Basic assumptions, core connotations, and path methods of model modification—using confirmatory factor analysis as an example
Structural equation modeling (SEM) is a widely used statistical method in social science. However, many published articles employing SEM appear to contradict its underlying principles and assumptions, which undermines the scientific rigor of the research. Model modifications should be data-driven an...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-02-01
|
| Series: | Frontiers in Education |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/feduc.2025.1506415/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Structural equation modeling (SEM) is a widely used statistical method in social science. However, many published articles employing SEM appear to contradict its underlying principles and assumptions, which undermines the scientific rigor of the research. Model modifications should be data-driven and clearly justified, rather than arbitrarily changing the relationships between variables. Removing measurement indicators can significantly reduce discrepancies between the sample data and the model. This approach is often considered optimal for model modification. Except for certain specific models, error correlations should only be established based on theoretical support to improve the model’s goodness-of-fit. Finally, any modifications to the model should undergo cross-validation to ensure its applicability to other sample datasets. |
|---|---|
| ISSN: | 2504-284X |