Table 2. Standard deviations* of the pooled set of standardized residuals from Children's Health Study participants' datasets for each model and target flow rate.
| Target flow rate (ml/s) | ||||
| Model | 30 | 50 | 100 | 300 |
| linP >30 | NA | 1.1 | 0.8 | 0.7 |
| linT >30 | NA | 0.7 | 0.8 | 1.3 |
| quadP | 1.4 | 1.0 | 0.8 | 0.6 |
| quadT | 0.8 | 0.9 | 0.9 | 1.5 |
| nonLin | 1.0 | 0.8 | 0.7 | 0.5 |
| nonLinLog | 0.7 | 0.8 | 0.8 | 0.9 |
| nonLinLogC | 0.7 | 0.8 | 0.8 | 0.9 |
The assumption of homoscedasticity (equal variance) is satisfied if the standard deviations are approximately the same across flow rates.