Valid confidence intervals and inference in the presence of weak instruments

被引：64

作者：

Zivot, E ^{[1
]}

Startz, R ^{[1
]}

Nelson, CR ^{[1
]}

机构：

[1] Univ Washington, Seattle, WA 98195 USA

来源：

INTERNATIONAL ECONOMIC REVIEW | 1998年 / 39卷 / 04期

关键词：

D O I：

10.2307/2527355

中图分类号：

F [经济];

学科分类号：

02 ;

摘要：

We investigate confidence intervals and inference for the instrumental variables model with weak instruments. Confidence intervals based on inverting the LM, LR, and Anderson-Rubin statistics perform far better than the Wald. Performance of the LM and LR statistics is improved by a degrees-of-freedom correction in the overidentified case. The practice of 'pre-testing' by looking at the significance of the first-stage regression and then making inference based on the LM or LR statistics leads to better results than pre-testing and using the Wald statistic.

引用

页码：1119 / 1144

页数：26

共 36 条

[31]

RUBIN H, 1948, THESIS U CHICAGO

[32] EXACT SAMPLING DISTRIBUTION OF ORDINARY LEAST SQUARES AND 2-STAGE LEAST SQUARES ESTIMATORS [J].