Point pattern analysis: an application to the loyalty networks of chain-stores
At the 41th congress of ERSA., Staufer-Steinnocher (2001) proved that kernel density estimation, a technique of spatial analysis belonging to the point pattern methods, can be usefully applied to geomarketing. Following this point of view, the aim of this paper is to show that other point pattern techniques (center-grahic statistics, global and local autocorrelation indexes, clustering methods, 'nearest neighbour index', 'Ripley's K statistic', etc.) are able to suggest important considerations in marketing researches, making explicit the "geographical knowledge" embedded in available informations. Namely this argument is demonstrated, analysing spatial distributions of big stores chained in a promotional network, finalised to improve fidelity in the consumers.