01 / 21

Scales of
Analysis

Learn how geographers examine patterns at local, national, regional, and global scales to uncover different spatial relationships and draw richer conclusions.

AP Human Geography

What Is a Scale of Analysis?

A scale of analysis is the spatial extent at which a geographer studies a phenomenon. It determines the level of detail you see and the patterns that emerge. The same data can tell very different stories depending on whether you look at a single neighborhood or the entire planet.

The Four Primary Scales

Local
Neighborhoods, cities, and counties: the most detailed view of a place.
National
An individual country: policies, demographics, and cultural trends within borders.
Regional
Multi-country areas: Sub-Saharan Africa, Southeast Asia, the European Union.
Global
The entire world: worldwide patterns in trade, migration, and development.

Why Scale Matters

Changing the scale of analysis can change the conclusions you draw. A country may appear wealthy on a national scale, but zooming into specific regions or neighborhoods may reveal deep pockets of poverty that the national average hides.

GLOBAL REGIONAL NATIONAL LOCAL Most detail Broadest patterns

Core Principle: Scale Determines Pattern

This is sometimes called the ecological fallacy when broad-scale patterns are incorrectly applied to individuals, or the modifiable areal unit problem (MAUP) when results change based on how geographic boundaries are drawn.

Visualizing Scale Shifts

Local High detail Small area National Moderate detail One country Regional Less detail Multi-country Global ZOOM OUT → detail decreases, area increases ← ZOOM IN detail increases, area decreases

Scale of Analysis vs. Map Scale

Scale of Analysis Map Scale
Definition The spatial extent at which a phenomenon is studied The ratio between distance on a map and distance on Earth
Example "Studying urbanization at the national scale" "A map at \(1:50{,}000\)"
Purpose Determines what patterns emerge Determines how much area is shown
Conceptual? Yes: a framework for analysis No: a mathematical ratio

How to Apply Scales of Analysis

1
Identify the question. What spatial phenomenon are you investigating? (e.g., income inequality, migration, language diffusion)
2
Choose the appropriate scale. Does the question require local detail, or a broader regional/global overview?
3
Collect and analyze data at that scale. Use census tracts for local, country-level statistics for national, etc.
4
Consider other scales. Would the pattern look different at a smaller or larger scale? Always note limitations.
5
Draw conclusions carefully. Avoid applying broad-scale findings to individuals (ecological fallacy).

Example: Urban Poverty at Different Scales

Global "Poverty is declining worldwide" National "Country X has average income of $40K" Local "Neighborhood Y: 35% below poverty line" Same place, three different stories

Example: Language Patterns

Global Scale
English, Mandarin, Spanish, and Arabic dominate as global lingua francas.
Regional Scale
In Sub-Saharan Africa, Swahili serves as a regional lingua franca across East Africa.
Local Scale
Individual neighborhoods in Nairobi may use Kikuyu, Luo, or Sheng (a local creole).

Example: Population Density

Density (people/km²) 59 Global 347 Japan 6,363 Tokyo Scale of Analysis

The Ecological Fallacy

The ecological fallacy occurs when you apply conclusions drawn from aggregate data (large scale) to individuals (small scale). Just because a country has a high average income does not mean every person in that country is wealthy.

Fallacious
"Country X has high GDP per capita, so everyone there must be rich."
Correct
"Country X has high GDP per capita, but income distribution varies across regions and individuals."

The Modifiable Areal Unit Problem

The MAUP describes how statistical results can change depending on how the geographic boundaries of the study area are drawn. Redrawing boundaries can create, hide, or alter spatial patterns.

Boundary A 4 2 2 4 → Redraw Boundary B 5 4 3 Same data, different boundaries = different patterns

Real-World Application: Gerrymandering

Gerrymandering is a powerful real-world example of the MAUP. By redrawing the boundaries of voting districts, political parties can manipulate which candidates win elections: the same voters produce different outcomes depending on how district lines are drawn.

Scale Connection
At the state level, voter preferences may be balanced. At the district level, boundaries can make one party dominate.
Key Takeaway
Boundaries matter. The geographic units we choose can determine the conclusions we draw from data.

Common Mistakes

✕ Confusing Scale Types
Mixing up scale of analysis (conceptual framework) with map scale (mathematical ratio like \(1:24{,}000\)).
✕ Ecological Fallacy
Assuming broad-scale data applies to individuals: "The country is rich, so all citizens are wealthy."
✕ Ignoring MAUP
Forgetting that changing geographic boundaries can alter statistical results and spatial patterns entirely.

Summary Table

Scale Detail Level Example Topic Pitfall
Local Highest Neighborhood crime rates Too narrow to generalize
National Moderate Fertility rates by country Hides regional variation
Regional Lower Trade blocs (EU, ASEAN) Obscures national differences
Global Lowest Climate change impacts Ecological fallacy risk

Question 1 of 5

A geographer studying income inequality finds that a country has an average GDP per capita of $50,000. Upon closer examination, the geographer discovers that one region accounts for 80% of the nation's wealth. This discrepancy best illustrates which concept?

Question 2 of 5

Which of the following is the best definition of the ecological fallacy?

Question 3 of 5

A researcher changes the boundaries of census tracts in a study area and finds that poverty rates in each new zone are significantly different from the original zones. This problem is known as:

Question 4 of 5

At which scale of analysis would a geographer most likely study the spread of Swahili as a lingua franca across East Africa?

Question 5 of 5

How does scale of analysis differ from map scale?