Regression Test

Go to Terms and Definitions

A Regression Test is a Test performed after a Change to determine whether the Change adversely affected previously verified behavior.

Regression Tests may be needed after:

• A defect correction
• A software enhancement
• A Configuration Change
• A data-model or schema Change
• A dependency Change
• Replacement of a software or hardware component
• Modification of a Test Asset
• A Change to an external service or operational environment

The applicable Regression Tests may be selected using the Maintained Item, its identified dependencies, expected Change effects, applicable Requirements, Maintainability Criteria, and conditions of the applicable Change Authorization.

test performed after a change to determine whether previously verified behavior has been adversely affected

DIDO Reference Architecture, Regression Testing, adapted

ISO/IEC/IEEE 29119-1:2022, Software and systems engineering — Software testing — Part 1: General concepts, supporting source

Regression Testing is the activity of selecting and performing one or more Regression Tests. A Regression Test is an individual Test used in that activity.

A Regression Test examines behavior that was verified before the Change. Verification of the intended Change itself remains a separate activity.

An applicable Regression Test may be newly defined or may reuse an existing Test. Change-impact analysis may determine that only a subset of available Regression Tests applies to a particular Change.

A successful Regression Test indicates that the examined previously verified behavior was not adversely affected under the Test conditions. It does not establish that no other regression exists.

A Change replaces a Test Executable dependency. A previously successful Test is performed again to determine whether the replacement altered the Test Executable’s previously verified behavior. In this use, the repeated Test is a Regression Test.