• From the latent variables:
• From the scores of Table 1:
• From the scores of Table 2:
• Narrative conclusion:
○ Umami is a complex flavor anf can be expressed under different variations like alliaceous, or other savoriness.
○ Bitter is not a wanted taste, thus it is most likely to be negatively correlated with the other attributes.
○ Acidic is also present in all products and correlates with most other attributes, especially floury (starchy) and spicy (pungency).
○ Fatty is has the faintest effect among all chosen basic tastes. Perhaps, the research that fatty as a basic tastes in human perception needs to be studied further before we can fully accept Fatty as a basic taste.
I think PLSC did a great job at showing the relationships of the different attributes based on the basic tastes. The perception of Flavors (taste and smell) has a lot of intrinsic subjectivity and I think projecting all attributes back the well-researched, well-established “basic tastes” can provide new valuable new insights into how to best analyze our data. PLSC is the back-bone of many modern analytical approaches thanks to its ability to relate multiple data tables that matched on similar observations.