Abstract: processing of research results using methods of mathematical statistics is carried out in many fields of science, including sociology. One of the distinguishing features of sociological data is the predominance of categorical or non-numeric variables in the array of primary information. In this regard, an important task is the task of introducing into a sociologist’s practice methods that allow one to analyze variables of this type, to investigate both two-dimensional and multidimensional relationships between them. The set of such methods is limited. These include, first of all, a multiple correspondence analysis, which allows not only numerically, but visually examine large frequency tables, build correspondence maps, and identify multidimensional relationships between categorical variables. The purpose of this article is to demonstrate the capabilities of the multiple correspondence analysis method in interpreting the results of sociological studies. The method of multiple correspondence analysis was applied in processing the results of a survey on the study of political activity of Russians. In particular, the problem was solved of assessing the degree of respondents’ interest in politics depending on their socio-demographic characteristics, such as gender, age, income level, and others. As a result, the Bert matrix was calculated, containing 36 submatrices, which are contingency tables of individual variables. Based on the data of the Bert matrix, a two-dimensional correspondence map was constructed, which allowed visualizing the arrangement of the categories of each variable, estimating the distances between them, and also identifying and describing four clusters of respondents depending on their attitude to politics.
Keywords: multiple correspondence analysis, categorical variables, processing of results of sociological studies