correlation

Correlation

A statistical measure of how strongly two variables move together — ranges from -1 (perfect inverse relationship) through 0 (no relationship) to +1 (perfect co-movement).

When you'd see it: Anywhere two variables are compared for relationship: marketing spend vs revenue, NPS vs retention, support volume vs churn. Correlation matrices show all pairwise correlations across many variables simultaneously.

Why it matters: Correlation is how most exploratory data analysis begins — identifying which variables move together is the first step toward understanding the system. But correlation is not causation. Finding a correlation is the start of an investigation, not the conclusion.

Common mistakes: Treating correlation as evidence of causation. Two variables can be highly correlated because one causes the other, because both are driven by a third variable, or by pure coincidence — especially with small samples or many tested pairs. Correlation identifies candidates for investigation; causation requires experimental evidence.

Related Pearson correlation, Spearman correlation, covariance, linear regression, statistical significance