A Factorial Approach for Sorting Task data ( FAST)

Abstract

The aim of this paper is to present a new approach to analyze categorization data called FAST that stands for Factorial Approach for Sorting Task data. This approach, based on multiple correspondence analysis (MCA), provides an optimal representation of the products, an optimal representation of the consumers, which are to be interpreted jointly. It provides also elements of validation based on confidence ellipses. In the case of “labelled” categorization, where a verbalization task is asked to describe the groups of products, it provides an optimal representation of the words which is directly linked to both representations of the products and of the consumers. The FAST approach will be illustrated by an example where 98 consumers were asked to group 12 luxury perfumes.