Local fit evaluation of structural equation models using graphical criteria.

Abstract

Evaluation of model fit is critically important for every structural equation model and sophisticated methods have been developed for this task. Among them are the χ 2 goodness-of-fit test, decomposition of the χ 2 , derived measures like the popular RMSEA or CFI, or inspection of residuals or modification indices. Many of these methods provide a global approach to model fit evaluation: A single index is computed that quantifies the fit of the entire SEM to the data. In contrast, graphical criteria like d-separation or trek-separation allow to derive implications that can be used for local fit evaluation, an approach that is hardly ever applied. We provide an overview of local fit evaluation from the viewpoint of SEM practitioners. In the presence of model misfit, local fit evaluation can potentially help in pinpointing where the problem with the model lies. For models that do fit the data, local tests can identify the parts of the model that are corroborated by the data. Local tests can also be conducted before a model is fitted at all, and they can be used even for models that are globally under-identified. We discuss appropriate statistical local tests, and provide applied examples. We also present novel software in R that automates this type of local fit evaluation.

Separators and adjustment sets in causal graphs: Complete criteria and an algorithmic framework

“…In such cases, even simpler basis sets can be derived. For instance, Pearl and Meshkat [31] discuss basis sets for hierarchical linear regression models (also known as path models or structural equation models [32]). We here consider the vertices V = {X 1 , .…”

Preventive interventions for children in organized team sport tackling aggression: Results from the pilot evaluation of “Fairplayer.Sport”

“…Finally, we did not include local tests of misfit of our models. As Thoemmes, Rosseel, and Textor (2018) recommended, local tests of misfit might serve as an addition to global fit indices. They might be useful especially in situations when proposed models did not converge and after the fitting of models to actual data did result in a bad overall fit.…”

Description, prediction and causation: Methodological challenges of studying child and adolescent development

“…Due to space limitations, we are not able to discuss the ins and outs of causal inference research in more detail. A more comprehensive treatment of this topic would include a discussion of the actual act of estimating the causal effect (e.g., through the use of techniques such as inverse probability weighting, matching, or g-estimation), and a discussion of the ways researchers can test the assumptions on which their causal model is based (such as sensitivity analysis (e.g., Hernán and Robins, 2020 ), and local (mis)fit analysis (e.g., Shipley, 2003 ; Thoemmes et al, 2018 )). For a more elaborate discussion of the diverse steps in causal research, we refer the reader to the recent work of Grosz et al (2020 ; see also Ahern, 2018 ; Hernán and Robins, 2020 ).…”