On the Relationship of General Designs and Moments of Estimators in Simple Linear Regression
作者:
Saeed Maghsoodloo,
JamesN. Hool,
期刊:
IIE Transactions
(Taylor Available online 1984)
卷期:
Volume 16,
issue 2
页码: 170-176
ISSN:0740-817X
年代: 1984
DOI:10.1080/07408178408974682
出版商: Taylor & Francis Group
数据来源: Taylor
摘要:
Methods are developed for determining sampling designs in simple linear regression (SLR) when residuals are non-normal that permit use of the classical normality-based analyses. General designs (such as asymmetrical), wherein both sample size and location of independent variable values are determined, are considered. The methods are based on behavior of standardized third and fourth moments of regression estimators. All other regression assumptions are assumed to be true.
点击下载:
PDF (430KB)
返 回