THE COMPREHENSIVE DEVELOPMENT OF THE COUNTRY FROM A MULTIVARIATE ANALYSISSummary: The work is structured in two parts: the first attempt to define the concept of "comprehensive development of countries" in order to select appropriate indicators to introduce the model to be established in the thumb. This is done through the decomposition of the concept in six (6) dimensions of four (4) specific fields each, which materialize this concept. These dimensions and fields are deducted from the consideration of the human being, in its evolution individually and collectively; of society interrelationship hombre-sociedad; of the economy, basic means development of culture, basic vehicle development and launch one aspect of quality, reflecting the kind of development generally recognized by the international community. In the second part of thumb, using multivariate techniques most suitable, we analyze the data for 72 countries selected from the whole set of different levels of human development on the basis of united nations (HDI). In this selection are introduced to represent countries at all levels of development. It establishes criteria interpretability of the development process of countries. An estimated development indices comprehensive original data and standardized factors conceptual models and significant. The multivariate techniques applied prove useful for this purpose. These techniques are: factor analysis, cluster analysis, discriminant analysis and structural equation model. The factorial analysis enables the reduction of 96 selected indicators to 22 factors conceptual sustain the bulk of the information contained in the initial model. It also provides a reduction of 22 factors in 4 factors dimensional conceptual level implementation. Analysis of conglomerates, through the application of the method of successive hierarchical and non-hierarchical (resignation) provides a categorical dependent variable (4 / 5 level) indicative of the degree of development, expanding the classification of nations united to 4 or 5 levels: high , middle, middle, middle and low. The discriminant analysis will lida the results of the analysis of clusters and allows rank countries with a higher percentage of correct answers to 90%. They apply three types of derivation of the discriminant function (simultaneous, lambda-wilks and d2 of mahalanobis). The structural equation model confirms and improves the exploratory factor analysis of 22 factors. By analyzing future of the initial model, the standard (transformation of variables) and reespecificación latter, it gets a significant equivalent model, which reduces the number of variables involved a 66% (64 indicators of initials), which is considered satisfactory. The analysis suggests the importance of all relevant dimensions considered and the relative influence of some larger fields over others. These results are likely to improve through a process of successive approximation, removing and incorporating those indicators that are most appropriate to present concepts to be measured and reespecificando models structural equations pair wing obtaining an optimal model.
