Research Methods

Correlation vs Causality

A correlation exists when there appears to be a dependent relationship between two variables. That is to say, two variables (or ÔthingsÕ) appear to change at the same time. This would therefore appear to suggest (but crucially does necessarily prove) a link between the two variables. For example, a sociologist may identify that girls perform better in a single sex education setting. We can therefore say that there is a correlation between single sex education and girlsÕ exam performance. However, further investigation would be needed to prove that there exists a causal
relationship here. That is to say, that attending a single sex school directly leads to (causes) girls to perform better. Hence, identifying a correlation is useful as it can provide the basis for an investigation into a causal relationship. But it is important to always remember that a correlation between two variables does not necessarily mean that one causes the other. More evidence is usually needed to prove causality.

Using an example taken from WebMD, this Khan Academy video (10 minutes long) explains the difference between correlation and causality by examining an article that looks at breakfast eating habits and obesity:

Discover More Understanding why correlation does not imply causality

From this, students could find other examples of studies (perhaps referred to in the health sections of newspapers) and identity whether the author seems to be suggesting that causality exists, when it is in fact merely correlation.