Una procedura sfocata per la costruzione di un modello fattoriale

Abstract

In this study the author presents a new method to obtain a factor model. It is founded on the properties of the hierarchical fuzzy clusters of single-linkage for grouping variables. It has the following qualifications: a) it uses very simple algorithms for calculation; b) questions on indeterminateness aren't brought up; c) it allows know the interdependence between factors without having to solve problems of rotation; d) it is possible to know, the nesting of the variables through the graphical representation of convex adjacency matrix.