Evaluating a Choropleth Map
Learn how to read, evaluate, and critically interpret choropleth maps—how color patterns reflect data choices, why class breaks matter, and when a map actually supports its conclusion.
What you'll do in this lesson
A voice-first session with the Crimsora tutor on Evaluating a Choropleth Map, then targeted practice and FRQs — with the tutor adapting to where you get stuck.
What this lesson covers
A choropleth map uses color shading to show how values change across regions. You've probably seen one showing population density by state, or income levels by county. But here's the catch: the same real-world data can look dramatically different depending on how the mapmaker set up the colors. In this lesson, you'll learn to read a choropleth carefully, spot how design choices shape the story it tells, and decide whether a map's colors actually prove what someone claims.
Reading a Choropleth Map: Classes and Break Values
A choropleth map groups regions into classes—each class is assigned a color, and all regions in that class get the same shade. The break values are the numbers that divide one class from another. For example, a map might show "infant mortality per 1,000 births" with classes like: 0–10 (lightest), 11–25 (light), 26–50 (dark), 51–100 (darkest). The break values here are 10, 25, and 50.
To read a choropleth, first identify the classes and their ranges. Then locate your region on the map, find its color, and match that color to the legend to see what class it belongs to. This tells you the region's data falls somewhere within that range—but not the exact value. That's important: a choropleth shows approximate relationships, not precision.
When you're given a region's actual value, you can determine which class it belongs to by comparing that value to the break values. A country with a literacy rate of 78% belongs in the 75–90% class if the breaks are at 60, 75, and 90. This skill lets you check whether a map is drawn correctly and understand which regions truly share the same category.
To read a choropleth, first identify the classes and their ranges. Then locate your region on the map, find its color, and match that color to the legend to see what class it belongs to. This tells you the region's data falls somewhere within that range—but not the exact value. That's important: a choropleth shows approximate relationships, not precision.
When you're given a region's actual value, you can determine which class it belongs to by comparing that value to the break values. A country with a literacy rate of 78% belongs in the 75–90% class if the breaks are at 60, 75, and 90. This skill lets you check whether a map is drawn correctly and understand which regions truly share the same category.
How Class Breaks Shape the Pattern
The same data can look completely different if you move the break values. Imagine a map of test scores across 20 schools. If you use breaks at 50, 60, 70, and 80, you might see most schools in the 70–80 range (a moderate problem). But if you move the breaks to 72, 74, 76, and 78, suddenly the same schools split into many classes, and the map looks more fragmented and urgent.
Why does this matter? Because the visual pattern influences how viewers feel about the data. When breaks are too far apart, real differences disappear into a single color—and viewers miss important variation. When breaks are too close, minor differences look like major divides. A well-designed choropleth uses breaks that reflect genuine clustering or meaningful policy thresholds (like the World Health Organization's malnutrition cutoff lines).
When you evaluate a choropleth, always ask: Are these break values reasonable? Do they hide important differences, or do they highlight meaningful ones? A mapmaker who wants to downplay a problem might choose breaks that compress the variation; one who wants to alarm viewers might spread them out. Neither is necessarily dishonest—it's a choice—but it's your job to notice it and judge whether the choice fits the question being asked.
Why does this matter? Because the visual pattern influences how viewers feel about the data. When breaks are too far apart, real differences disappear into a single color—and viewers miss important variation. When breaks are too close, minor differences look like major divides. A well-designed choropleth uses breaks that reflect genuine clustering or meaningful policy thresholds (like the World Health Organization's malnutrition cutoff lines).
When you evaluate a choropleth, always ask: Are these break values reasonable? Do they hide important differences, or do they highlight meaningful ones? A mapmaker who wants to downplay a problem might choose breaks that compress the variation; one who wants to alarm viewers might spread them out. Neither is necessarily dishonest—it's a choice—but it's your job to notice it and judge whether the choice fits the question being asked.
Counts Versus Rates: A Different Pattern, Same Numbers
A crucial decision in choropleth mapping is whether to show counts or rates. A count is a total: "5,000 people in poverty." A rate adjusts for population size: "20% of the population in poverty." The same poverty crisis can disappear or reappear depending on which one you map.
Example: State A has 50,000 people in poverty and a population of 500,000 (10% rate). State B has 100,000 people in poverty and a population of 5,000,000 (2% rate). A choropleth of counts shows State B with a darker color because it has more poor people. But a choropleth of rates shows State A darker because a higher percentage of its population is poor. Which tells the true story? That depends on your question. If you care about total aid needed, use counts. If you care about the severity of poverty where people actually live, use rates.
Mapmakers sometimes choose count maps to make large regions look worse, or rate maps to make small regions seem more prosperous. When you see a choropleth, check the legend: Is it showing a count or a rate? Then ask yourself whether that choice makes sense for the question being answered.
Example: State A has 50,000 people in poverty and a population of 500,000 (10% rate). State B has 100,000 people in poverty and a population of 5,000,000 (2% rate). A choropleth of counts shows State B with a darker color because it has more poor people. But a choropleth of rates shows State A darker because a higher percentage of its population is poor. Which tells the true story? That depends on your question. If you care about total aid needed, use counts. If you care about the severity of poverty where people actually live, use rates.
Mapmakers sometimes choose count maps to make large regions look worse, or rate maps to make small regions seem more prosperous. When you see a choropleth, check the legend: Is it showing a count or a rate? Then ask yourself whether that choice makes sense for the question being answered.
Does the Map Support the Conclusion?
A map is sometimes presented as evidence for a claim: "Europe is wealthier than Africa," or "Climate change is making droughts worse worldwide." Your job is to judge whether the map actually proves that claim.
Here's how: First, check what data the map actually shows. A map titled "GDP per capita by country" tells you about average income, not wealth distribution or quality of life. Second, look at the data choices. If the map shows only a five-year window, it might miss long-term trends. If the breaks are set to emphasize one region, the visual pattern might mislead.
Third, check the regions covered. A map showing "global poverty" that only includes countries with available data might be missing the poorest regions. Finally, ask whether other explanations fit the pattern. A map showing high fertility rates in Africa and low rates in Europe could support "development causes fertility decline," but it could also reflect differences in education access, women's workforce participation, or healthcare availability—all shown on different maps.
A good choropleth is evidence, not proof. It shows a pattern worth investigating, but by itself it cannot rule out alternative explanations or confirm cause and effect.
Here's how: First, check what data the map actually shows. A map titled "GDP per capita by country" tells you about average income, not wealth distribution or quality of life. Second, look at the data choices. If the map shows only a five-year window, it might miss long-term trends. If the breaks are set to emphasize one region, the visual pattern might mislead.
Third, check the regions covered. A map showing "global poverty" that only includes countries with available data might be missing the poorest regions. Finally, ask whether other explanations fit the pattern. A map showing high fertility rates in Africa and low rates in Europe could support "development causes fertility decline," but it could also reflect differences in education access, women's workforce participation, or healthcare availability—all shown on different maps.
A good choropleth is evidence, not proof. It shows a pattern worth investigating, but by itself it cannot rule out alternative explanations or confirm cause and effect.
Common Mistakes When Reading Choropleth Maps
Students often misread choropleths in a few predictable ways. First, confusing the class with the exact value. If a region is colored dark because it falls in the 50–100 class, that doesn't tell you whether the value is 51 or 99. Don't imagine false precision.
Second, forgetting that color intensity doesn't always match importance. A map colored from white to black will make viewers think darker regions are "more" of something, but if the map uses arbitrary colors like orange for one class and purple for another, the color itself carries no meaning—only the legend does.
Third, assuming correlation is causation. If a map shows both deforestation and wildlife loss in the same regions, that's a useful pattern, but it doesn't prove deforestation causes the loss. There might be a third factor (extreme poverty, weak enforcement, recent settlement) driving both.
Fourth, ignoring null data. Many maps leave some regions blank or gray because data wasn't available. These missing regions might not be "zero"—they might just be unmeasured. A choropleth of literacy rates that omits conflict zones will look rosier than reality.
Finally, treating a snapshot as a trend. A 2022 map of unemployment is a moment in time. Without maps from 2021 and 2023, you can't tell whether unemployment is rising, falling, or stable.
Second, forgetting that color intensity doesn't always match importance. A map colored from white to black will make viewers think darker regions are "more" of something, but if the map uses arbitrary colors like orange for one class and purple for another, the color itself carries no meaning—only the legend does.
Third, assuming correlation is causation. If a map shows both deforestation and wildlife loss in the same regions, that's a useful pattern, but it doesn't prove deforestation causes the loss. There might be a third factor (extreme poverty, weak enforcement, recent settlement) driving both.
Fourth, ignoring null data. Many maps leave some regions blank or gray because data wasn't available. These missing regions might not be "zero"—they might just be unmeasured. A choropleth of literacy rates that omits conflict zones will look rosier than reality.
Finally, treating a snapshot as a trend. A 2022 map of unemployment is a moment in time. Without maps from 2021 and 2023, you can't tell whether unemployment is rising, falling, or stable.
Key terms
- Choropleth map.
- A map that uses color shading to show how a variable's value changes across regions. Regions are grouped into classes, and each class is assigned a distinct color.
- Class (in choropleth mapping).
- A category that groups regions together based on their values. For example, a class might be 'population density 50–100 people per square km,' and all regions in that range share one color.
- Break value.
- The number that marks the boundary between two classes. In a map with breaks at 25, 50, and 75, regions with values 0–25 form one class, 26–50 form another, and so on.
- Rate.
- A value adjusted for population or area size, expressed as a proportion or percentage. Examples: 'deaths per 100,000 people' or 'unemployment rate.' Rates allow fair comparison between regions of different sizes.
- Count.
- The total number of something in a region, unadjusted for population or size. Example: 'total number of people in poverty.' Large regions usually have larger counts, even if rates are lower.
- Legend.
- The key on a map that explains what each color represents, including the class ranges and break values.
- Null data.
- Missing or unavailable data for one or more regions, often shown as blank or gray on a map. Null data can hide patterns or create misleading impressions.
- Visual hierarchy.
- The way a map's design (colors, shading intensity, size) influences what viewers notice first. A darker color naturally draws the eye and can make a problem seem more urgent, even if the data doesn't warrant that impression.
Worked example
The map below shows "infant mortality rate per 1,000 live births" by region. The classes and break values are: 0–20 (light blue), 21–50 (medium blue), 51–100 (dark blue), and 101+ (black). The actual values for five regions are: Region A = 18, Region B = 52, Region C = 45, Region D = 105, Region E = 19. (a) Which regions share the same class? (b) If the mapmaker moved the breaks to 0–40, 41–80, 81–120, how would the pattern change? (c) Would a map of infant mortality count (total infants) instead of rate likely show the same regional pattern? Explain.
(a) Start by matching each value to its class. Region A = 18 falls in 0–20 (light blue). Region B = 52 falls in 51–100 (dark blue). Region C = 45 falls in 21–50 (medium blue). Region D = 105 falls in 101+ (black). Region E = 19 falls in 0–20 (light blue). So Regions A and E share the same class (light blue, 0–20).
(b) Now check the new breaks. Using 0–40, 41–80, 81–120: Region A = 18 is in 0–40 (still light, same class). Region B = 52 is in 41–80 (now medium, not dark). Region C = 45 is in 41–80 (medium, same as before). Region D = 105 is in 81–120 (still dark, same class). Region E = 19 is in 0–40 (still light, same class). The new breaks would shift Region B from dark to medium, making the map appear less urgent—fewer regions in the highest class. The overall pattern would look less severe.
(c) A count map shows total infant deaths in each region, not the rate. A large region with a population of 10 million and a rate of 25 per 1,000 births would have roughly 250,000 infant deaths. A small region with a population of 100,000 and a rate of 80 per 1,000 births would have roughly 8,000 infant deaths. The count map would show the large region in a darker class, even though its rate is lower. This inverts the pattern—the areas with the worst outcomes (highest rates) might appear lighter because they are smaller. A rate map is fairer for comparing the severity of child health across regions of different sizes.
(b) Now check the new breaks. Using 0–40, 41–80, 81–120: Region A = 18 is in 0–40 (still light, same class). Region B = 52 is in 41–80 (now medium, not dark). Region C = 45 is in 41–80 (medium, same as before). Region D = 105 is in 81–120 (still dark, same class). Region E = 19 is in 0–40 (still light, same class). The new breaks would shift Region B from dark to medium, making the map appear less urgent—fewer regions in the highest class. The overall pattern would look less severe.
(c) A count map shows total infant deaths in each region, not the rate. A large region with a population of 10 million and a rate of 25 per 1,000 births would have roughly 250,000 infant deaths. A small region with a population of 100,000 and a rate of 80 per 1,000 births would have roughly 8,000 infant deaths. The count map would show the large region in a darker class, even though its rate is lower. This inverts the pattern—the areas with the worst outcomes (highest rates) might appear lighter because they are smaller. A rate map is fairer for comparing the severity of child health across regions of different sizes.
Practice questions
A choropleth map shows 'percentage of the population with internet access' in 30 countries, with classes 0–25%, 26–50%, 51–75%, and 76–100%. Country X has an actual value of 48%. Which class does Country X belong to, and how would you know if the map is drawn correctly?
Answer: Country X belongs to the 26–50% class because 48% falls within that range. To check if the map is correct, you would locate Country X on the map, identify the color of the 26–50% class in the legend, and verify that Country X is colored that way. If it is, the map is drawn correctly; if it shows a different color, there's an error.
This tests your ability to match a value to its class range using break values. It also emphasizes the key skill of reading a legend and validating what you see. The break values here are 25, 50, and 75. Any value from 26 to 50 belongs to the second class. This is not about algebra—just careful comparison of the value (48) to the boundaries (26 and 50).
A mapmaker creates two versions of a choropleth showing 'unemployment by county.' Version 1 uses break values 3%, 6%, 9%, and 12%. Version 2 uses break values 5.5%, 7%, 8%, and 9%. Both are based on the same actual unemployment data. (a) Which version would make the unemployment problem appear more urgent? (b) What question would you ask to judge whether the chosen breaks are fair?
- Version 1; ask whether the breaks match natural clusters in the data
- Version 2; ask whether the breaks are spaced equally
- Version 1; ask what story the mapmaker is trying to tell
- Version 2; ask how many counties fall into each class
Answer: Version 1; ask whether the breaks match natural clusters in the data
Version 2 uses breaks much closer together (5.5% to 9%, a range of 3.5 percentage points), creating more classes and visual fragmentation. This makes small differences look like big divides, which can heighten alarm. Version 1 is more spread out and typically clearer. However, spreading out breaks isn't automatically 'fair'—the key is whether the breaks reflect real boundaries in the data or meaningful policy thresholds, not whether they're equally spaced. A fair question asks whether the breaks match natural clustering (e.g., do counties naturally cluster at 3–6%, 6–9%, etc.?) or policy importance, not just equal spacing. Version 2 being more spaced equally doesn't make it fairer; it makes it harder to interpret fine differences.
FAQ
- If two regions have the same color on a choropleth, does that mean they have exactly the same value?
- No. Regions with the same color belong to the same class—they fall within the same range of values. For example, if two regions are both colored 'medium blue' for the 21–50% class, one might be 25% and the other 48%. The color tells you the range, not the exact value. That's why choropleths show approximate relationships, not precision. If you need exact values, you need a data table, not just the map.
- Why would a mapmaker use a rate instead of a count?
- Rates adjust for population size, so they show fair comparison. If you map the count of people in poverty, large cities or states will almost always be darker (more people = more poverty in total). But that doesn't mean poverty is worse there—it just means there are more people. A rate map shows the percentage in poverty, so you can compare 'severity where people actually live' fairly. Use counts when you care about total resources needed; use rates when you care about how severe a problem is for the people living there.
- A map shows 'COVID-19 deaths by country,' but three countries have blank (gray) regions because they didn't report data. Does this mean the virus wasn't there?
- No. Null data (missing information) doesn't mean zero. Those countries might have had deaths they couldn't measure or chose not to report. A choropleth with blank regions is incomplete. When you see blank regions, note them and remember that the map doesn't show the full picture. The actual pattern might be different if those countries' data were included.
- Can a choropleth map prove that one thing causes another?
- No. A choropleth shows correlation—that two patterns appear together in the same regions. But correlation is not causation. If a map shows high deforestation and high poverty in the same regions, that's interesting and worth investigating. But it could mean deforestation causes poverty, poverty causes deforestation, or both are caused by a third factor like weak government enforcement or recent settlement. You'd need additional evidence—data over time, comparisons across similar regions, or case studies—to prove cause and effect.
Learn this with a teacher, not a page
The Crimsora tutor teaches Evaluating a Choropleth Map live — explaining on a whiteboard, asking you questions, and adapting to where you get stuck.