On the so-called "Huber Sandwich Estimator" and "Robust Standard Errors"

被引：326

作者：

Freedman, David A. ^{[1
]}

机构：

[1] Univ Calif Berkeley, Dept Stat, Berkeley, CA 94720 USA

来源：

AMERICAN STATISTICIAN | 2006年 / 60卷 / 04期

关键词：

maximum likelihood; specification error; White's correction;

D O I：

10.1198/000313006X152207

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

The "Huber Sandwich Estimator" can be used to estimate the variance of the MLE when the underlying model is incorrect. If the model is nearly correct, so are the usual standard errors, and robustification is unlikely to help much. On the other hand, if the model is seriously in error, the sandwich may help on the variance side, but the parameters being estimated by the MLE are likely to be meaningless-except perhaps as descriptive statistics.

引用

页码：299 / 302

页数：4

共 16 条

[11] The effect of information on voter turnout: Evidence from a natural experiment [J].

Lassen, DD .

AMERICAN JOURNAL OF POLITICAL SCIENCE, 2005, 49 (01) :103-118

[12]

Lehmann E, 2003, THEORY POINT ESTIMAT

[13]

LEHMANN FL, 1998, ELEMENTS LARGE SAMPL

[14]

Rao C. R., 1973, Linear statistical inference and its applications, V2

[15]

van der Vaart A. W., 1998, ASYMPTOTIC STAT, DOI 10.1017/CBO9780511802256

[16] A HETEROSKEDASTICITY-CONSISTENT COVARIANCE-MATRIX ESTIMATOR AND A DIRECT TEST FOR HETEROSKEDASTICITY [J].

WHITE, H .

ECONOMETRICA, 1980, 48 (04) :817-838

← 1 2 →