Refers to the behaviour of a software application or component as it is recorded during a test. Frequently, this actual state does not conform to the specifications on the software.
When the actual state is not as expected or required, work is undertaken to attain the target state.
Acceptance testing is performed on software by the customer following delivery. The aim of acceptance testing is to help the customer who commissioned the software to decide if all the requirements of the software have been fully implemented and to see whether to accept the software or not.
This is testing by future customers and users - or independent teams of testers - in a simulated environment or environment that has been specially designed for testing.
Alpha testing can be understood as "internal acceptance testing".
This is a tool used to check formalised testing objects (e.g. source code for software applications) for certain attributes and quality hallmarks.
Any deviation from the requirements formulated at the design stage, from documentation or standards based on clients'/users' experience.
Anomalies may be discovered through testing, analysis, reviews or through use of the software.
This is the independent evaluation of software carried out by an external organisation checking conformity to guidelines, standards, specifications and procedures that have been developed based on objective criteria.
Documents relating to the form, functions and contents of the product to be produced are also audited, along with written process for production and the specifications showing conformity with standards and guidelines or make conformity measurable.
Cf. audit in Wikipedia.