The problem of allowing correlated errors in structural equation modeling: concerns and considerations

Computational Methods in Social Sciences

卷:3

Richard HERMIDA¹ 

[1] George Mason University, 3575 Owasso Street, Shoreview, MN, USA, 55126;

关键词: Structural Equation Modeling;    Confirmatory Factor Analysis;    Measurement;    Research Methods;    Statistics.;   

DOI  :  

来源: DOAJ

【 摘 要 】

Results of structural equation models may be negated by inappropriate methodological procedures. Model fit is known to be improved by the addition of pathways. Some pathways are added due to modification indices. These a-theoretical pathways will improve model fit at the expense of theory and reduction in parameter value replication. Furthermore, some additions to the model like correlating measurement errors are usually theoretically unjustifiable. This quantitative review examines the frequency of correlating measurement errors and examines the reasons, if any, for having these pathways in the model. Additionally, this quantitative review examines the consequences of correlating measurement errors in structural equation modeling.

【 授权许可】

Unknown