Introduction to Correlation and Causation
Welcome to one of the most exciting and useful topics in GCSE Statistics! Have you ever noticed that when the weather gets hotter, ice cream sales go up, and people also get more sunburns? Does eating ice cream cause sunburn? Of course not! But there is clearly a link between them.
In this chapter, you will learn how to describe links between variables and, more importantly, how to decide whether one thing actually makes another happen. This is a crucial skill not just for your exam, but for making sense of news reports, scientific claims, and everyday data.
Don't worry if this sounds a bit tricky right now. We will break everything down into simple, manageable steps with plenty of real-world examples!
1. Understanding Correlation
Correlation measures the strength and direction of a linear relationship between two variables.
Think of correlation as a measure of co-relation: when one variable changes, does the other variable tend to change as well?
Types of Correlation
When you plot bivariate data (data involving two variables, \(x\) and \(y\)) on a scatter diagram, you will typically see one of three patterns:
• Positive Correlation: As the independent variable (\(x\)) increases, the dependent variable (\(y\)) also tends to increase.
Example: The number of hours you revise and your test score. More revision generally leads to higher marks.
• Negative Correlation: As one variable (\(x\)) increases, the other variable (\(y\)) tends to decrease.
Example: The temperature outside and the amount of heating fuel used in a house. As it gets warmer, heating use goes down.
• Zero Correlation (No Correlation): There is no linear relationship between the two variables. The points are scattered randomly.
Example: A student's shoe size and their score on a maths test.
Strength of Correlation
Correlation can also be described by how closely the points cluster around a straight line of best fit:
• Strong Correlation: The data points lie very close to a straight line.
• Weak Correlation: The points follow a general upward or downward trend, but are spread out widely.
Memory Trick:
• Positive: Both go UP together (like an uphill slope /).
• Negative: One goes UP, the other goes DOWN (like a downhill slope \).
Key Takeaway: Correlation simply tells us that two variables move together in a recognisable pattern. It does not explain why they move together.
2. Understanding Causation (Cause and Effect)
Causation (also called cause and effect) means that a change in one variable directly causes a change in another variable.
Here are some clear examples of true causal relationships:
• Pushing down the accelerator pedal causes a car to speed up.
• Turning on the tap causes the water level in the sink to rise.
• Freezing water below \(0^\circ\text{C}\) causes it to turn into ice.
In each case, there is a clear physical or biological mechanism showing that variable \(A\) directly produces the outcome in variable \(B\).
Key Takeaway: Causation requires a proven, direct mechanism where one event makes the other event happen.
3. The Golden Rule: Correlation Does NOT Imply Causation
This is the single most important rule in statistical reasoning:
Just because two variables show a strong correlation, it does NOT mean that one variable causes the other.
Whenever you see a correlation between two variables, there are three main possibilities to consider:
Possibility 1: Variable \(A\) truly causes Variable \(B\)
Example: An increase in rainfall causes the water level in a reservoir to rise.
Possibility 2: A Third Variable (Lurking or Confounding Variable) causes both
Very often, two variables appear linked only because a third, hidden variable affects them both simultaneously.
Classic Example: There is a strong positive correlation between ice cream sales and the number of drowning incidents at the beach.
Does buying ice cream make people drown? No!
The third variable is the weather / temperature.
When the weather is hot and sunny:
1. More people buy ice cream.
2. More people go swimming in the sea, which leads to more drowning accidents.
Possibility 3: It is Pure Coincidence (Spurious Correlation)
Sometimes, two completely unrelated sets of data happen to follow the same trend by pure chance.
Did you know? Over several years, there was a strong positive correlation between the number of films starring Nicolas Cage and the number of people who drowned in swimming pools. They have absolutely nothing to do with each other—it is pure coincidence!
Quick Summary Box:
When you see correlation, ask yourself:
1. Does \(A\) cause \(B\)?
2. Does \(B\) cause \(A\)?
3. Is there a hidden third variable affecting both?
4. Is it just a complete coincidence?
Key Takeaway: Never jump to the conclusion that \(A\) causes \(B\) without controlled scientific experiments and evidence of a direct mechanism.
4. Answering Exam Questions: Reasoning and Discussing Results
In your CCEA GCSE Statistics exam, you will often be asked to evaluate statements made about data or scatter graphs.
How to Spot Faulty Reasoning
Look out for statements where someone sees a graph and makes an unjustified claim using causal words like "causes", "leads to", "makes", or "proves".
Example Exam Scenario:
A researcher plots a scatter diagram showing a strong positive correlation between the number of mobile phones owned in a household and the life expectancy of people in that household. The researcher concludes: "Buying more mobile phones will make you live longer."
How to answer:
1. State clearly that correlation does not mean causation.
2. Identify a sensible third variable (confounding factor). In this case, household income or wealth is the third variable: wealthier families can afford more mobile phones and also have better access to healthcare, nutrition, and living conditions.
3. Conclude that buying more phones will not directly increase life expectancy.
Phrases to Use in Your Exam
To score full marks, use clear, precise statistical language:
• "There is a positive/negative correlation between the two variables, but this does not imply causation."
• "The relationship could be influenced by a confounding variable, such as..."
• "The data shows an association, but there is no evidence of a direct causal link."
Common Mistakes to Avoid
• Mistake 1: Saying "There is no link" when a graph shows a strong correlation. There is a correlation; it just isn't necessarily a causal link.
• Mistake 2: Forgetting to suggest a realistic third variable when asked to explain why a claim might be misleading.
• Mistake 3: Using definitive words like "This proves that..." instead of "This suggests a correlation between...".
Key Takeaway: Always challenge claims of cause and effect in exam questions unless a valid scientific mechanism or controlled experiment is provided.
5. Quick Revision Checklist
Before you move on to practice questions, make sure you can answer these key questions:
• Can you define correlation in your own words?
• Can you define causation and give a real-life example?
• Can you explain why correlation does not guarantee causation?
• Can you identify a third (confounding) variable in a given real-world context?
• Can you explain the difference between a real causal link and a coincidental correlation?