Prostorový dosah volebních preferencí

Abstract

This bachelor thesis analyzes spatial autocorrelation of electoral results in the Czech Republic using geoinformatics methods. Based on detailed polling district data, differences between presidential and municipal elections are evaluated. Spatial relationships are quantified using the global Moran’s I index with various spatial weight definitions (fixed distance, KNN, topological adjacency). The study also addresses the impact of spatial aggregation (MAUP). The results confirm that spatial autocorrelation strongly depends on scale and neighborhood definition, and that different election types exhibit distinct spatial patterns.