%X This thesis develops a new methodology for associating spatial zoning characteristics to point features. Doing so, it aims to provide an applicable approach that can be used to measure location characteristics more accurately. While doing so it should reduce bias caused by the Modifiable Areal Unit Problem. Whereas traditional methods use statistical areas directly to associate spatial phenomena to point features, our suggested approach advocates for disaggregation of spatial data and utilizing optimal zoning systems created around individual properties in order to obtain functional statistical areas. By means of a case study, in which prices of 3287 commercial real estate properties across western Europe were estimated using multiple socio-economic indicators measured through varying statistical areas, it was found that our suggested approach could improve model accuracy up to 18%. Furthermore, functional statistical areas enable researchers to determine the spatial scales and shapes at which location characteristics should be measured. This should provide more control over data and allow for better comparable results.
%A Philibert Weenink
%D 2022
%L theses_frw3806
%T Overcoming the Modifiable Areal Unit Problem (MAUP) of socio-economic variables in real estate modelling