Just because some variables associate does not necessarily mean that they cause each other. This is critical to remember.

In the first of the above examples, water sales in the retail chain do not cause ice-cream sales. Rather, both are more likely caused by the outside temperature – hot people tend to buy things that will cool them down and hydrate them, like water and ice cream. Blond hair does not cause people to have blue eyes or vice versa – it’s a pigment that causes both and creates an association between them.

On the other hand, we can sometimes infer that certain variables may cause each other when they are found to associate. Finding that customer size associates with sales may have an underlying causal logic. Perhaps the size of the customer partly dictates their needs and resources, both things that may affect sales of services to the customer. We will discuss more about these relationships later in the book, in sections such as the ones on regression.