The idea is to suggest that statistical techniques can be applied in the situation. Imagine a situation where food and/or wine is set out for a taste evaluation. Participants are given paperwork enabling them to record their conclusions. You will need to determine what results you want. Perhaps it is, in each case, like-dislike on a scale of one to ten. Putting results in preference order is more complicated.
One of many interesting elements to a taste activity is "to enable comparisons to be carried out". Participants will conduct their own comparisons. If it is worthwhile, these can be brought together to see if participants agree with one another. Other analyses can be carried out as well. Perhaps how the men fare compared with the women. That's not the same as agreeing with each other or concordance. Perhaps men are not so good at determining the fragrance of a dish. You will need to qualify fragrance.
In outline, decide what you want answers to, devise a way of obtaining the info, select a tool to analyse the info and then be sure that your results have statistical validity.
Organisers might discuss preliminaries such as:
Stats applied within the Food & Beverage situation can highlight, to many unfamiliar with what stats might help with, the extent to which men and women differ in their broad taste preferences can be tested. There was not much emphasis on this in the Gastronomy etext account.
If you want to test the assertion that more men than women favour "umami" [savouriness] , it would not be done [ a commmon thought within our groups of students in the seventies/eighties ] in tandem with wine assertions such as more men than women prefer red wine. [Any chance of contact with wine was to be promoted, no doubt. "Umami" as a term was created after the seventies.]
Meanwhile, a brief look at a light-hearted treatment of umami might suffice. Go to "G for All" and see the side panel.