Key points are not available for this paper at this time.
Abstract Four current controversies in structural equation modelling are reviewed; the “causal” nature of structural equation models, the assessment of global fit, the use of specification searches to enhance model fit, and the call for a return to “simpler” forms of analysis. Within each area the nature of the controversy is reviewed and suggestions made to improve future practice.
E. Kevin Kelloway (Mon,) studied this question.