Confounding variable
A confounding variable is a third factor that influences both the explanatory and outcome variables, creating a misleading association between them that is not causal. It is why correlation does not imply causation.
See it move
Ice-cream sales rise from €3,000 to €18,000 and AC repair calls rise from 6 to 42 in the same months, looking like ice cream breaks air conditioners. The real driver is outdoor temperature, which independently pushes up both ice-cream demand and air-conditioner strain, the confounding variable behind a non-causal correlation.
Check yourself
Across 40 retail branches, those with more customer loyalty-card sign-ups also have higher average transaction values: branches with over 500 sign-ups average €62 per transaction, while branches with under 200 sign-ups average €41 per transaction. A manager concludes that pushing more loyalty-card sign-ups will raise average transaction value. Which statement best identifies why this conclusion may be flawed?