Do two things tend to change together?
Correlation analysis measures whether changes in one thing are reliably associated with changes in another. The easiest correlations to think about are those where one measurement increases, and then a second one also consistently increases.
Negative Correlations
Negative correlations are present when an increase in one variable is associated with a decrease in another. For example, in our B2B Buyer Experience Research, we found that as the importance of the solution being purchased increased, the duration of the buyers’ purchase journey decreased (more important solutions were purchased in less time). We cannot say that Solution Importance itself caused the shortened buying cycle, but we do know they reliably move together.
Positive Correlations
In our own research on B2B marketer compensation, we saw marketers with more service time at their companies said it would take more money to get them to switch companies. This illustrates a positive correlation, meaning that both variables move in the same direction (longer tenure, more money required to switch companies).
Correlation is not causation
What correlations are not is an indication that one thing causes another. In the example above, while longer tenures tended to coincide with higher compensation expectations for switching companies, tenure itself didn’t cause the increased price, nor did the increased price cause longer tenures. During longer tenures, employees are likely to be establishing friendships, building loyalty to the organization, finding their routines, among much else. Thus, while more time allows for such experiences, it is not the time itself causing the relationship. All we can say, as with any correlation, is that changes in one thing (time at a company) are reliably associated with changes in another (money required to switch companies).