2.8 Design of Experiment
For response surface design, the following table was the procedure for
the experimental setup for central composite design. The design of
experiment was adopted to study the responses which are percentage free
fatty acid content and total vitamin E content of the hydrolysed sample.
The independent variables that determined the experimental system were
weight of water (\(H_{2}O)\), weight of lipase, and reaction time.
Variables were coded at five levels: -\(\alpha\), -1, 0, 1, and\(\alpha\). The range of values that describe the magnitude of both
actual and coded are shown in Table 2.1. In the central composite
design, five start-points and six replicate runs were incorporated to
provide for estimation of model curvature and allow for the estimation
of pure error respectively Chu et al. (2003). All experiments within
this setup were done in a random order to minimise the effect of
unexplained variability in observed responses due to extrinsic factors.
Statistical analyses for the optimization procedure was done using the
Design Expert (Version 11) software. A quadratic polynomial was assumed
for predicting the responses of the regression models.
Table 2.1 Experimental Protocol for Central Composite Design for
enzymatic hydrolysis of palm fatty acid distillate