Gemeindetypisierung mit Hilfe quantitativer statistischer Verfahren (Beispiel: Regierungsbezirk Köln)

Abstract

This work examines the use of factor analysis and distance grouping as an aid to as objective as possible a classification of Gemeinden (local authorities). The 237 Gemeinden of the Cologne Regierungsbezirk (Government District) were used as an example. Out of 40 demographic, economic and constructional variables drawn from official statistics, 7 variables of higher order (the factors) were computed, which underlay or caused the interdependencies between the selected characteristics. For specific problem formulation it would be necessary to substitute for the groups of variables presented here a combination more oriented towards the problem in question. The factor loadings (correlations between variables and factors) enabled interpretation to be made, after the varimax rotation, of the 7 factors, which served as the new descriptive dimensions of the Gemeinden. Factor values were computed for the Gemeinden which gave a measure of whether a particular factor is characteristic for one Gemeinde or not. Following that, it was possible to comprehensively describe the structure of every Gemeinde in terms of 7 factor values. In a last section, a distance grouping in multidimensional space was under taken in order to attempt the classification of similar Gemeinden into types. This indicated that it was unnecessary to take all 7 factors into account. Because of this, two groupings were undertaken (once for economic structure and once for age structure). The map at the end shows a combination of both the types thus derived.