M.A.D

John Tukey was the one who first separated Multivariate Statistical Analysis from Classical Statistics, naming it Data Analysis and presenting it as an independent science. According to him Data Analysis includes:

1. Ways to design and collect data to make analysis easier, more correcte, and more accurate.

2. Analysis procedures.

3. Interpretation techniques and results

Question 2: For which data is it necessary to use data analysis methods ?

For qualitative data

Question 3 : What does MAD software include ?

Includes Classical statistical analysis methods, such as hypothesis testing, analysis of variance, simple and multiple regression, time series, non-parametric tests, etc., as well as ten methods of multivariate statistical data analysis, such as Factorial Analysis of Correspondences, Principal Components , Ascending Hierarchical Classification, Class Analysis, Discriminant Factor Analysis, Inductive Statistics and Hierarchical Analysis, Generalized Hierarchical Classification, the BENKAR method and Composite Hierarchical Classification. The menu includes most of the data transformation procedures before the various analysis methods are applied, such as creating a logical 0-1 table, or a BURT table, and multidimensional table processing procedures, such as converting a three-dimensional data table into a table suitable for statistical analysis.

Question 4 : What does the KARAP method offer ?

The KARAP method, on the one hand, helps to interpret the factorial level 1x2, on the other hand, with the new term "closest relative" established by Professor Dr. Dimitrios Karapistoli, it ensures the researcher solid classes, in terms of the uniqueness of the behavior of the "objects" of each cluster of variables, since it only includes objects whose profile is linked to specific variables of each cluster.Also by finding the "nearest relatives" of each "object" the image presented by the 1x2 factorial level is "cleaned up", after the main connections between the "objects" and the variables have been clarified, since the KARAP method offers 100% of the information drawn from the data matrix, preventing untimely connections of "objects" and variables, due to reduced inertia presented by the two factorial axes.

Question 5 : What are "orphaned" objects ?

Orphan objects are those points that in the three-dimensional subspace are not associated with any variable. It is therefore an element to be investigated.

Question 6 : What does the term "nearest kinsman" of an "object" introduce?

This term considers for an "object" Ii as the closest relative the variable Xk when the profile of Ii has with the profile of the variable Xk the shortest distance from any other variable Xs of the sample using the set of factor axes, thus offering for the set of the connections between "objects" and variables information rate 100%.

Question 7: What is the necessity of using the Generalized Ascending Hierarchical Classification ?

Based on the method of Factorial Analysis of Correspondences and the KARAP method, it answers the question of identifying the typology of variables in combination with the "objects" with which each variable is most associated .

Question 8 : What do KARAPISTOLI charts draw ?

Karapistoli diagrams graphically draw the three-dimensional subspace in the plane, so that the researcher does not need to perform classification in order not to confuse neighboring points that belong to different subspaces of the Rⁿ

Question 9: What exactly is achieved with the BENKAR method ?

With the BENKAR method we evaluate the classification of N objects that came from the application of the Ascending Hierarchical Classification with the FACOR method or the VACOR method.

Question 10 : What is achieved with SVM machine learning ?

With the SVM learning machine, every form of classification is objectively evaluated by whatever metric it was created, helping the researcher to compare the results between two different classification methods, but also to establish the homogeneity of the classes of each classification.

Question 11 : Why do we need to evaluate the classes of an Ascending Hierarchy using the KARAP method ?

The proposed procedure answers the question: "based on the different importance of the variables observed in each class of the Ascending Hierarchy, which class presents, for the "subjects" participating in it, the strongest presence in the formed typology"

Question 12: What does the new Complex Hierarchical Classification method offer ?

The Complex Hierarchical Classification method proposes a new method of classification, whose philosophy focuses on the obvious, that is, that any evaluation must, like the criteria used, have a single importance directly from the beginning, which results from the system of values ​​that shape the elements endogenously of the evaluation and not from extraneous factors that may arise from arbitrary calculations or a priori assumptions.

Question 13: What does Graph Classification offer ?

The classification based on the theory of graphs is for the researcher to identify the processes that realize in the connections mainly between the variables.