Business & Growth
A/B Testing
Comparing two alternatives by assigning eligible experimental units to different versions and measuring a predefined outcome. Random assignment helps separate the version change from other differences between groups. A larger observed result alone is not proof of a reliable improvement; sample size, uncertainty, exclusions and the testing plan matter.
Reviewed
Sources
Free complete lesson
This definition and the complete practical lesson are free. The interactive reader loads the lesson examples without a membership.