We use multiple correspondence analyses on tables of three or more categorical variables. This expands the study of a simple correspondence analysis from the two-way table to more variables. One downside of this technique is the loss of how rows and columns relate.

The data that we will look at is based on a study of students and their gender—whether they live in urban, suburban, or rural locations—and their goals to be popular, that is, being good at sports or getting good grades.

This is a simplified dataset that is taken from the original located at DASL. The full dataset can be found at the following location:

http://lib.stat.cmu.edu/DASL/Datafiles/PopularKids.html

We will tally the results of the columns to observe the value order of the columns and then we need to create a categorical list of the labels for each item in the factor columns. We will list these figures in the fourth column in the correct value order...