Sha256: c0ae23d41dddfaa879fb5989fec0dc68f6f228c323c551d01f4d36cb9829e0a8
Contents?: true
Size: 1.64 KB
Versions: 7
Compression:
Stored size: 1.64 KB
Contents
# Verdict Concepts Understanding the following Verdict concepts can help you implement experiments and discuss them with peers. - [Experiment](#experiment) - [Qualifiers](#qualifiers) - [Subject](#subject) - [Group](#group) - [Assignment](#assignment) ### Experiment `Experiment`s are defined in `app/experiments` and encapsulate the details of the test including metadata such as the name, description or owners, as well as any [qualifiers](#qualifiers) or the [groups](#groups) available in the test. ### Qualifiers Qualifiers are methods that take two parameters, `subject` and `context`. They return `true` if the `subject` can take this experiment, or `false` if the `subject` is not eligible. Experiments can have multiple qualifiers which are evaluated in succession. If a qualifier returns `false`, no further qualifiers are checked. Examples of qualifiers might be subjects that: - Originate from certain browsers or devices - Have a specific `Accept-Language` header - Have added items to their cart. ### Subject A subject is the thing that will be [qualified](#qualifiers), and if successfull, [assigned](#assignment) to a particular [group](#group). For example, a subject might represent a user. ### Group Each [experiment](#experiment) will have two or more groups, usually a control group and a test group. Qualified [subjects](#subject) are assigned to one of the groups in the experiment. ### Assignment An assignment occurs when a qualified [subject](#subject) is placed into a [group](#group). Assignments are persisted so repeat visits by the subject skip [qualification](#qualifiers) and re-assignment and simply returns the existing assignment.
Version data entries
7 entries across 7 versions & 1 rubygems