In our study, we would like to suggest a statistical
procedure that is known to be able to handle the problems
of nonnormality. Known as the modified Ft statistics, this
procedure is categorized under robust statistics. Robust
statistics combine the virtues of both parametric and
nonparametric approaches. In nonparametric inference,
few assumptions are made regarding the distribution from
which the observations are drawn. In contrast, the
approach in robust inference is different wherein there is a
working assumption about the form of the distribution,
but we are not entirely convinced that the assumption is
true. Robustness theories can be viewed as stability
theories of statistical inference and signify insensitivity to
small deviations from the assumptions [5]. What is