Table 3 Fit measures for the different models used to predict the EQ-5D-Y utilities based on YGTSS scores. (n = 144)
Models without demographics | Models with demographics | |||
|---|---|---|---|---|
Model 1 Scale total scores | Model 2 Subscale scores | Model 3 Model 1 plus demographics with P < 0.05 | Model 4 Model 2 plus demographics with P < 0.05 | |
GLM | ||||
Variables | YT | YM + YV | YT + Sex | YM + YV + Sex |
Parameters, β (SE) | ||||
Intercept | 0.9881 (0.0139) | 0.9856 (0.0164) | 1.0201 (0.0220) | 1.0179 (0.0238) |
YT | −0.0035 (0.0008) | −0.0036 (0.0008) | ||
YM | −0.0032 (0.0013) | −0.0033 (0.0013) | ||
YV | −0.0037 (0.0010) | −0.0038 (0.0010) | ||
Sex | −0.0357 (0.0191) | −0.0356 (0.0191) | ||
Predicted index | ||||
Mean (SD) | 0.9313 (0.0302) | 0.9313 (0.0303) | 0.9313 (0.0325) | 0.9313 (0.0327) |
Range | 0.8410 to 0.9881 | 0.8404 to 0.9856 | 0.8346 to 0.9988 | 0.8341 to 0.9982 |
Fit measures | ||||
AIC | −321.80 | −320.04 | −323.34 | −321.56 |
BIC | −312.89 | −308.16 | −311.46 | −306.71 |
Adjusted R2 | 0.125 | 0.120 | 0.140 | 0.135 |
Predictive accuracy | ||||
MAE | 0.0599 | 0.0601 | 0.0583 | 0.0585 |
RMSE | 0.0775 | 0.0775 | 0.0766 | 0.0765 |
Beta | ||||
Variables | YT | YM + YV | YT + Sex | YM + YV + Sex |
Parameters, β (SE) | ||||
Intercept | 3.0751 (0.1661) | 3.1013 (0.1963) | 3.4553 (0.2696) | 3.5025 (0.2913) |
YT | −0.0281 (0.0088) | −0.0290 (0.0087) | ||
YM | −0.0314 (0.0156) | −0.0342 (0.0156) | ||
YV | −0.0259 (0.0123) | −0.0256 (0.0122) | ||
Sex | −0.4149 (0.2357) | −0.4206 (0.2358) | ||
Predicted index | ||||
Mean (SD) | 0.9304 (0.0175) | 0.9304 (0.0176) | 0.9303 (0.0193) | 0.9302 (0.0194) |
Range | 0.8694 to 0.9559 | 0.8697 to 0.9569 | 0.8610 to 0.9638 | 0.8613 to 0.9643 |
Fit measures | ||||
AIC | −533.58 | −531.70 | −535.02 | −533.22 |
BIC | −524.67 | −519.82 | −523.14 | −518.37 |
Adjusted R2 | 0.101 | 0.094 | 0.112 | 0.105 |
Predictive accuracy | ||||
MAE | 0.0596 | 0.0595 | 0.0586 | 0.0585 |
RMSE | 0.0775 | 0.0775 | 0.0768 | 0.0768 |