DESIGN AND OPTIMIZATION FOR HAZARDOUS JAROSITE WASTE RECYCLING IN ENVIRONMENTALLY SUITABLE COMPOSITE PRODUCTS USING RESPONSE SURFACE METHODOLOGY - Management of Hazardous Residues Containing Cd - page 75

Table 23. Model summary statistics for response density

r 2 Adjusted r 2 Predicted r 2

Source Std.

Dev

Press

Linear 0.072 0.7744 0.7334 0.6004 0.10

Quadratic 0.025 0.9796 0.9669 0.9256 0.019

Special Cubic 0.027 0.9799 0.9628 0.8776 0.031

Cubic 0.029 0.9869 0.9576

Table 24. Model summary statistics for response cadmium leaching
Source Std.

Dev.

Adjusted
R-Squared

Predicted
R-quared

R-Squared

Press

Linear 0.080 0.8550 0.8286 0.7854 0.10

Suggested

Quadratic 0.066 0.9281 0.8831 0.8383 0.078

Special
Cubic

0.071 0.9282 0.8667 0.7601 0.12

Cubic 5.545E-

003

0.9997 0.9992

Suggested

Table 25. Model summary statistics for response lead leaching

Source Std.

Dev.

Adjusted
R-Squared

Predicted
R-Squared

R-Squared Press

Linear 3.30 0.6752 0.6161 0.5042 182.54

Quadratic 1.82 0.9277 0.8825 0.8430 57.82 Suggested

Special
Cubic

1.78 0.9398 0.8882 0.7908 77.02

Cubic 0.069 0.9999 0.9998 Suggested

From Table 26 of ANOVA for mixture special cubic model for response
compressive strength Prob > F value implies the model is significant and there
is only a 0.01% chance that a "Model F-Value" this large could occur due to
noise. However. When "Prob > F" less than 0.05 indicate that the model terms
are significant. And values greater than 0.1 indicate the model terms are not
significant. In this study Linear Mixture Components, AB, AC, BC, ABC are
significant model terms. The analysis showed that the "Lack of Fit F-value" of
0.38 is not significant relative to the pure error and there is a 77.06% chance
that a "Lack of Fit F-value" this large could occur due to noise. For model
fitting, Non-significant lack of fit is good and desired one. For other response