Sterling and convert.com collaboration: Confounding Variables

The new convert.com glossary is out and David Dias explains a real case about confounding variables.

‍

Confounding variables are external factors that distort the relationship between an experiment and its outcome, creating false causality. Even statistically significant results can be misleading if other changes—like marketing campaigns or traffic shifts—impact performance at the same time. This article explains why confounders occur, how they bias A/B tests, and what teams can do to reduce their impact through better experiment design, segmentation, and validation.

‍

Here's the article: https://www.convert.com/glossary/confounding-variables/
‍

Here's the glossary: https://www.convert.com/glossary/

‍