Dynamic measure of development

Abstract

The article presents a possible using of the author's method – normalisation with respect to the pattern – in the construction of synthetic measure. When a stimulant (destimulant) is normalized, for each object the share of its distances from the maximum (minimum) in the total distance from the maximum (minimum) of all objects is determined. Such transformation meets the requirements of normalisation - deprives variables their units and unifies their ranges. Normalisation with respect to the pattern has properties suggested in the literature - preserves skewness, kurtosis and the Pearson correlation coefficients. Moreover, although the current data are the sole data used to convert variables, normalized diagnostic variables are comparable across time. This feature gives an advantage of pattern normalisation over other methods in dynamic analysis of complex phenomena. The article uses normalisation with respect to the pattern in construction of Hellwig’s measure of development, in which Euclidean distances from an abstract ideal point are calculated. Since normalized diagnostic variables become destimulants with the minimum value equals 0, the ideal point used to construct a synthetic measure is constant over time. So, the values of modified measures are comparable both across objects and time. One can compare the positions of objects in the rankings as well as the values of the measures themselves (calculate the increments of values, descriptive characteristics, etc.).