Correlated Measurement Errors and the Least Squares Estimator of the Regression Coefficient
作者:
JohnJ. Chai,
期刊:
Journal of the American Statistical Association
(Taylor Available online 1971)
卷期:
Volume 66,
issue 335
页码: 478-483
ISSN:0162-1459
年代: 1971
DOI:10.1080/01621459.1971.10482288
出版商: Taylor & Francis Group
数据来源: Taylor
摘要:
This article explores, for a set of survey conditions, the effect of correlated measurement error on the ordinary least squares estimator of the regression coefficient for a finite bivariate population, when both variables are subject to correlated measurement error. The sample estimates for some selected housing variables show that the average of the correlated measurement error components have little effect, but the simple response variance and covariance have pronounced effect, on the ordinary regression estimator. In particular, the estimates of the simple response covariance studied are far from negligible. Finally, the article examines the sensitivity of the bias factor of the estimator.
点击下载:
PDF (562KB)
返 回