EXPLANATORY MULTIVARIATE METHODS: THE IMPORTANCE OF PERFORMANCE MATRIX: DATA ANALYSIS BY THE SPHINX SOFTWARE

Multivariate analyses are characterized by a higher degree of technicality and a stronger returning with renewed vigor to mathematics; moreover, Multivariate analyses meet three objectives: explanation, synthesis and classification. In our contribution we are interested in one of the named explanatory multivariate techniques; importance-performance matrix. The latter makes it possible to understand how a variable is possibly influenced by a series of other variables: which have no influence. The illustration of this technique is based on a raw survey file on the satisfaction of pensioners with the services provided by the national pension fund. The analysis of survey data is done by the Sphinx software.