Translator Disclaimer
February 2014 New heteroskedasticity-robust standard errors for the linear regression model
Francisco Cribari-Neto, Maria da Glória A. Lima
Braz. J. Probab. Stat. 28(1): 83-95 (February 2014). DOI: 10.1214/12-BJPS196

Abstract

Linear regressions fitted to cross-sectional data oftentimes display heteroskedasticity, that is, error variances that are not constant. A common modeling strategy consists of estimating the regression parameters by ordinary least squares and then performing hypothesis testing inference using standard errors that are robust to heteroskedasticity. These tests have the correct size asymptotically regardless of whether the error variances are constant. In finite samples, however, they can be quite size-distorted. In this paper, we propose new heteroskedasticity-consistent covariance matrix estimators that deliver more reliable testing inferences in samples of small sizes.

Citation

Download Citation

Francisco Cribari-Neto. Maria da Glória A. Lima. "New heteroskedasticity-robust standard errors for the linear regression model." Braz. J. Probab. Stat. 28 (1) 83 - 95, February 2014. https://doi.org/10.1214/12-BJPS196

Information

Published: February 2014
First available in Project Euclid: 5 February 2014

zbMATH: 06291462
MathSciNet: MR3165430
Digital Object Identifier: 10.1214/12-BJPS196

Rights: Copyright © 2014 Brazilian Statistical Association

JOURNAL ARTICLE
13 PAGES


SHARE
Vol.28 • No. 1 • February 2014
Back to Top