liu.seSearch for publications in DiVA
Change search
ReferencesLink to record
Permanent link

Direct link
Reducing Regression Testing Feedback Cycle Times Through Improved Testing Techniques
Linköping University, Department of Computer and Information Science, Software and Systems. Linköping University, The Institute of Technology.
2014 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

Software is continually and rapidly evolving with constant risk of introducing faults. Software testing has long been used to aid in the detection of faults, and agile development strategies have been driving the use of automated tests and regression testing specifically. As development continues, test suites eventually grow in the number of test cases to the extent that the execution time is extensive. When it has increased to the point that it prevents efficient software engineering, a regression testing technique is required to reduce the feedback cycle times - the times for receiving feedback from tests on changes.

This thesis has investigated regression testing techniques presented in previous research. The focus has been on test case selection techniques - for selecting a subset of all test cases for execution - and test case prioritization techniques - for determining the execution order of test cases. With some evaluation criteria in mind, a safe modification-based selection and prioritization technique was chosen and a proof-of-concept implementation was developed. First, the implemented technique was evaluated for robustness using an example application. Following, a case study was conducted on an existing software development project, where the perceived problems with regression testing were documented by interviewing a software developer. The technique was then integrated with the project's existing regression testing and its efficiency was evaluated.

It was concluded that a regression testing technique is, to some extent, practical to implement, although difficult to verify for complete correctness. Empirical evaluations in the case study showcased reduced feedback cycle times of 60% or more compared to when not using the technique - assuming a uniform distribution of failing test cases. However, it was stated as important to evaluate the efficiency of the technique on a per-project basis.

Place, publisher, year, edition, pages
2014. , 109 p.
National Category
Computer Science
URN: urn:nbn:se:liu:diva-110676ISRN: LIU-IDA/LITH-EX-A--14/053--SEOAI: diva2:748107
External cooperation
Subject / course
Computer and information science at the Institute of Technology
2014-09-05, Donald Knuth, Linköpings universitet, 581 83 Linköping, 13:15 (English)
Available from: 2014-09-19 Created: 2014-09-18 Last updated: 2014-09-19Bibliographically approved

Open Access in DiVA

fulltext(872 kB)393 downloads
File information
File name FULLTEXT01.pdfFile size 872 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Search in DiVA

By author/editor
Lövgren, Viktor
By organisation
Software and SystemsThe Institute of Technology
Computer Science

Search outside of DiVA

GoogleGoogle Scholar
Total: 393 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 2415 hits
ReferencesLink to record
Permanent link

Direct link