Table 3
The 
of alternative least-squares linear regression models in the training set
| Index | Methods | Haigis | Hoffer Q | Holladay1 | SRK/T |
| 1 | Formula LR | 0.377 | 0.541 | 0.579 | 0.394 |
| 2 | ML LR | 0.376 | 0.442 | 0.426 | 0.378 |
| 3 | Formula & ML LR | 0.425 | 0.622 | 0.605 | 0.482 |
The outlier cases were removed before calculating the above values. The largest
among three methods is marked in bold for each formula. P-values of all correlations were < 0.05.
