The Application Of Local Indicators For Categorical Data (LICD) In The Spatial Analysis Of Economic Development

Abstract

<jats:p>The paper makes an attempt to apply local indicators for categorical data (LICD) in the spatial analysis of economic development. The first part discusses the tests which examine spatial autocorrelation for categorical data. The second part presents a two-stage empirical study covering 66 Polish NUTS 3 regions. Firstly, we identify classes of regions presenting different economic development levels using taxonomic methods of multivariate data analysis. Secondly, we apply a join-count test to examine spatial dependencies between regions. It examines the tendency to form the spatial clusters. The global test indicates general spatial interactions between regions, while local tests give detailed results separately for each region. The global test detects spatial clustering of economically poor regions but is statistically insignificant as regards well-developed regions. Thus, the local tests are also applied. They indicate the occurrence of five spatial clusters and three outliers in Poland. There are three clusters of wealth. Their development is based on a diffusion impact of regional economic centres. The areas of eastern and north western Poland include clusters of poverty. The first one is impeded by the presense of three indiviual growth centres, while the second one is out of range of diffusion influence of bigger agglomerations.</jats:p>

Same-same, but different

“…The usage of static areas in investigations of geographical distributions is common in studies based on typologies of areas. In the domain of the EU, the use of Nomenclature of Territorial Units for Statistics (NUTS) or local administrative units (LAUs) is the standardized way to understand social geography (European Commission, 2007;Pietrzak et al, 2014). The reason the EU decided to implement standardized units of space was to make cross-country analysis more homogenous by securing similar population sizes within each area unit.…”

The application of Local Indicators for Categorical Data (LICD) to explore spatial dependence in archaeological spaces

“…Therefore, for irregular lattice, higher order windows are required, and cell classification becomes more complex. Pietrzak et al (2014) developed a simple method to estimate local spatial dependency for cells with variable size (Polish NUTS regions), by applying JCS to contiguity matrices (or k-nearest neighbour matrix) detemined for each region. The advantage of this method over Boots' method is that it does not rely on a regular moving window and it is less computationally demanding (for large areas).…”

Wpływ wyboru metody klasyfikacji na identyfikację zależności przestrzennych – zastosowanie testu join-count

“…2010Pietrzak i in. 2014a;2014b]. Rozkład przestrzenny wartości zmiennej niemetrycznej może być losowy bądź wykazywać tendencję do przestrzennego grupowania się.…”

“…metodę trzech średnich, metodę trzech median oraz metodę opartą na średniej arytmetycznej i odchyleniu standardowym zob. [Nowak 1999]). Wyniki zaprezentowano na rys.…”