Another important topic that we have discussed in CS-443, or Software Quality Assurance and Testing, is known as Pairwise Testing. Pairwise Testing is yet another form of testing, but this type is a little bit different than the rest. Pairwise Testing, sometimes known as all-pairs testing, tests each pair of input parameters in order to make sure that the functions in the system run correctly no matter what the input is, guaranteeing that it will run for every combination. Pairwise Testing is known as a Permutations and Combinations (P&C) based software testing technique. A blog that I found to be really helpful in explaining Pairwise Testing is known as Pairwise Testing | What It Is, When & How to Perform by Kiruthika D. In the blog, she gives an example that helped me understand more. She states “Let’s say you have an application that allows users to enter two numbers, and the application will output the sum of the two numbers. You can use pairwise testing to test all possible combinations of two numbers, such as (1, 2), (2, 3), (3, 4), (4, 5), etc. By testing all the combinations of two numbers, you can be sure that the application is working correctly and will not fail when given different numbers.” This shows that you don’t actually test every single combination, but you test every single input with another input. This way, it makes sure that all of the inputs work instead of testing a potentially infinite amount of combinations.
The actual purpose/use of Pairwise Testing is exactly what I previously stated. It is used to make sure that all combinations of inputs are possible, but you don’t need to test every single combination. It can be extremely helpful as it reduces the amount of time it takes to test the program as well as the amount of effort. While Pairwise testing is a great testing technique, you obviously can’t use it all the time as it involves pairs. As for when to use it, Kiruthika states “Pairwise testing is helpful when testing complex systems that have multiple input parameters and multiple possible values for each parameter. It can significantly reduce the number of test cases that need to be created while ensuring that all possible discrete combinations of parameters are tested. This can help reduce test case creation time and cost and improve the software’s overall quality. Pairwise testing is not appropriate for all types of software testing. As we discussed, it is most effective for systems with multiple parameters and multiple possible values for each parameter. If a system has only a few parameters and a small number of possible values for each parameter, pairwise testing may be unnecessary. Pairwise testing, also, will not be useful if the values of inputs are inappropriate.” Essentially she is saying that Pairwise testing is used for functions that have multiple parameters with multiple values, and the order of parameters doesn’t matter. On top of that, depending on the type of parameter, the technique might not work either. While I personally don’t see myself using this technique in the future, I think that it has the opportunity to be very useful in certain situations, so I’m glad that I was able to understand it more in case I ever need to use it.
One pixel at a time 