Assessing the adequacy of structural equation models: Golden rules and editorial policies

被引:62
|
作者
Goffin, Richard D. [1 ]
机构
[1] Univ Western Ontario, Dept Psychol, Social Sci Ctr, London, ON N6A 5C2, Canada
关键词
Structural Equation Modeling; goodness of fit indices; assessment of fit; chi-square; partial least squares;
D O I
10.1016/j.paid.2006.09.019
中图分类号
B84 [心理学];
学科分类号
04 ; 0402 ;
摘要
Recommendations for enhancing the present journal's editorial policy with regards to the publication of Structural Equation Modeling (SEMing) studies have been provided in Barrett (2007). Wisely, these recommendations oppose strict reliance on "cut-off' values that have been proposed for well-known Approximate Fit Indices (AFIs). However, the present article critically evaluates other aspects of the recommended editorial policy in light of recent literature and finds a number of weaknesses: AFIs are ignored altogether, predictive accuracy suggestions do not take advantage of the SEMing literature, the chi-square test is overemphasized, power in SEMing analyses is underemphasized, and a number of important aspects of model assessment are not considered. Suggestions for a more comprehensive editorial policy regarding SEMing articles are provided. (c) 2006 Elsevier Ltd. All rights reserved.
引用
收藏
页码:831 / 839
页数:9
相关论文
共 50 条