9.5 - Correlation & Cause

Correlation is not the same as cause

• A correlation is a relationship between two variables, where a change in one is linked to a change in the other.
• A correlation does not by itself prove that one thing causes the other.

Three types of correlation

1. Positive - As one variable increases, the other increases.
2. Negative - As one variable increases, the other decreases.
3. No correlation - There is no relationship between the two.

A correlation can arise for three reasons

1. Chance - The link is coincidental.
2. A third factor - Something else affects both variables.
3. A genuine cause - One variable really does cause the change in the other.

Identifying the cause:

• Only the last is a true cause, and a correlation alone cannot tell you which of the three you are seeing.
• Example: ice cream sales and drownings both rise in the same months, so they are correlated - but neither causes the other. A third factor, hot weather, raises both.

Showing that one factor causes another

• A correlation on its own is not enough to show that one factor causes a change in another.
• Two conditions have to be met before a causal link can be accepted:
  1. The other variables controlled - The investigation rules out the other factors that could explain the change, so the link is not down to something else.
  2. A plausible mechanism - There is a sensible explanation for how one factor brings about the change in the other.

Applying the two conditions

• Smoking and lung cancer - Meets both: controlled studies ruled out other factors, and there is a known mechanism by which the chemicals in tobacco smoke damage lung cells.
• Ice cream and drowning - Meets neither, so it stays a correlation only.