Many software systems are required to process huge combinations of input data, each of which deserves to be tested. Since we rarely have time to create and execute test cases for all these combinations, the fundamental problem in testing is how to choose a reasonably-sized subset that will find a large percentage of the defects but that can be performed within the limited time and budget available.
Pair-wise testing is the most effective test design technique to deal with this problem but, unfortunately, is not understood by many testers. The answer is not to attempt to test all the combinations for all the values for all the input variables but to test all pairs of variables. This significantly reduces the number of tests that must be created and run while still finding a large percentage of the defects. Lee shares examples of the effectiveness of pair-wise testing and demonstrates the technique through the use of orthogonal arrays, James Bach’s all-pairs algorithm, and Microsoft’s PICT tool - and gives you a good introduction to this GREAT technique.
