Features of statistical methods application in rating construction

Abstract

The article is devoted to solving the problem of rating construction methodology, including issues of comparability in dynamics and inconsistency. The study analyzes approaches to the procedures for aggregating private components of the rating. Attention is focused on the shortcomings of ratings used in international and domestic practice as integral meters, among which are the incompleteness of information on the private components of ratings, an unacceptably high level of error in the development of ratings as integral meters and the lack of evaluation of their quality. The approach to rating construction on the basis of average weighted sum of ranks is offered. The ratio of the sum of the squares of the rank paired coefficients of the private rating to the total sum of the squares of the rank coefficients for all private ratings (the matrix of squares of rank correlations) is used as weights. The approach proposed by the authors, in contrast to the existing methodological approaches, makes it possible to build a rating in the absence of information about private components. The novelty of the study results lies in the method of estimating weights of private ratings as basic components of the rating. The article presents the results of construction a rating of the regions of the Central Federal District in Russia for achieving national goals in 2021 based on private ratings, which confirmed the adequacy of the proposed approach. The article may be of interest to a wide range of researchers in the field of statistics and regional economics.