Data shows a correlation if the change in a factor is similar to the change in an outcome. Scientists must find a scientific explanation to conclude that a particular factor causes an outcome.

A correlation shows a connection between a factor and an outcome.

To describe a correlation in words state:

- the change in the factor
- how the outcome changes as this happens
- how the pattern of change in each factor is similar or different

Here is an example, with each marking point shown as [1]:

'As carbon dioxide levels increase [1], the global average temperature increases [1] and the increases follow a similar pattern [1].'

To spot correlations in data, look at how the numbers change in each column of the data table.

Factor | Outcome |
---|---|

As the values of the data here increase… | …the values of the data here increase in a similar way. |

The trend in the outcome is the same as the factor.

Factor | Outcome |
---|---|

As the values of the data here increase… | …the values of the data here decrease in an opposite way. |

The trend in the outcome is the opposite of the trend in the factor.

Data which does not show a correlation has no similar or opposite trend in the outcome.

Factor | Outcome |
---|---|

As the values of the data here increase… | …the values of the data here go up, down or do not change. |

The upwards trend is the factor not present in the outcome.

Graphs can also reveal correlations.

This graph shows increasing carbon dioxide levels in the atmosphere.

This graph shows a similar increase in global average temperature.

It is important that the graphs show similar trends. This is most easily seen if graphs are overlaid on top of each other.