLMF

This page was built for publication: Hypothesis testing near singularities and boundaries

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q85619)