Table 6. Analysis of variance (ANOVA) of central composite design for improving biomass of L. brevis SCML 432
| Source | Coefficient estimate | Sum of squares | Degree of freedom | Mean square | Standard error | F-value | p-valueProb>F1) |
| Model | 8.234792 | 116.9112 | 9 | 12.99014 | 0.166718 | 204.4687 | 9.13E-07 |
| X1-Molasses | 2.675625 | 114.5435 | 1 | 114.5435 | 0.063013 | 1802.95 | 1.14E-08 |
| X2-Yeast ext. | -0.10312 | 0.170155 | 1 | 0.170155 | 0.063013 | 2.678292 | 0.15284 |
| X3-Maltose | 0.276875 | 1.226553 | 1 | 1.226553 | 0.063013 | 19.30632 | 0.004599 |
| X1X22) | -0.05458 | 0.023834 | 1 | 0.023834 | 0.089115 | 0.375154 | 0.562686 |
| X1X3 | -0.12792 | 0.130903 | 1 | 0.130903 | 0.089115 | 2.060455 | 0.201176 |
| X2X3 | 0.29625 | 0.702113 | 1 | 0.702113 | 0.089115 | 11.05147 | 0.015918 |
| X12 | -0.0325 | 0.0169 | 1 | 0.0169 | 0.063013 | 0.266011 | 0.624467 |
| X22 | 0.013333 | 0.002844 | 1 | 0.002844 | 0.063013 | 0.044773 | 0.839427 |
| X32 | 0.050833 | 0.041345 | 1 | 0.041345 | 0.063013 | 0.650775 | 0.45062 |
| Residual | | 0.381187 | 6 | 0.063531 | | | |
| Lack of Fit | | 0.381137 | 5 | 0.076227 | | 1524.548 | 0.019442 |
| Pure Error | | 5E-05 | 1 | 5E-05 | | | |
| Cor. Total3) | | 117.2924 | 15 | | | | |
| Standard deviation | 0.25 | Press4) | 3.09 |
| Mean | 8.27 | R-square | 0.99 |
| Coefficient of variation | 3.05 | Adjusted R-square | 0.99 |
| | Adequate precision | 53.710 |
Response model was suggested quadratic model through the lack of fit tests.
1)P-value, probability distribution value. p-value less than 0.05 indicate that the term is significant.
2)X1X2, X1X3, X2X3 represent the interaction effect of variables X1, X2, and X3. X12, X22, and X32 are the squared effects of the variables.
3)Cor Total, corrected total sum of squares.
4)Press, the predicted residual sum of squares for the model.