Key findings:
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.