3.3. Multiple Factor Analysis
The MFA allows a quick analysis of the behavior of the samples over
the storage time using a two dimensional map (Fig. 1). Each colored
point represents a distinct moment of evaluation (Day zero, Day 60
and Day 120) and the black one is a mean point. The distance between
each point enables to evaluate similarities and differences among each
sample and each evaluation time, i.e. if a sample is distant from another,
this represents that these two samples are different and if the scenery is
the opposite and the samples are very close, this represents that the
samples are similar.
The dimension 1 of MFA explains 71.3% of the variation among
samples and separated them in two groups: a group with control
sample (Sucrose) and samples that used sucralose to sweeten the
mango nectar and other group that used Stevia, Neotame and Blend.
The dimension 2 explains 12.8% and separated the samples Thaum/
Sucr and Stevia of their first group. The less distance between Sucrose
and Sucralose represents more degree of similarities in sensory profile
among these samples while the greater distance between Sucrose and
Stevia represents that these samples are very different.
The colored points, asmentioned, represent each evaluation time.
The red point represents Day zero, i.e. the day of sample elaboration.
In all samples that were noted the red point (Day zero) is far from
blue (Day 60) and green points (Day 120) and these two points are
close together. The changes in sensory profile of the mango nectar
samples were higher among Day zero and Day 60 than Day 60 and
Day 120. The greater distance between Day zero and Day 60 represents
that the sensory characteristics of all mango nectar samples suffered
changes in storage. In addition, these changes in sensory perception
were lower after the 60th day of storage and that was showed in
MFA by the proximity of the points that represent Day 60 and Day
120.
The Rv coefficientmeasures the agreement between each evaluation
time, i.e. a greater value of RV means that the configuration of this
evaluation time is similar than one another. The RV coefficient of the
sample configurations in Day 60 and Day 120 was 0.90 and was the
higher agreement between configurations. While the RV coefficient of
Day zero–Day 60 and Day zero–Day 120was 0.55 and 0.48, respectively.
This confirms what were observed in MFA, that the changes in sensory
profile were higher in the first half of the research (Day 0 to 60th) than
in the second half (Day 60th to 120th).