Glossary

Minimum Detectable Effect

Minimum Detectable Effect

Minimum Detectable Effect

Minimum Detectable Effect (MDE) is the smallest improvement estimate you want the statistical experiment to detect. MDE can be used to understand the probable relationship between cost and likely value, and effort and impact for your experiment. By calculating MDE, you can prioritize experiments according to expected ROI.

Share


Related Links

AB testing is a data-driven approach to compare two different…