The strongest correlation to response variable was found on the
reaction time (X3). Tersely, the dependent variable (Y) was also positively
correlated with the molar ratio of methanol to palmitic acid in
feed (X1), reaction temperature (X2), and reaction time (X3). The
remaining independent variables were slightly negatively related
to the response variable. Furthermore, all independent variables
poorly intercorrelated to each other. The main effects at a confidence
interval of 95% were plotted as Fig. 3. Explicitly, the first three factors,
i.e. m, T and t, affected significantly on the conversion efficiency (Y)
of biodiesel from palmitic acid and methanol. In contrast, minimal
effects on biodiesel conversion were imposed by catalyst loading
(w) and stirring speed (s). Therefore, the higher levels of X1, X2 and
X3 would be expected to result in a higher conversion efficiency of
palmitic acid to methyl palmitate. Above all, the effect related to
the reaction time (t) had the largest slope, which implied that the reaction
time possessed the strongest effect on biodiesel production
efficiency (Y).
To explicitly estimate the significance of each factor, the ANOVA
analysis on the 2-level FFD was carried out. After performing
ANOVA analysis for the full model, the fact that three main factors
and three two-way interactions were found to have p-values less
than 5% indicated that they were significantly different from null at