Skip to content
VirtusAcademy

Correlation and Causation

FoundationHigherAQAEdexcel

Learn Correlation and Causation for GCSE Statistics with this free worksheet and full mark scheme — Foundation and Higher exam-style questions with worked answers for AQA and Edexcel. Correlation between two variables does not prove that one causes the other.

Free downloads

These worksheets and mark schemes are original, written for Virtus Academy and checked against the current AQA and Edexcel specifications. Every worksheet comes with a full mark scheme.

Foundation

WorksheetFoundation · PDFMark SchemeFoundation · PDF

Higher

WorksheetHigher · PDFMark SchemeHigher · PDF

Topic overview

Correlation between two variables does not mean that one causes the other. This distinction is examined directly and repeatedly.

A correlation may arise for several reasons. One variable may genuinely cause the other. Both may be caused by a third variable, which is called a confounding variable. Or the correlation may be pure coincidence, particularly with small samples.

Ice cream sales and drowning incidents correlate strongly, but neither causes the other — hot weather causes both, and is the confounding variable. To establish causation rather than correlation, a controlled experiment is needed, in which other variables are held constant.

Revision notes

Why correlation is not causation

A correlation between two variables can arise in three ways.

One genuinely causes the other. Both are caused by a third, confounding variable. Or the correlation is coincidental, which is more likely with small samples.

Confounding variables

A confounding variable causes changes in both of the variables being studied.

Ice cream sales and drowning incidents correlate, but hot weather causes both. Identifying a plausible confounding variable is a standard exam requirement.

Establishing causation

A controlled experiment is needed, in which all other variables are held constant.

Only then can a change in the response variable be attributed to the explanatory variable. Observational data alone can suggest a relationship but cannot establish cause.

Key points

  • Correlation does not imply causation.
  • One variable may genuinely cause the other.
  • A third variable may cause both.
  • This is called a confounding variable.
  • Correlation may also be coincidence.
  • A controlled experiment is needed to show causation.

Worked examples

Example 1

Ice cream sales and drowning incidents show strong positive correlation. Explain why this does not mean one causes the other. [2 marks]

Working

Both are caused by a third variable, hot weatheridentify the confounding variable
so the correlation arises without either variable causing the otherexplain why causation is not shown

Example 2

Explain what is meant by a confounding variable. [2 marks]

Working

A confounding variable is a third variable that causes changes in both of the variables being studieddefine the term
producing a correlation between them even though neither causes the otherexplain its effect

Example 3

Explain what would be needed to establish that one variable causes another. [2 marks]

Working

A controlled experiment in which all other variables are held constantstate what is required
so that any change in the response variable can be attributed to the explanatory variable aloneexplain why that establishes causation

Common mistakes

  • Concluding causation from correlation.

    Correlation shows a relationship only, never a cause.

  • Not naming a plausible confounding variable.

    Questions usually want a specific suggestion.

  • Saying correlation is always meaningless.

    It may reflect a genuine causal link — it simply does not prove one.

  • Forgetting coincidence as a possibility.

    Especially with small samples, correlation can arise by chance.

Exam tips

  • Never conclude causation from correlation alone.
  • Suggest a specific confounding variable when asked.
  • Name the controlled experiment as the way to show causation.
  • Learn the ice cream and drowning example.

Key terms

Causation
One variable genuinely causing a change in another.
Confounding variable
A third variable causing changes in both.
Coincidence
A correlation arising purely by chance.
Controlled experiment
A study holding other variables constant.

Related topics

Written and reviewed against the current AQA and Edexcel specifications. Spotted an error? Let us know.