A Practical Geometry of Data-Model Fit and Modification

• In Event: Past, Present, and Future of Model Fit: Modern Solutions to Historical Challenges

Abstract

Structural equation model modification following unsatisfactory data-model fit, while common practice, typically involves little to no understanding of the interrelations/redundancy among potential modifications involving fixed/constrained parameters nor those modifications’ potential impact on free parameters already an active part of a model. The current work uses a geometric framing to (1) help better understand data-model fit, (2) explicate the nature and degree of relations among both constrained and free model parameters, and (3) present a software package that generates several modeling tools to help visualize model fit and modification, and in turn embed this within a more thoughtful and informed post hoc modeling process.

Authors

• Christian Meyer, University of Maryland

• Gregory R. Hancock, University of Maryland