/eɪ bi ˈtɛstɪŋ/
A randomized experiment designed to gauge user preference, typically in web design.
In a typical A/B test, two versions of something are created and assigned randomly to users. Version preference is measured implicitly using metrics such as click rate. This method could be used to analyze user purchasing decisions between two versions of a website where specific product information is included or omitted, which can support inclusion of navigational and display attributes.
Display Attribute, Navigation Attribute