This article covers 3.1.4 Research methods of the AQA GCSE Psychology specification.
Correlation and Co-Variables
- A correlation is an association between two variables.
- The two variables being measured are called co-variables.
- Neither variable is manipulated; both are simply measured as they are.
- You then look at whether the two co-variables tend to rise and fall together.
Key Idea
- A correlation looks at how two things vary together.
- You do not need to calculate a coefficient at this level.
Positive, Negative and Zero Correlations
- Positive correlation
- As one co-variable rises, so does the other.
- Negative correlation
- As one co-variable rises, the other falls.
- Zero correlation
- There is no clear relationship between the two co-variables.
Example
- Negative: the more time a teenager spends on a screen at night, the fewer hours they sleep.
- Positive: the more hours a pupil revises, the higher their test score tends to be.
Reading a Scatter Diagram
Definition
Scatter diagram
A graph that plots two variables against each other, with one point for each participant, used to show the correlation between them.
- A scatter diagram plots each pair of co-variable scores as a single point.
- Points rising to the right show a positive correlation.
- Points falling to the right show a negative correlation, and a shapeless cloud shows none.
Common Mistake
- Do not say a correlation proves that one variable causes the other.
- Do not confuse a scatter diagram with a bar chart or a line graph.
Strengths and Weaknesses of Correlations
- Strengths
- You can study variables that could not be manipulated in an experiment for practical or ethical reasons.
- A scatter diagram shows the direction and rough strength of a relationship at a glance.
- It is a useful starting point that can flag links worth testing later with an experiment.
- Weaknesses
- It cannot show cause and effect, because neither co-variable is controlled.
- A third variable may be driving both co-variables to change.
- It only picks up straight-line links, so a curved relationship can be missed and read as zero.
Note
- A strong correlation can be tempting to read as proof of cause, but it never is on its own.
- To test cause, you would need to manipulate one variable in a controlled experiment.
Correlation Is Not Causation
- A correlation shows two variables are linked, not that one causes the other.
- A third factor could be driving both variables to change.
- Only a controlled experiment can properly test for cause.
Self review
- What is a correlation, and what are co-variables?
- Describe a positive and a negative correlation.
- What does a scatter diagram with points rising to the right show?
- Give one strength and one weakness of using a correlation.
- Why does correlation not prove causation?