Software testing sizing in incremental development: A case study

The paper presents a case study on software testing sizing of 20 project releases delivered by a technical team through incremental development. Different sizing metrics including KLOC, development effort, test case number and function sizes are reviewed and evaluated. The results show that KLOC doesn´t behave well, development effort surprisingly act as a good measure, test case number is an excellent testing sizing method, and existing function size ways are not suitable in our case. In the meanwhile, several sizing evaluation models are proposed during the discussion.