An unsupported correlation is a claimed link between two patterns that the evidence does not justify. This appears in data response questions, particularly when counts are compared across districts of different size. It belongs to graphs, photographs and spatial data.
How do you test a claimed link?
- State the pattern in plain words: “districts with more of X have more of Y”.
- Check the type of data. Are both columns counts, and do the places differ in size or population?
- Convert to rates where possible, such as per 1 000 people or per km².
- Look for a third factor that could drive both.
- Word the conclusion cautiously and say what further evidence would help.
Worked example
All data is invented. Six districts have these figures for tall buildings and heat complaints in a year, along with population.
| District | Tall buildings | Heat complaints | Population (thousands) |
|---|---|---|---|
| 1 | 2 | 11 | 20 |
| 2 | 5 | 17 | 30 |
| 3 | 9 | 24 | 40 |
| 4 | 14 | 33 | 60 |
| 5 | 20 | 41 | 80 |
| 6 | 26 | 55 | 100 |
Step 1, the pattern. Districts with more tall buildings report more complaints. It looks like a strong link.
Step 2, check the size. Population also rises down the table, so larger districts will report more of almost anything.
Step 3, convert to a rate (complaints per 1 000 people, so complaints ÷ population in thousands):
- District 1: 11 ÷ 20 = 0.55
- District 2: 17 ÷ 30 ≈ 0.57
- District 3: 24 ÷ 40 = 0.60
- District 4: 33 ÷ 60 = 0.55
- District 5: 41 ÷ 80 ≈ 0.51
- District 6: 55 ÷ 100 = 0.55
Step 4, conclude. The rates stay between 0.51 and 0.60 and do not rise with tall buildings. The pattern in the counts is explained largely by population size, so the data does not support a link between tall buildings and heat complaints.
What mistake do students make?
Mistaken answer: “Tall buildings cause heat complaints, because districts with more tall buildings have more complaints.”
This reads a cause into a pattern in raw counts. Population drives both columns, and the rates show no matching rise.
The correction is to use cautious wording: “The number of complaints is higher where there are more tall buildings, but population is also higher there. Rates per 1 000 people are similar, so the data does not show that tall buildings are the cause.”
Check yourself
All data is invented.
1. A district has 9 complaints and 3 000 people. What is the rate per 1 000 people?
Show answer
3 000 people = 3 thousand. 9 ÷ 3 = 3 complaints per 1 000 people.
2. Name a third factor that might explain why districts with more cafes also have more bicycle thefts.
Show answer
Population or visitor numbers, or the amount of commercial land in the district. Any one of these is acceptable if you link it to both variables.
3. Rewrite “More schools cause more traffic” more cautiously.
Show answer
“Districts with more schools also have more traffic, which may be linked to population size or road layout. More evidence, such as traffic per 1 000 people, would be needed.”
Where this leads next
You now have five evidence skills, so test them together in the mixed practice set. Record any missed question in the mistake log and retest queue. If you want a teacher to question your conclusions and sharpen your wording, see our online one-to-one Geography tuition.