The Behrens-Fisher problem with covariates and baseline adjustments

DOI10.1007/S00184-019-00729-2zbMATH Open1436.62322arXiv1808.08986OpenAlexW2963635442WikidataQ127490029 ScholiaQ127490029MaRDI QIDQ2303752FDOQ2303752

Authors: Cong Cao, Markus Pauly, F. Konietschke 

Publication date: 5 March 2020

Published in: Metrika (Search for Journal in Brave)

Abstract: The Welch-Satterthwaite t-test is one of the most prominent and often used statistical inference method in applications. The method is, however, not flexible with respect to adjustments for baseline values or other covariates, which may impact the response variable. Existing analysis of covariance methods are typically based on the assumption of equal variances across the groups. This assumption is hard to justify in real data applications and the methods tend to not control the type-1 error rate satisfactorily under variance heteroscedasticity. In the present paper, we tackle this problem and develop unbiased variance estimators of group specific variances, and especially of the variance of the estimated adjusted treatment effect in a general analysis of covariance model. These results are used to generalize the Welch-Satterthwaite t-test to covariates adjustments. Extensive simulation studies show that the method accurately controls the nominal type-1 error rate, even for very small sample sizes, moderately skewed distributions and under variance heteroscedasticity. A real data set motivates and illustrates the application of the proposed methods.

Full work available at URL: https://arxiv.org/abs/1808.08986

Recommendations

• Bayesian and non-Bayesian solutions to analysis of covariance models under heteroscedasticity
• Heteroscedastic ANOVA: old \(p\) values, new views
• A review of the Behrens-Fisher problem and some of its analogs: does the same size fit all?
• Accuracy of p-values of approximate tests in testing for equality of means under unequal variances
• An improvement in the Welch approximate \(t\) test

zbMATH Keywords

heteroscedasticitysemiparametric methodsnon-normalityANCOVA designs

Mathematics Subject Classification ID

Nonparametric hypothesis testing (62G10) Linear regression; mixed models (62J05)

Cites Work

• Box-Type Approximations in Nonparametric Factorial Designs
• Title not available (Why is that?)
• Some Theorems on Quadratic Forms Applied in the Study of Analysis of Variance Problems, I. Effect of Inequality of Variance in the One-Way Classification
• A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity
• Bootstrap procedures under some non-i.i.d. models
• Jackknife, bootstrap and other resampling methods in regression analysis
• Bootstrap and wild bootstrap for high dimensional linear models
• Title not available (Why is that?)
• Title not available (Why is that?)
• Bootstrapping heteroskedasticity consistent covariance matrix estimator
• Weak convergence of the wild bootstrap for the Aalen-Johansen estimator of the cumulative incidence function of a competing risk

Cited In (1)

• Analysis of covariance under variance heteroscedasticity in general factorial designs

This page was built for publication: The Behrens-Fisher problem with covariates and baseline adjustments

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2303752)