Testing components of web browsers and other graphical interface software can be extremely expensive because of the need for human review of screen appearance and interactive behavior. Combinatorial testing has been advocated as a method that provides strong fault detection with a small number of tests, although some authors have disputed its effectiveness. This paper compares the effectiveness of combinatorial test methods with exhaustive testing of discretized inputs for the Document Object Model Events standard. More than 36,000 tests all possible combinations of equivalence class values were reduced by more than a factor of 20 with an equivalent level of fault detection, suggesting that combinatorial testing is a cost-effective method of assurance for web-based interactive software.
Citation: Software Quality Professional
Pub Type: Journals
combinatorial testing, covering arrays, interoperability testing