Revision Library
Choose a subject
Menu
GCSE Statistics · Topic guide

Correlation and Causation (Higher)

Correlation and causation are two different ideas: correlation is a statistical relationship where two variables change together, measured by a coefficient between -1 and 1, while causation means one variable directly causes a change in the other. In GCSE Statistics, Higher tier questions often give a correlation coefficient and ask why a strong value still does not prove causation.

HigherCorrelation and Time SeriesEdexcelAQACCEA

Before you start

Make sure you're comfortable with these topics first:

Method

  1. Read the correlation coefficient given (or calculated from Spearman's rank or the PMCC) and describe its strength and direction.
  2. State that the coefficient only measures how closely the two variables change together, not whether one causes the other.
  3. Look for a plausible confounding (third) variable that could explain both variables changing together.
  4. Consider whether the direction of any causal link could be reversed, or whether the relationship could be coincidental.
  5. Check the sample size and time period the coefficient was calculated from; a coefficient based on very little data is less reliable evidence of a relationship.
  6. Use the context given in the question to write a specific explanation, rather than a generic statement that correlation is not causation.

Common mistakes

  • MistakeTreating a correlation coefficient close to zero as proof that the two variables have no relationship of any kind.

    InsteadA value near zero shows only that the points have no steady upward or downward trend. The variables could still be related in another way, such as a U-shaped pattern, so write 'no correlation' and avoid 'no relationship'.

  • MistakeAnswering with the stock sentence 'correlation does not mean causation' and nothing more.

    InsteadApply it to the context: name a specific third variable, a reversed link or a coincidence, using the words of the question, and say how it could produce both patterns.

  • MistakeSuggesting a third variable that could not affect both variables, such as the price of ice cream to explain a link between ice cream sales and sunburn at a beach.

    InsteadTest a suggestion by asking whether it could change both variables. Hot, sunny weather could raise both the ice creams sold and the cases of sunburn, so it is a sensible third variable.

Worked examples

Example 1

A researcher calculates a correlation coefficient of 0.92 between the number of libraries and the number of universities in different regions of a country. A politician claims that building more libraries causes more universities to be founded. Explain why this conclusion is not necessarily correct.

  1. A coefficient of 0.92 shows a very strong positive correlation between the two variables.
  2. This only shows the two variables tend to increase together across the regions studied, not that one causes the other.
  3. Identify a plausible third variable: regions with a larger population or greater investment in education are likely to have both more libraries and more universities.
  4. This third variable (population size or education investment) is a more likely common cause of both increases.
  5. Final answer: the strong correlation (0.92) does not prove causation; a third variable such as regional population size or education investment is more likely to explain why both figures rise together.

Example 2

For each of 10 consecutive years, a researcher records the average number of streaming subscriptions per household and the average price of a takeaway meal in the same country. The correlation coefficient between the two variables is 0.94. A blogger writes: 'Streaming subscriptions are making takeaways more expensive.' Describe the correlation and explain why the data do not convincingly support the blogger's claim.

  1. A coefficient of 0.94 is close to 1, so there is a strong positive correlation: in years with more subscriptions, the average takeaway price tends to be higher.
  2. Both variables were recorded over the same years, so each may simply have risen as time passed. A shared upward trend over time can give a high coefficient even when neither variable affects the other.
  3. A third variable is also plausible: general price rises over the period would push up takeaway prices whatever happened to streaming, while growing use of the internet would push up subscriptions.
  4. There are only 10 pairs, all from one country, which is a small amount of evidence for a lasting relationship. Nothing in the data shows that changing one variable would change the other.
  5. Final answer: there is a strong positive correlation (0.94), but it is more likely to reflect both variables changing over time, together with rising prices generally, than streaming making takeaways dearer, so the data do not convincingly support the blogger's claim.

Practice questions

Q1[1 mark]A correlation coefficient of 0.95 is calculated between two variables. What does this value tell you about the strength of the relationship?Show answer

Answer: It shows a very strong positive correlation (close to 1).

Got it right?
Your answerWork it out on paper, then show the answer and mark yourself.
Q2[1 mark]State the term used for a variable that affects both variables in a study, potentially producing a correlation without direct causation.Show answer

Answer: A confounding (third) variable.

Got it right?
Q3[1 mark]A study calculates a correlation coefficient of -0.85 between hours of television watched and exam grade. Does this prove watching television causes lower grades? Give a reason.Show answer

Answer: No; correlation does not prove causation, a third variable (e.g. less time spent revising) could explain both.

Got it right?
Your answerWork it out on paper, then show the answer and mark yourself.
Q4[1 mark]Two variables have a correlation coefficient of 0.08. What does this suggest about a causal link between them?Show answer

Answer: Very little to no linear relationship, so there is no evidence here to support a causal link.

Got it right?
Your answerWork it out on paper, then show the answer and mark yourself.
Q5[1 mark]A correlation coefficient of 0.99 is found between two variables using only 4 data points. Give one reason to be cautious about this result.Show answer

Answer: The sample size is very small, so the strong correlation could be due to chance rather than a real relationship.

Got it right?
Your answerWork it out on paper, then show the answer and mark yourself.
Q6[2 marks]Explain why a strong correlation coefficient calculated over a very short time period should be treated with caution when claiming causation.Show answer

Answer: A short time period may not capture enough variation and could reflect a temporary coincidence rather than a genuine, lasting relationship.

Got it right?
Your answerWork it out on paper, then show the answer and mark yourself.

Exam-style questions

Written in the style of a GCSE Statistics exam paper, with a full mark scheme.

Exam tips

  • Use cautious wording that matches the evidence: 'suggests', 'is consistent with' and 'does not show that' are safer than 'proves' or 'causes'.
  • Use every detail the question gives you, such as a small number of pairs, both variables rising over several years, or a group that may not be typical. Each is evidence you can comment on.
Q1[3 marks]A study calculates a correlation coefficient of 0.78 between the number of police officers and the number of reported crimes in different towns. A councillor claims that employing more police officers causes more crime. Assess this claim.Show mark scheme

Tick each line you got.

Nothing ticked yet - 3 available

Got it right?
Q2[2 marks]Give two reasons why a strong correlation coefficient between two variables might not indicate that one causes the other.Show mark scheme

Tick each line you got.

Nothing ticked yet - 2 available

Got it right?
Q3[4 marks]A gym chain calculates a correlation coefficient of 0.85 between the number of coffee shops and the number of gym memberships in different towns. The marketing team claims that coffee shops encourage people to join gyms. (a) State what a correlation coefficient of 0.85 indicates about the relationship. (b) Explain whether the marketing team's claim is justified.Show mark scheme

Tick each line you got.

Nothing ticked yet - 4 available

Got it right?

See real GCSE Statistics past-paper questions, with official mark schemes 

This topic in real past papers

We have tagged some boards' past papers question by question (not every board or year yet). This topic comes up in:

Free printable worksheet

Want more practice on paper? Download the Correlation and Causation (Higher) worksheet pack - 26 pages of exam-style questions with a full mark scheme. Free to download, no card or account. Print it for personal and classroom use.

This topic is chapter 11 of GCSE Statistics Higher Essentials 2, the whole course as one free printable PDF. Its worksheet is also chapter 3 of GCSE Statistics Higher Topic Workbook 6 and chapter 21 of GCSE Statistics Higher 10-Minute Tests 2, free printable practice books.

Other versions of this worksheet:

Keep going

Not quite what you needed?

Tell us what is missing on Correlation and Causation (Higher), or which topic to write up next.

Build a full practice pack.