A/B test
An A/B test compares variants assigned to groups to estimate their effect on a defined outcome
What it means, how it is used and where it comes up, with its related concepts.
What it is and how it works
A rigorous design sets out the hypothesis, metric, assignment and analysis criteria before interpreting results. Sample size, duration and uncertainty must be considered. A difference observed over a few days does not in itself prove that one variant is better.
Topics
Variants and linked concepts
- MVT. A multivariate test combines changes to several elements to study their effects. It requires a design and sample suited to its greater complexity.