Scales of Analysis

By the end of this lesson, students will be able to (per CED LOs PSO-1.C and PSO-1.D):

• Define the four scales of analysis used by geographers — global, regional, national, and local — and identify the scale at which a given map, dataset, or image operates (per EK PSO-1.C.1 and Skill 5.A).
• Distinguish scale of analysis from map scale (the cartographic ratio).
• Explain how patterns and processes at different scales reveal variations in, and different interpretations of, the same data (per EK PSO-1.D.1).
• Identify the Modifiable Areal Unit Problem (MAUP) when it appears in a map or table and describe its scale and zoning components.
• Recognize an ecological fallacy — the error of attributing a group-level statistic to every individual in the group — and connect it to scale of analysis.
• Apply multi-scale thinking to a real geographic problem (e.g., Brazilian deforestation, US obesity rates, Chinese inequality, Kenyan migration) and identify what each scale reveals or hides.

Per CED EK PSO-1.C.1, "scales of analysis include global, regional, national, and local." Per CED EK PSO-1.D.1, "patterns and processes at different scales reveal variations in, and different interpretations of, data." These two essential knowledge statements together do something powerful: they tell you both which scales to use and why using more than one matters.

Scale vs. scale of analysis

Geographers use the word "scale" two different ways, and the AP exam will test the difference. Map scale (covered in lesson 1.1) is the cartographic ratio between distance on a map and distance on the ground — for example, 1:24,000 means one unit on the map equals 24,000 units on the ground. Scale of analysis is the geographic level at which a question is studied: global, regional, national, or local. The two ideas are related (a small-scale map shows a large area, which lends itself to global or regional analysis) but the AP CED treats scale of analysis as the topic 1.6 concept.

Why scale matters

The same data set will tell different stories at different scales. The United States has a fairly high per-capita income at the national scale — but at the local scale, neighborhoods of severe poverty sit beside neighborhoods of extreme wealth. China is "growing rapidly" at the global scale — but at the regional scale, coastal provinces grow far faster than interior ones. Aggregation hides internal variation; disaggregation reveals it. This is what EK PSO-1.D.1 means by "different interpretations of data."

Two errors caused by single-scale thinking

The Modifiable Areal Unit Problem (MAUP) describes how statistical results change when you change either the size of the spatial units (the scale effect) or their boundaries (the zoning effect). The ecological fallacy describes the error of assuming what is true of a group must be true of every individual in it — reading a group-level statistic as a personal one. Both errors get worse the further you operate from the scale at which the underlying process actually unfolds. Multi-scale analysis is the standard correction.

Per CED EK PSO-1.C.1. The CED names exactly these four; the AP exam expects you to identify them in maps, tables, and images per Skill 5.A.

The MAUP, formally identified by Stan Openshaw in 1984, is the most important methodological warning that comes out of scale-of-analysis thinking. It states that statistical results computed for the same underlying data can change dramatically depending on how the data are aggregated into spatial units.

The MAUP is one reason geographers insist on examining patterns at multiple scales, and it is the basis for the political fight over gerrymandering — the deliberate redrawing of electoral district boundaries to favor one party. It is also why a "best place to live" ranking based on counties might give a wildly different answer than the same ranking based on metro areas.

Ecological fallacy

The ecological fallacy is the related but distinct error of assuming what is true of a group must be true of every individual in the group. A country with a high average income contains many low-income people. A region that voted overwhelmingly for one candidate still contains substantial minorities for the other. A school district with strong test scores still has struggling students. The fallacy is committed when a group statistic is read as a personal one. Working at too coarse a scale of analysis is one of the most common causes.

Four cases that show how the same phenomenon looks fundamentally different at different scales of analysis.

Brazil · Amazon Deforestation

~17% of original Amazon cleared

The same forest-loss problem looks fundamentally different at each scale.

• Global: One of the world's largest carbon sinks at risk — a planetary climate problem.
• National: A Brazilian sovereignty and policy debate over how to develop the Amazon legally.
• Regional: Concentrated in the "arc of deforestation" along Pará, Mato Grosso, and Rondônia.
• Local: Specific cattle ranches, soy farms, and illegal mining sites, often on indigenous land.

Country page →

United States · Adult Obesity Rates

~42% national adult obesity rate (CDC)

Obesity is a textbook MAUP example: the rate looks very different by unit size and by region.

• Global: The US is among the highest-rate developed countries.
• National: A single ~42% adult figure that hides enormous variation.
• Regional: Higher across the South and Midwest, lower in the West and Northeast.
• Local: Census-tract differences driven by income, food access, walkability, and ethnicity.

Country page →

China · Economic Inequality

Coastal per-capita GDP often 3-4x interior

National GDP growth obscures regional and local divergence.

• Global: The world's second-largest economy and growing.
• National: Strong national growth and fast urbanization.
• Regional: Coastal provinces (Shanghai, Guangdong, Jiangsu) far richer than interior provinces (Gansu, Guizhou, Yunnan).
• Local: Within coastal cities, migrant-worker neighborhoods have very different conditions from registered-resident ones.

Country page →

Kenya · Internal Migration

Nairobi's population grew ~4% per year recently

Migration looks like one process at the national scale and many different processes at the local scale.

• Global: Part of broader Sub-Saharan urbanization.
• National: Strong rural-to-urban flow toward Nairobi and Mombasa.
• Regional: Different sending regions for different cities (western Kenya vs. coastal).
• Local: Specific informal settlements (Kibera, Mathare) absorb most rural arrivals; their demographics differ sharply from formal Nairobi neighborhoods.

Country page →

1. The CED names exactly four scales of analysis: global, regional, national, local. Why these four and not others? Where would "city" or "household" or "continental" fit, and are they implicit in the four CED names?
2. Use one current global event in the news. Describe how it looks different at each of the four scales. Which scale most accurately represents what is happening?
3. Gerrymandering is the political application of the zoning effect side of the MAUP. If a state's voters are evenly divided 50-50 between two parties, can district lines be drawn that produce 8 of 10 seats for one party? Sketch how, and discuss whether the resulting map represents the will of the voters.
4. An ecological fallacy says you cannot infer individual behavior from group statistics. Identify a recent news headline (about a country, region, or demographic group) that commits an ecological fallacy. Rewrite it more carefully.
5. Suppose you are designing a public-health intervention for adult obesity in the United States. Should the intervention be national, regional, or local? Why? What does the case study suggest about which scale will produce real change?

Draft alignment — pending educator review. AP HG codes correspond to the official College Board Course and Exam Description (Effective Fall 2020, V.1). Statements below are paraphrased in CountryReports' own voice; refer to the College Board's published CED for verbatim wording.

AP® and Advanced Placement® are registered trademarks of the College Board. The College Board was not involved in the production of this material and does not endorse it. Standards statements above are paraphrased; codes refer back to the official College Board CED, the NCSS C3 Framework, the Common Core State Standards, and other cited frameworks.