Influence of Different Spatial Aggregations on Variables Implemented in Macroscopic Road-Safety Modeling

Abstract

Spatially aggregated data are prone to the effects of the modifiable areal unit problem, especially when the phenomenon of interest is not used as a criterion to define the areas of analysis. Further, these conditions can result in statistical inconsistency. The purpose of this study is to assess the influence of different scales of analysis on the statistical variables employed in macroscopic modeling of road safety for the city of Fortaleza, Brazil. Two sets of spatial configurations were created with different aggregation criteria as a means of detecting changes in the averages, standard deviations, correlations, and spatial autocorrelations of the variables. The results show that the statistical variation is not only dependent on the change of scale, but also on the format of the variable produced by the aggregation process.