Correlation vs Causation

A correlation is a pattern between two variables but without evidence suggesting that one causes the other. If there is a provable explanation that one variable influences the other, the correlation is causal, but without an explanation, the correlation may be non-causal.

For example, there is a correlation between pleasant summer days and ice-cream sales but also a correlation between the sales of ice cream and umbrellas. The first example is causal since the nice weather causes more people to go out and enjoy the sun than buying ice cream. The second example is non-causal since umbrella sales don't cause ice-cream sales to go down or vice versa; both are caused by the weather.

Correlations can even be coincidental, where there is no relation between the variables at all.