A study of the effectiveness of regression test case selection and scheduling strategies in a system test environment

Includes bibliographical references (pages [71]-70)Today’s world has become increasingly dependent on software. The complex nature of our world has caused software to become more and more complex. This increase in use and complexity has made it quite a challenge to keep software defect rates low and customer satisfaction high. Software reliability engineering concepts are designed to address this challenge. These concepts can be applied to areas of testing to remove defects and improve customer satisfaction. Much literature has been written on applying software reliability engineering concepts to testing at the code level, with a lack of application in the system test area. System testing plays an important role in the software development process. Unit and integration testing only test the code to make sure it works as written. This is not enough. System testing allows the system to be tested based on its expected behavior, which is very important to the customer. After reviewing software testing literature on Modified Code Coverage Using Data Flow, Statistical Based Testing, and Cost Models and Analysis, it was decided that more work is needed in regression test selection and scheduling in a system test environment. In this thesis, a simple regression test case selection strategy is proposed, along with recommended scheduling strategies based on levels of Customer Risk, Failure Tendency, and Common Setup.M.S. (Master of Science