 In this week's lesson you'll be learning about the modifiable area unit problem and get hands-on experience as to how it works. Essentially it's been long recognized as a problem since the 1970s and occurs when the spatial pattern of a variable can be completely modified according to the level of aggregation of spatial units. Its use has been well documented during elections when political boundaries are deliberately manipulated to improve the electoral chances of a particular party. So to give a working example quickly, I've created points in ArcGIS. These are randomly distributed throughout the state of Pennsylvania and then aggregated them using different aggregation boundaries. First, small size grid cells were put across Pennsylvania and the data aggregated. These grid cell sizes were increased in size and then aggregation was made by county boundaries. As you can see, the spatial patterns change between each of these maps. This is so essentially the modifiable area unit problem occurs when different aggregations of counts produce different spatial patterns when boundaries are drawn in different ways.