Tasting session dashboard

Work in progress

This site is still under construction. Please come back later for further information and documentation.

PCA on tasters

The first PCA performed provides information on the agreement between tasters and indications on how each single taster is able to differentiate between samples.

PCA performed on assessors as features. Each of the points are representative of one of the tasters, who overall show a relatively good agreements

What to look for

There are 4 cases to consider:

1. Tasters in good agreement will be close to each others. You therefore might be able to find patterns between group of tasters who seem to share the same opinion;
2. Tasters with strong disagreement will lie on opposite side of a straight line through the origin of the chart. It would reveal that their overall opinion on the samples has been quite diverging, liking and disliking being reversed;
3. Tasters who do not specially agree nor disagree will lie relatively far from each other, but not on opposite side of a straight line as above;
4. Finally, tasters who are not able to really differentiate between samples and whose opinion does not really impact the variation between sample will fall close to the center. It could be due to a poor use of the scales in the tasting protocols, or because the samples are indeed very similar.

PCA on tasting criteria

The second PCA aims at finding links between the evaluated sensory characteristics and how relevant they are in differentiating between samples.