Calibrated percentile double bootstrap for robust linear regression inference

DOI10.5705/SS.202016.0546zbMATH Open1406.62076arXiv1511.00273OpenAlexW2963009677MaRDI QIDQ4558601FDOQ4558601

Authors:

Publication date: 22 November 2018

Published in: STATISTICA SINICA (Search for Journal in Brave)

Abstract: We consider inference for the parameters of a linear model when the covariates are random and the relationship between response and covariates is possibly non-linear. Conventional inference methods such as z-intervals perform poorly in these cases. We propose a double bootstrap-based calibrated percentile method, perc-cal, as a general-purpose CI method which performs very well relative to alternative methods in challenging situations such as these. The superior performance of perc-cal is demonstrated by a thorough, full-factorial design synthetic data study as well as a real data example involving the length of criminal sentences. We also provide theoretical justification for the perc-cal method under mild conditions. The method is implemented in the R package `perccal', available through CRAN and coded primarily in C++, to make it easier for practitioners to use.

Full work available at URL: https://arxiv.org/abs/1511.00273

Recommendations

• Unusual properties of bootstrap confidence intervals in regression problems
• Bootstrapping robust estimates of regression
• Bootstrapping MM-estimators for linear regression with fixed designs
• Calibration Intervals in Linear Regression Models
• Can we trust the bootstrap in high-dimensions? The case of linear models

zbMATH Keywords

bootstrapconfidence intervalsEdgeworth expansionresamplingsecond-order correctnessrobust linear regression inference

Mathematics Subject Classification ID

Nonparametric robustness (62G35) Linear regression; mixed models (62J05) Nonparametric tolerance and confidence regions (62G15)

Cited In (6)

• perccal
• Assumption Lean Regression
• Models as approximations. I. Consequences illustrated with linear regression
• Rejoinder: Models as approximations
• Content-adjusted tolerance intervals via bootstrap calibration
• Improving coverage probabilities for parametric tolerance intervals via bootstrap calibration

Uses Software

• R
• CRAN

This page was built for publication: Calibrated percentile double bootstrap for robust linear regression inference

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4558601)