Questionnaire: Cooking and you

Authors: Cédric GLEMIN and Cécile KOUBA

Our aim is to find a link between student's lifestyle and student's mode of food consumption. For that, we built a questionnaire for students to collect information about what their favourite meats, starchy food or hobbies were.

As first analysis, we made a Multiple Correspondence Analysis. We chose to put in the foreground (active variables) the different modes of student's food consumption. In the background (illustrative variables), we decided to analyse their personality, lifestyle…
Our analysis led us to interpret 2 axes:
The first axis explains 6,6% of total inertia and the second axis explains 5,3% of total inertia. We did not take axes 3 and 4 into account due to their low percentage of inertia.
Moreover, we found that the first axis represents the attraction students could have concerning cooking and the second axis characterizes students according to their different modes of food consumption.

As our questionnaire was directed only towards students, our analysis led us to clusters of students which are quite similar concerning their lifestyles and their modes of food consumption. However, beside their similarities, we can distinguish 3 clusters of students. We obtained those 4 clusters with a Hierarchical Clustering on Principal Components:

  • The first cluster is composed of students whose alimentation is varied and who eat fruits and vegetables often a day. Most of these students cook for the pleasure, they have cookbooks and kitchenware
  • The cluster 2 is composed students who do not like cooking, do not have any cookbook and whose aim of cooking is to feed themselves. They eat especially meat and fish and do not eat season fruits and vegetables.
  • The third cluster is composed of students who have a quite varied alimentation but who do not enjoy cooking very much

Thanks to this survey, we saw that there is a real typology of the habits. Each different group of student has a particular way of cooking and feeding themselves.

Find here the data set and the R script:

The results of the ENMCA() function are the following:

Beware the results of HCPC() and ENMCA() are not quite the same since the data are ventilated and the method of classification is a bit different.