AP® Human Geography · Unit 1 · Topic 1.6
Scales of Analysis and the Modifiable Areal Unit Problem
Scale of analysis is the level you study data at, from local to global, and patterns can look completely different at each one. The modifiable areal unit problem means that even at the same scale, redrawing the boundaries around identical data can change the results.
A map is a summary, and summaries depend on where you draw the line between groups. Scale of analysis means the level at which you're grouping data: individual, neighborhood, county, state, country, world. The same raw numbers can produce opposite-looking patterns at different scales because aggregation averages things together and can hide or erase the variation that was sitting inside each unit. A county map of a vote can show a sea of red while the state it belongs to went blue, because red counties can be large in land area but low in population, while a small number of densely packed blue counties hold enough votes to swing the statewide total.
The modifiable areal unit problem, often shortened to MAUP, is a related but distinct issue: it's not about zooming between scales, it's about how boundaries are drawn at the very same scale. Take one city's crime data and carve it into three different sets of neighborhood boundaries, shifting the lines a few blocks each time. Not one crime record changes, but which neighborhood looks most dangerous can change completely, because each new boundary sweeps different addresses into different groups. The data is modifiable in the sense that the units analyzed are a human choice, not a natural fact.
Both ideas connect directly to why projections and symbolization choices earlier in this unit matter: maps always involve decisions, and MAUP and scale effects show that even raw statistics, not just visual style, bend under those decisions. Going forward, this is why geographers report what scale a finding applies to and stay cautious about assuming a pattern zooms up or down cleanly.
Key terms
- Scale of analysis
- The level at which data is grouped and studied, such as local, regional, national, or global.
- Modifiable areal unit problem (MAUP)
- The way statistical results can change when the same data is grouped using different boundary lines, with nothing in the data itself changing.
- Aggregation
- Combining individual data points into larger groups, which can average away internal variation.
- Scale effect
- The phenomenon where a pattern or relationship found at one scale does not necessarily appear at a different scale.
Watch it explained
- How it worksHow does aggregating county data up to states erase a pattern you could see before?
- How it worksWalk me through the modifiable areal unit problem using one dataset carved into three different grids.
- Worked exampleFind how a 60-40 statewide vote can still produce a map where most counties show the other color.
- Going furtherWhat changes about a poverty map when you switch from the national scale to the neighborhood scale?
- Common mistakeWhy isn't a pattern that holds at the global scale guaranteed to hold in one town?
Check yourself
End of topic.