ABSTRACT
In the literature on correspondence analysis (CA), there is often a difference between the notation used in the French and the English texts. Whereas English works often use matrix algebra, the Benzé?cri-school favors linear algebra. Each row or column point can also be thought of as a local barycenter, or a centroid, that is an expression for a weighted average profile for the given category; the point sums up the average position of all the respondents from the given country. Once the position of the barycenter is known, the associations in the table can be presented in a barycentric coordinate system. Once the chi-square distances between the category points and the barycenter are known, one can easily find the value of the inertia. Many researchers want to look at the relations between three or more variables, or between larger sets of variables, and when this is the case, to do a CA is not necessarily an appropriate solution.
