They challenge the idea that the number of parameters solely determines model complexity.
Their study reveals that a model’s functional form (how parameters are arranged) is critical to fit diverse data patterns, impacting its generalizability and interpretability.
Their study reveals that a model’s functional form (how parameters are arranged) is critical to fit diverse data patterns, impacting its generalizability and interpretability.
Comments
1. Bifactor models with fewer specific factors have higher FP, making them more likely to overfit data.
2. Exploratory factor analysis (EFA) models outperform confirmatory bifactor models in fitting diverse data, even with the same parameters.