In a series of records that have been published in the 1914-1916 period, Corrado Gini has laid the foundations of the discussion on a methodology for studying the statistical relations concerning the association of two characters. Gini's theory of concordance, which makes a distinction between the notion of absolute and relative maximum, allowed the definition of a multiplicity of correlation indexes and homophilia which are to be connected to the various cognitive applications that are to be found in the substantial research context. Gini's theory, at the same time, has performed a critical deepening which permitted to obtain a complete framework together with a reinterpretation of already known indexes such as Bravais-Pearson's correlation coefficient and Spearman's rank correlation. In our work, by extending Gini's concordance theory to multiple statistical distribu-tions and in the light of the simplest algebraic language, some partial correlation indexes, which have been previously defined by means of the geometric instrumenta-tion of tensorial calculus, are here reconsidered, and, therefore, the partial homophilia indexes are introduced.
Read full abstract