Table 2.
Numbers of factors chosen through various selection criteria
| Criterion | Number of factors | ||
|---|---|---|---|
| CDI | FACES | DQOLY | |
| Recommended no. | 1 | 2 | 3 |
| Maximum-likelihood extraction | |||
| Smallest no. with P-value > 0.05 for test that no. of factors is sufficient | 5 | 8 | 13 |
| Minimum AIC | 5 | 9 | 9 |
| Minimum BIC | 1 | 2 | 2 |
| Maximum LCVa | 1 | 2 | 3 |
| Principal component extractionb | |||
| Largest no. with eigenvalue > 1 | 10 | 8 | 15 |
| Largest no. with % of total eigenvalue > 5% | 5 | 4 | 3 |
| Largest no. with % of total eigenvalue > 10% | 1 | 1 | 2 |
| Smallest no. with cumulative % of total eigenvalue > 70% | 10 | 10 | 14 |
| Smallest no. with cumulative % of total eigenvalue > 80% | 14 | 14 | 8 |
| One-step principal factor extractionb | |||
| Largest no. with eigenvalue > 1 | 4 | 4 | 11 |
| Largest no. with % of total eigenvalue > 5% | 6 | 6 | 4 |
| Largest no. with % of total eigenvalue > 10% | 4 | 2 | 2 |
| Smallest no. with cumulative % of total eigenvalue > 70% | 4 | 4 | 8 |
| Smallest no. with cumulative % of total eigenvalue > 80% | 5 | 5 | 11 |
a
Using 10 folds for CDI and FACES and 15 folds for DQOLY.
b
Applied to the correlation matrix.