1st International Conference on Spatial Statistics - Mapping Global Change, Enschede, Hollanda, 23 - 25 Mart 2011, cilt.7, ss.287-292
This study, using two methods of spatial data assessment: 1) spatial autoregression (SAR) models (Cliff and Ord, 1981) and 2) geographically weighted regression (GWR) (Fotheringham et al., 1998), identifies the relationship between shopping centers attributes and trade area (TA) characteristics in Ankara, Turkey. Ankara has the highest level of shopping center gross leasable area per capita in Turkey, for this reason, is a unique case study. The two models provide information on distinctive characteristics of shopping center locations. The first one depicts the global relationship between shopping center supply, assessed by total gross leasable area in each district, and demand, assessed by demographic and socioeconomic characteristics of TAs, such as population, income, homeownership while accounting for the spatial dependence across the TAs. The second one, on the other hand, assesses local demand supply relations at the district level. These two models do not substitute but complement each other. SAR model results show that there is a positive relationship between shopping center supply and median age, and a negative relationship between supply and household size. These are expected results in compliance with the literature findings. The level of homeownership variable, however, illustrates a distinctive picture, unique to Ankara, in that there is a negative relationship between homeownership and shopping center supply. The GWR results show that it is easier to explain the level of variations in selected parameters in the suburbs than in inner city neighbourhoods, therefore, as expected, in car dependent suburbs stronger relations with shopping center locations are identified than mixed-use inner city neighbourhoods. The results are essential for identifying the spatial network of retail outlets in a city or region which guides urban developers, investors and public policy decision makers in site selection. (C) 2010 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of Spatial Statistics 2011