Software testing is an extremely creative and intellectually challenging task. When testing follows the principles given below, the creative element of test design and execution rivals any of the preceding software development steps.
1. Testing shows the presence of bugs
Testing an application can only reveal that one or more defects exist in the application, however, testing alone cannot prove that the application is error free. Therefore, it is important to design test cases which find as many defects as possible.
2. Exhaustive testing in impossible
Unless the application under test (UAT) has a very simple logical structure and limited input, it is not possible to test all possible combinations of data and scenarios. For this reason, risk and priorities are used to concentrate on the most important aspects to test.
3. Early testing
The sooner we start the testing activities the better we can utilize the available time. As soon as the initial products, such the requirement or design documents are available, we can start testing. It is common for the testing phase to get squeezed at the end of the development lifecycle, i.e. when development has finished, so by starting testing early, we can prepare testing for each level of the development lifecycle.
Another important point about early testing is that when defects are found earlier in the lifecycle, they are much easier and cheaper to fix. It is much cheaper to change an incorrect requirement than having to change a functionality in a large system that is not working as requested or as designed!
4. Defect clustering
During testing, it can be observed that most of the reported defects are related to small number of modules within a system. i.e. small number of modules contain most of the defects in the system. This is the application of the Pareto Principle to software testing: approximately 80% of the problems are found in 20% of the modules.
5. The pesticide paradox
If you keep running the same set of tests over and over again, chances are no more new defects will be discovered by those test cases. Because as the system evolves, many of the previously reported defects will have been fixed and the old test cases do not apply anymore. Anytime a fault is fixed or a new functionality added, we need to do regression testing to make sure the new changed software has not broken any other part of the software. However, those regression test cases also need to change to reflect the changes made in the software to be applicable and hopefully fine new defects.
6. Testing is context dependent
Different methodologies, techniques and types of testing is related to the type and nature of the application. For example, a software application in a medical device needs more testing than a games software. More importantly a medical device software requires risk based testing, be compliant with medical industry regulators and possibly specific test design techniques. By the same token, a very popular website, needs to go through rigorous performance testing as well as functionality testing to make sure the performance is not affected by the load on the servers.
7. Absence of errors fallacy
Just because testing didn’t find any defects in the software, it doesn’t mean that the software is ready to be shipped. Were the executed tests really designed to catch the most defects? or where they designed to see if the software matched the user’s requirements? There are many other factors to be considered before making a decision to ship the software.
Other principles to note are:
- Testing must be done by an independent party.
Testing should not be performed by the person or team that developed the software since they tend to defend the correctness of the program.
- Assign best personnel to the task.
Because testing requires high creativity and responsibility only the best personnel must be assigned to design, implement, and analyze test cases, test data and test results.
- Test for invalid and unexpected input conditions as well as valid conditions.
The program should generate correct messages when an invalid test is encountered and should generate correct results when the test is valid.
- Keep software static during test.
The program must not be modified during the implementation of the set of designed test cases.
- Provide expected test results if possible.
A necessary part of test documentation is the specification of expected results, even if providing such results is impractical.