tendency to follow a straight line. It is positive because the straight line has a
positive gradient – it is pointing in a northeasterly direction.
Scatter diagramCause and correlation
Finding a strong correlation between two variables is not sufficient to claim
that one causes the other. There may be a cause and effect relation between the
two variables but this cannot be claimed on the basis of numerical evidence
alone. On the cause/correlation issue it is customary to use the word ‘association’
and wise to be wary of claiming more than this.
In the sunglasses and ice cream example, there is a strong correlation
between the sales of sunglasses and that of ice cream. As the sales of sunglasses
increase, the number of ice creams sold tends to increase. It would be ludicrous
to claim that the expenditure on sunglasses caused more ice creams to be sold.
With correlation there may be a hidden intermediary variable at work. For
example, the expenditure on sunglasses and on ice creams is linked together as a
result of seasonal effects (hot weather in the summer months, cool weather in
the winter). There is another danger in using correlation. There may be a high
correlation between variables but no logical or scientific connection at all. There