Correlation Does Not Mean Causation
Two metrics moving together does not mean one drives the other. In UX research, a spike in clicks after a redesign does not prove the redesign caused it. The footgun is treating every coincidence as proof your change worked.
Why it exists
Humans are pattern-seeking animals. When two things happen together, our brains automatically weave a story of cause and effect because that narrative feels safer and more predictable than randomness. The concept exists as a guardrail against that instinct, forcing evidence before we act on a relationship between variables.
The mental model
Think of correlation as a neighborhood watch report and causation as a conviction in court. The witness saw two people near the scene at the same time, but that alone does not prove either person committed the crime. You need motive, opportunity, and a mechanism before you can sentence someone; likewise, you need a controlled mechanism before you can claim one metric drives another.
How it works
Correlation measures statistical association, usually expressed as a coefficient between negative one and positive one. Causation requires three harder conditions: the cause must precede the effect, the two must covary, and every alternative explanation must be ruled out. In practice, that means running randomized controlled experiments, using instrumental variables, or applying causal inference frameworks rather than simply eyeballing a dashboard trend.
When to use it
Use this distinction when interpreting analytics, A/B test results, survey cross-tabs, or ethnographic observations. Any time a stakeholder says we changed X and Y went up, so X caused Y, pause and map whether a hidden variable Z could explain both. It is especially vital during quarterly reviews when teams mine historical data for success stories without experimental controls.
When not to use it
Do not invoke the phrase to dismiss every observed relationship as meaningless. Correlation is still a powerful signal for generating hypotheses and deciding where to invest experimental bandwidth. The error is stopping at correlation, not starting there.
One canonical example
Ice cream sales and drowning deaths both spike in summer. Banning ice cream would not reduce drowning; the hidden variable is hot weather. In UX, a similar trap occurs when mobile app crashes and five-star reviews both rise after a release. The crashes did not cause the praise; a simultaneous marketing push attracted new users who rated the app before hitting the bug.
Interview question
When metric Y rises after a design change to X, why should a researcher pause before claiming the change caused the improvement?
- a.Because an unmeasured variable Z could be driving both the change and the metricCorrect
- b.Because A/B tests are the only valid way to measure any user behavior
- c.Because positive trends in dashboards often reverse themselves the following week
- d.Because correlation coefficients above 0.5 are usually statistical flukes
Why? this is the answer
The card explicitly warns that a hidden variable Z might explain both the redesign and the metric shift, making coincidence look like causation. The tempting distractor about A/B tests is wrong because the card notes correlation is still a valuable hypothesis signal and lists several causal inference methods, not only experiments.
Just read this? Test yourself on what you have been reading.
Read the original → en.wikipedia.org
- #ux research
- #data literacy
- #experiments
- #causality
- #metrics
Put your scrolling time to good use
Learn one idea, try a quiz and save useful cards for revision. Tezvyn makes it easy to learn and stay current in your tech field, a few minutes at a time.
The iPhone app is on the way
We are building it. Until it lands, nothing here is held back from you: every interview card, your saved cards, streaks and the job board all work in Safari, plus hundreds of free practice quizzes of thirty questions each. Sign in and it all carries over to the app the day it arrives.
Want it as an icon? Tap Share at the bottom of Safari, then Add to Home Screen. It opens full screen and the cards you have read stay available offline.
We are hiring for this. Open roles that interview on ux research — each one lists the topics its interview covers.
See open roles