3.2.1. Model fitting and analysis of variance (ANOVA)
To optimize the effect of extraction parameters on oil yield, RSM
with CCD was used. The experimental range and model bound of
638 A.S. Reshad et al. / Fuel 150 (2015) 636–644
the independent parameters studied are discussed in Section 3.1.3.
A 2n factorial CCD model suggests conducting total 13 experiments,
considering four axial, four factorial and five centre points. The
model predicted parameters and experimental generated data are
shown in Table 2. The oil yield data obtained from the experiments
were fitted to develop the model. The significant of each parameter
were evaluated from the p-value by ANOVA which reflects the
strength of significances of the parameters. RSM suggested
quadratic model with F-value and p-value of 40.03 and 0.0001
respectively. Accordingly, the ANOVA of the suggested model
was analysed and presented in Table 3.The p-value was observed
to be lesser than 0.0001; which signifies the reliability of the model. Similarly, the smaller the p-value for the independent parameter, the more significant it is in the suggested model for
response [23,25,35]. The suggested model’s p-value (