New research tools with multidimensional statistic analysis methods

Data Analysis, established by the French School, encloses numerous different statistical methods. We can distinguish two main directions. a) Factor Analysis, which allows us to discover the structure and dependencies of a set, by examining a factorial level, with Correspondence Analysis being the predominant method, created by Professor J.P. Benzecri of the University Paris VI. The main advantage of the researcher using these methods as nonparametric ones, is that it is not required to take account of any hypothesis as to the parameters of the phenomenon studied. The great advantage of Data Analysis, is that it has another great ability, to take direct qualitative variables into account and use them in combination with quantitative ones. The factor analysis therefore represents graphically the relationship between two assemblies which shows on various levels in order to facilitate interpretation of these relations. This approach is clearly of quality content and is achieved by pure analytical methods of Linear Algebra b) Automatic Classification, which is to classify the statistical units of a population into homogenous groups, which are simultaneously affected by several factors.