Table 4.
Prediction of the scores on the original scales. Results for the shortened scales performing best for each of the four domains.
| Scale | Items in short scale | Correct (%)a | Mean diff. (SD)b | Correlationc | Kappac |
|---|---|---|---|---|---|
| BRBI | |||||
| Training set | 10,11,12 | 87.10 | 1.74 (5.40) | 0.982 | 0.90 |
| Testing set | 86.44 | 1.94 (5.36) | 0.983 | 0.89 | |
| BRST | |||||
| Training set | 2,3,6,8 | 70.38 | 0.77 (5.49) | 0.921 | 0.74 |
| Testing set | 69.76 | 0.80 (5.69) | 0.916 | 0.73 | |
| BRBS | |||||
| Training set | 20,21,22 | 100.00 | 0.15 (4.79) | 0.963 | 1.00 |
| Testing set | 100.00 | 0.18 (5.01) | 0.960 | 1.00 | |
| BRAS | |||||
| Training set | 17,18 | 100.00 | −2.04 (6.79) | 0.951 | 1.00 |
| Testing set | 100.00 | −2.17 (6.97) | 0.951 | 1.00 | |
aPercent correctly predicted scale scores; bMean difference between predicted and observed scale scores; cCorrelation and weighted kappa between predicted and observed scores.