POSSIBILITIES OF THE CLUSTER ANALYSIS METHOD FOR CLASSIFYING THE POPULATION OF A REGION BY SOCIAL POVERTY INDICATORS

Abstract

The purpose of the article is to characterize the possibilities of cluster analysis for studying poverty as a social phenomenon that accompanies a person at all stages of their existence and touches all spheres of public life. This method allows grouping respondents with similar social characteristics into associations, revealing relatively homogeneous social agglomerations. The use of an integrative multidimensional hierarchical approach for the analysis of poverty as a social phenomenon is substantiated. This methodological approach involves the use of a set of equivalent objective criteria, such as cash income, available savings, housing status and car ownership, social and professional criteria, such as educational level, prestige of a professional occupation, governing powers and the degree of intellectual work, a subjective criterion expressed on empirical level as a self-assessment of the situation in society. Based on these criteria, according to the results of a sociological study, the methodology for applying cluster analysis is substantiated, its testing is presented in the process of classifying the population by poverty indicators, and five clusters are presented: “Absolute poor”, “Working poor”, “Poor pensioners”, “Mass layer”, “Well-off”, which differ in the level of poverty; their distinctive characteristics are given that make it possible to compile a social portrait of poverty in the studied region.