ANALYSIS OF PROXIMITY AND SIMILARITY AMONG THE PROVINCES OF TÜRKIYE BASED ON 24 VARIABLES USING HIERARCHICAL CLUSTERING

Abstract

In this study, a hierarchical cluster analysis was carried out based on 24 economic and social variables selected for the provinces of Türkiye. The main aim of the study is to determine the similarity and closeness levels of 81 provinces according to these 24 variables, to group provinces with similar characteristics under the same clusters, and to analyze the spatial socio-economic differentiation of Türkiye. The variables used in the study were formed from economic, social, and demographic data and were made comparable through a preliminary preparation process. The 81 provinces were analyzed by using the statistical method called hierarchical clustering, and the relative positions of the provinces were determined. In this context, a proximity matrix, a dendrogram, and similarity coefficients in the agglomerative schedule were produced. According to the results of the hierarchical cluster analysis, cluster groups were identified at distances of 1, 5, 11, and 17 units. At a distance of 1 unit, 16 groups were formed; at 5 units, 5 groups; at 11 units, 3 groups; and at 17 units, 2 groups were identified. The findings show that the provinces of Türkiye form clear socio-economic clusters, and this clustering reveals regional development differences across the country. It was also observed that some provinces with similar socio-economic development levels are included in the same clusters even if they are not geographically close.