Ecological study
- Study design
It compares whole groups (municipalities, countries, years), not people. It is fast and cheap, but it says nothing about each individual.
What it is
In an ecological study the unit of analysis is a group: rates or average levels of exposure and disease are compared between populations or periods. It uses existing data (vital statistics, surveys, sales), so it is useful for exploring hypotheses and evaluating policies. Its big limitation is the ecological fallacy: an association between groups does not guarantee the same association between people.
How it is built
- The units are chosen: departments, countries, schools or years.
- Each unit gets a measure of exposure (for example, the percentage of smokers).
- Each unit gets a measure of disease (for example, the mortality rate per 100 000).
- The correlation or regression between the two measures is analysed at group level.
Worked example
- Four regions report their percentage of adult smokers and their lung cancer mortality per 100 000 people.
- Data: 10% → 20; 15% → 28; 20% → 37; 25% → 45 deaths per 100 000.
- Approximate slope: (45 − 20) / (25 − 10) = 25 / 15 ≈ 1.7 deaths per 100 000 per percentage point of smoking.
- So every 5 more points of smokers go with about 8 more deaths per 100 000 (1.7 × 5 ≈ 8).
- But the data do not say whether the people who died smoked: assuming so would be the ecological fallacy.
- Reading: an association between regions that justifies a cohort or case-control study with individual data.
Strengths and weaknesses
| Strengths | Weaknesses |
|---|---|
| Cheap and fast with existing data | Ecological fallacy |
| Compares countries, regions and periods | Cannot control individual confounding well |
| Useful for exposures that exist only at group level (laws, taxes, safe water) | Aggregate data may vary in quality between units |
| Generates hypotheses | Does not establish the time sequence for individuals |
When to use it
- Used to evaluate public health policies: for example, comparing rates before and after a law in several municipalities.
- In Colombia, DANE and SIVIGILA data by municipality allow ecological studies without contacting patients.
Common pitfalls
- Concluding that a smoker has X times the risk from region-level data.
- Forgetting that regions differ in many things at once (age, income, access to care) that can explain the association.
Test yourself
Think of the answer, then tap the question to see it.
What is the unit of analysis in an ecological study?
A group (municipality, country, school or period), not the person.
What is the ecological fallacy?
The error of assuming that an association seen between groups also holds for the individuals in those groups.
Countries that eat more of a food have more cases of a disease. Does this prove that people who eat it get sick more?
No. It is an ecological association; other features of those countries could explain it, and we do not know what the sick people ate.
Connections
Educational content on epidemiological and statistical methods; the examples use made-up numbers to teach the calculation. Not a clinical or legal guide: current Colombian rules are in the official sources.
Sources: The design, applications, strengths and weaknesses of descriptive studies and ecological studies (Health Knowledge (Faculty of Public Health), 2018); Principles of Epidemiology | Lesson 1 - Section 6 (Centers for Disease Control and Prevention, 2012).
Spotted a mistake?We are always improving for the community. Tell us what is wrong and our team will review it.Report a mistake
Representative model, for educational purposes only.