| Popis: |
The relevance of the problem is due to the fact that despite the many works devoted to the periodization of demographic processes in the Russian Federation and in a number of foreign countries, there are questions about the use of formal methods of phase separation and evaluation of the quality of periodization. The purpose of this work is to improve the methodology of periodization of demographic processes in the Russian Federation, represented by time series, using statistical tools. The article presents the results of five variants of periodization of the total fertility rate in the Russian Federation for 1960–2017. At the same time, various methods of phase separation, both in one-dimensional time series and taking into account informative factors, are used in the work, and approaches to the comparative evaluation of the results of periodization are proposed. To select homogeneous periods, the author uses onedimensional and multidimensional data arrays processed by phase and cluster analysis. As a result, several not previously established homogeneous subperiods – phases of development, significantly different from each other, were identified. Deviations from the regression model, parabolic trend and Holt model were used as fluctuations. Cluster analysis was carried out on two data sets differing in the composition of exogenous variables. In the dynamics of the analyzed time series, from three to five phases, including from 6 to 24 years, are distinguished. To select the optimal variant of periodization, meeting the scientific requirements, the author proposed an approach of analysis of the homogeneity within the selected phases, based on the coefficient of variation and the significance of differences between the output phases on the basis of F-Fisher criterion – Snedcor. A comprehensive assessment of the selected phases allowed us to establish that the best characteristics are the periodization of the total fertility rate in the Russian Federation for 1960–2017 years performed using phase analysis for deviations from the regression equation. Despite the methodological nature, the work has a practical aspect, as the analysis relates to real statistical data on demographic processes in Russia. The materials of the article can be used by researchers of the problems of periodization of socio-economic and demographic processes; regional and Federal authorities to assess the effectiveness of management decisions. |
