What SQA Resource are you searching for?

Data-Driven Testing

• Posted by George Nunez on May 17th, 2010 filed in SQA - Terms
• 1 Comment »

Data-driven testing is a term used to define software testing where one test script is used against several iterations of variable data; thereby, testing many possible conditions with one test case.   This re-use reduces maintenance and improves test coverage.  The data values are typically stored in one or more central repository: data sources (i.e. csv or Excel files) or databases.  Anything that has a potential to change or is variable (i.e. environment, end points, test data and locations, etc), is separated out from the test logic (scripts) and moved into an external repository.  As the need arises or new data becomes available,  the repository can be updated and the same test logic (script) can be re-used and executed against multiple data scenarios.

A best-use of this approach is to discover live test data that has caused problems in a production environment.  This “troublesome” test data can be used (after cleansing first of course) by the developers for Test Driven Development, and by the QA department for regression tests to assure that the “troublesome” data does not once again cause a problem in production.