ABSTRACT
Scale is a major component of multiscale geographically weighted regression, and this chapter discusses types of spatial scale and how MGWR can be used to investigate spatial process scale. We show how the bandwidth is a measure of the spatial scale over which different processes are relatively constant, and then we show how two major problems relating to spatial scale, the modifiable areal unit problem and Simpson’s paradox, can be better understood by alternatively viewing aggregations of data as aggregations of spatially heterogeneous processes.
