CONCERNING LEAST SQUARES ANALYSIS OF EXPERIMENTAL DATA

被引：336

作者：

OVERALL, JE

SPIEGEL, DK

机构：

[1] U. Texas, Medical Branch, Galveston

来源：

PSYCHOLOGICAL BULLETIN | 1969年 / 72卷 / 05期

关键词：

least squares regression method vs. analysis of variance; balanced vs. unequal cell frequency designs;

D O I：

10.1037/h0028109

中图分类号：

B84 [心理学];

学科分类号：

04 ; 0402 ;

摘要：

Several different least squares methods appear superficially similar to the conventional analysis of variance and yield identical results when applied to data from balanced experimental designs. When applied to complex problems involving unequal and disroportionate cell frequencies, the methods yield quite different results. In such cases, the effects tested by the different methods are different; hence, interpretation of results should depend upon the method used. (PsycINFO Database Record (c) 2006 APA, all rights reserved). © 1969 American Psychological Association.

引用

页码：311 / &

共 13 条

[1]

BOTTENBERG R, 1960, APPLIED MULTIPLE LIN

[2] MULTIPLE CLASSIFICATION ANALYSIS FOR ARBITRARY EXPERIMENTAL ARRANGEMENTS [J].