A correlation is a link between two things. Evidence is required to establish a correlation between a factor and an outcome. If an outcome happens when a factor is present, and does not happen when the factor is absent, there is a correlation. Other factors that could affect the outcome also need to be considered.
The graph shows that as the number of cigarettes a person smokes increases, the number of deaths from cancer also increases. It shows a correlation, suggesting that the more a person smokes the more likely they are to get lung cancer.
Individual cases do not provide convincing evidence for or against a correlation. For example, there may be a 90-year-old who has smoked 40 cigarettes a day and not developed lung cancer. There may also be a 40-year-old who has never smoked a cigarette but who has developed lung cancer. These are exceptions - they do not disprove the correlation. The 90-year-old is lucky as there has been a correlation between smoking and lung cancer.
The fact that there is a correlation between a factor and an outcome does not necessarily mean this factor causes the outcome. Data must be collected and must provide evidence to prove that the factor has caused the outcome.
There is a correlation between the pollen count in the air and the incidence of hay fever, for instance. The pollen count increases from spring onwards, reaching a peak in mid-summer. It is therefore possible that pollen causes hay fever.
There is also a correlation between the amount of ice cream sold during the summer and the number of hay fever cases. But nobody would suggest that eating ice cream causes hay fever.
To show a causal link, scientists must find evidence that scientifically explains the connection.
If there is no scientific explanation then there is only a correlation. It cannot be shown that the factor causes the outcome.
There is a correlation between carbon dioxide levels and global average temperatures. What scientific explanation could provide evidence for a causal link between these factors?
The greenhouse effect explains how increasing levels of carbon dioxide in the atmosphere leads to increases in temperature on Earth. This is evidence of a causal link.
Sometimes a change in a factor leads to an outcome, but not in all cases. Scientists say that a change in the factor increases the risk of the outcome. This is very common when discussing causes of ill-health.
For example, when nitrogen dioxide levels stay high for several days, more people have asthma attacks. However, not everyone has an asthma attack. This type of correlation describes how a factor is connected with an increase in the risk of a particular outcome.
Scientists then try to find a causal link to explain the connection. If no causal link is known, this remains a correlation. It cannot be scientifically shown that the factor causes the outcome.
Data shows that children who live near major roads are more likely to suffer from asthma.
Why is this finding a correlation and not a causal link?
These data show that living near a major road is associated with an increase in the risk of a child having asthma. This is a type of correlation. Living near a major road does not directly cause asthma, so there is not causal link between the two.
What research might scientists carry out to try to find a causal link between living near a major road and asthma?
Scientists could research the effects of air pollution on the breathing system of children.