There are several techniques for testing application software for quality assurance. One possible approach is to identify the most critical features of software, and place more emphasis on those, while putting less emphasis on the not-so-critical features. Another approach is to develop a statistical procedure for identifying which test cases to perform, without necessarily identifying which ones are critical, and which are not. Such random selection leads to a more representative sample of the overall features of particular software, and defects identified are presumed to be proportionate with the proportion of defects that would have been identified based on complete exhaustive testing. A third approach is to use the technique of Orthogonal Defect Classification (ODC). This technique is used to develop a model that breaks down all the features of the AUT into categories, then with the aid of standard orthogonal arrays, selects which test cases to run.
Keywords: quality assurance, software testing, regression testing, acceptance sampling