APPLICATION OF CHOSEN NORMALIZATION METHODS IN THE PROCESS OF CONSTRUCTION OF SYNTHETIC MASURURE IN INDIRECT CONSUMPTION RESEARCH

Abstract

The main purpose of this paper is to find the most effective method for the normalization of a final set of diagnostic variables for indirect consumption research with the use of synthetic consumption measures. The chosen normalization methods (standardization, unitarization and ratio transformation) were analyzed, both in the classical and the order approaches, with the use of different methods for multidimensional median vector construction (border and Weber median). Implementation of the multidimensional median construction is very important in the case of research objects with atypical characteristics of diagnostic variables (a separate analysis of Warsaw from the Mazowieckie Voivodeship). This kind of research units can cause asymmetry in empirical distribution of diagnostic variables which has a significant impact on the result of the linear order set of research objects. Additionally, implementation of the Weber median allows for considering the interactions in a set of diagnostic variables, which is crucial from the point of view of economic analysis.