Therefore, it was seen that variable weights were well represented
in the first two canonical coordinates for all FMPs. Hereafter,
as have stated before only those linear combinations of an original
variable with a variate in both significant canonical pairs identified.
But as clarification, a coefficient or weight value indicates the relative
contribution of each standardized variable, in the presence of
others, to a specific canonical variate by both magnitude and direction
(positive or negative) when the biplot representation is being
examined