Table 2.
(a) Coefficient estimates for different model terms appearing in the final equation for each response and their significance, together with each model’s lack of fit, type, and evaluation.
| a) | |
|
|
||||||
|---|---|---|---|---|---|---|---|---|---|
| |
1/PS |
ZP |
EE |
||||||
| Component | Coeff. Estimate | Df | p-value | Coeff. Estimate | df | p-value | Coeff. Estimate | df | p-value |
| Linear mixture: | 2 | <0.0001* | 2 | 0.003* | 2 | <0.0001* | |||
| A: Span 20 | 0.0043 | −36.23 | 85.96 | ||||||
| B: Cholesterol | 0.0050 | −38.44 | 85.12 | ||||||
| C: Cremophor RH | 0.0100 | −28.49 | 72.53 | ||||||
| AB | −0.0023 | 1 | 0.2109 | — | — | — | — | ||
| AC | 0.0129 | 1 | 0.0005* | — | — | — | — | ||
| BC | — | — | −23.16 | 1 | 0.0193* | −12.02 | 1 | 0.0554 | |
| ABC | −0.1482 | 1 | 0.0001* | 1964.67 | 1 | < 0.0001* | — | — | |
| Lack of fit | 3 | 0.3901 | 4 | 0.0564 | 5 | 0.1049 | |||
| Pure error | 4 | 4 | 4 | ||||||
| Model type | Reduced cubic | Reduced cubic | Reduced quadratic | ||||||
| Model evaluation | |||||||||
| RMSE | 0.000 | 1.826 | 1.344 | ||||||
| CV% | 8.155 | 6.735 | 1.658 | ||||||
| R2 | 0.980 | 0.991 | 0.953 | ||||||
| Adjusted R2 | 0.966 | 0.987 | 0.938 | ||||||
| Prediction R2 | 0.936 | 0.978 | 0.903 | ||||||