A strong correlation describes an association in the data. It does not, by itself, show that changing one variable causes the other to change. This is a different judgement from deciding whether a regression prediction is useful.
Bridge the explanation
Sales of cold drinks and numbers of swimmers are positively correlated. A plausible third factor is hot weather: it can increase both quantities. Buying more cold drinks has not thereby been shown to cause more people to swim.
| Statement | What the data support |
|---|---|
| “Higher cold-drink sales tend to accompany more swimmers.” | An association, if the observed data support it. |
| “Increasing cold-drink sales causes more people to swim.” | A causal claim that correlation alone does not establish. |
| “The association can help a prediction.” | Possibly, if the fitted model is suitable and the prediction conditions are comparable. |
Give a contextual alternative explanation, such as a shared influence, reverse direction, or how the observations were selected. Do not stop at the slogan “correlation is not causation”; explain why the proposed causal reading is unsupported here.
Exam wording: “The correlation alone does not establish causation. Hot weather may increase both cold-drink sales and the number of swimmers, producing the observed association.” Prediction reliability and causation answer different questions.
Need help? Join our JC Math tuition classes.
Learn more