So there's this thing called the Modifiable Areal Unit Problem -- when point-level data is aggregated into geographic districts, the resulting numbers can tell a story that isn't real. These stories change depending on the modifiable areal unit.
Imagine we have this map showing habitat for some species of beetle. 
In this scenario, the state on the right side contains most of the habitat, with a little spilling over into the left.
Here's more beetle habitat info. Each color represents a different species. What happens if I aggregate and show each state's top beetle habitat by land area?
What do we lose when we aggregate? And what changes if we aggregate across different boundaries?
Imagine the red lines here are contour lines delineating elevation changes. What story does the map tell now? And how much of that story is lost in the aggregated map above?
Obviously we need to aggregate spatial data to show patterns and tell stories. Knowing about the Modifiable Areal Unit Problem can help you tell stories that are actually meaningful and prevent drawing conclusions that are not actually there.
