Spatial association of population pyramids across Europe: The application of symbolic data, cluster analysis and join-count tests

被引:15
作者
Bivand, Roger S. [1 ]
Wilk, Justyna [2 ]
Kossowski, Tomasz [3 ]
机构
[1] Norwegian Sch Econ, Helleveien 30, N-5045 Bergen, Norway
[2] Adam Mickiewicz Univ, Dept Settlement Syst & Terr Org, Inst Socioecon Geog & Spatial Management, Ul B Krygowskiego 10, PL-61680 Poznan, Poland
[3] Adam Mickiewicz Univ, Spatial Econometr Lab, Inst Socioecon Geog & Spatial Management, Ul B Krygowskiego 10, PL-61680 Poznan, Poland
关键词
Symbolic data analysis; Cluster analysis; Classification; Population pyramid; Spatial autocorrelation; Spatial cluster; DISSIMILARITY MEASURES;
D O I
10.1016/j.spasta.2017.03.003
中图分类号
P [天文学、地球科学];
学科分类号
07 ;
摘要
Demographic processes across European regions show great diversity, but because of this diversity, it is hard to gain an overview of similarities and differences. This paper aims to examine the application of a new combination of existing approaches to the analysis of regional population pyramids to offer such an overview. Symbolic data analysis and cluster analysis are used to identify typical shapes of population pyramids, before applying join-count tests to examine the spatial distribution of these pyramid shapes. The data used are for 1397 NUTS regional units in 37 European countries in 2015. We find that Irish regions, Cyprus and some of the capital cities of Western Europe present the youngest population across Europe, while the population of Eastern Germany is the oldest and shrinking in size. Countries of East-central Europe are the most homogeneous in their demographic processes for the chosen period, while the large demographic discrepancies occur within Spain, France, the UK, Finland, and Sweden between NUTS regions. A spatial study indicated positive spatial autocorrelation, and the transnational character of demographic processes across Europe. For a detailed examination of East-central Europe, we applied local statistics, and revealed three transnational spatial clusters resulting from historical and socio-economic processes. (C) 2017 Elsevier B.V. All rights reserved.
引用
收藏
页码:339 / 361
页数:23
相关论文
共 66 条
[1]  
Anderberg M.R., 1973, Probability and Mathematical Statistics, DOI DOI 10.1016/C2013-0-06161-0
[2]  
[Anonymous], SYMBOLICDA ANAL SYMB
[3]  
[Anonymous], 2011, RELATIONSHIP DEMOGRA
[4]  
[Anonymous], 2006, Symbolic Data Analysis: Conceptual Statistics and Data Mining
[5]  
[Anonymous], B I INT STAT
[6]  
[Anonymous], MATH CLASSIFICATION
[7]  
[Anonymous], 1963, PRINCIPLES NUMERICAL
[8]  
[Anonymous], 1984, Cluster Analysis
[9]  
[Anonymous], ANAL SYMBOLIC DATA E
[10]  
[Anonymous], EUR MAN GUID