where yij is jth observation on a property for ith treatment, y:: is the
grand average of all the observations, yi: is the average of the observations
under the ith treatment, SSTreatment is the sum of square
for treatment, SSError is the sum of square for error, a is number of
treatments and N is total number of observations. The method
provides a test of a null hypothesis that each sample is drawn from
the same underlying probability distribution against the alternative
hypothesis that underlying probability distributions are not the
same for all samples. When F is larger than Fcritical, the null hypothesis
of different group means being equal will be rejected. If
the null hypothesis is rejected, Scheffe post-hoc procedure was
employed to distinguish which pair(s) of group is/are significantly
different. The following was employed (Montgomery, 2013)