Working Paper Series des Rates für Sozial- und Wirtschaftsdaten 99
Spatiality is an increasingly important dimension in the social sciences, as a new wave of recent publications shows. Intra-national comparisons have proved to be as fruitful as the better known inter-national analysis. Regional characteristics are found to have considerable influence on individual behaviour. This movement was fostered by methodological advances, e.g. in multi-level techniques. On the data side spatial analysis is supported by a good basic infrastructure in official and semi-official information, often provided in the way of easily usable DVDs. In addition, both scientific researchers and commercial marketing firms collect valuable information, partly on a very detailed local level of only a handful of households. However, there is ample room for improvement. Huge existing datasets (e.g. PISA E) are not open for spatial oriented scientific purposes; in many cases regional information is not sufficiently available (e.g. on criminal behaviour); systematic over-sampling in sparsely inhabited areas to allow additional regional analysis is not very common.