Correlation is a statistical association: when one variable changes, the other tends to change too, measured for example by Pearson's coefficient from -1 to 1. Causation means a change in one variable actually brings about a change in the other. Data can show correlation; establishing causation needs more.
Why it matters
Confusing the two leads to bad decisions. A third factor (a confounder) may drive both variables, the direction may be reversed, or the link may be coincidence. Randomised experiments such as A/B tests are the strongest way to show cause, since random assignment breaks confounding. Observational methods can help but need stronger assumptions.
Example
Ice cream sales and drowning incidents rise together in summer.
Confounder: hot weather increases both.
Banning ice cream would not reduce drowning.In product analytics, users who use a feature may retain better simply because they were already more engaged.
Common mistakes
- Concluding "X causes Y" from a dashboard trend.
- Ignoring confounders and reverse causation.
- Reading a correlation of 0 as no relationship. It only rules out a linear one.
Related
Get notified when this term updates
Plus occasional practical tips on uptime and monitoring, nothing else. Unsubscribe anytime.
This page is general information written by our editorial team and updated when documentation changes. Servers and providers differ, so test changes safely first. See our editorial policy and disclaimer.