Modern software represents a huge and complex space for testing which makes it difficult to find all the important bugs. A large and experienced testing team cannot cover all the issues which can happen to a product in real life, that’s why QA specialists have to use different tricks in their work, and repeating tests is one of them.
A repeated test means a test that includes some already known elements to be tested in other tests. That is, repeating the test suggests repeating one of the previous test’s aspects.
Present day testing includes a powerful diversity which is important, but even though arguments against repetition are valid in most time, there are some specific reasons why repeating a test is even important in some certain situations.
Here are the technical reasons.
- If you want to discover that point where some problem or behavior appeared, verify the fix by re-running your test or repeating it on several earlier builds.
- You need to repeat the test if there’s a risk that one correct test run won’t guarantee the bug discovery due to the introduction of important variables you cannot control.
- In case you suspect the test was run in a wrong way the other time, have several testers run it keeping the same instructions to check the results.
- You also need repetition if you’ve changed some part in the test while its another part is constant.
- The repetition is essential when the repeated tests are part of some performance standard which gets value from comparison with other executions of these exact tests.
The rest are business reasons.
- If your test has value and is rather inexpensive compared to new different tests, go for it again.
- If the problem to be discovered by this test has all chances to be much more important than those detected by some other tests, you ought to re-run it.
- You may also want to repeat the tests if they seem the only tests worth executing.
- There are cases when you are just forced to execute the same tests according to contract or regulation.
- Finally, it may be needed to re-run the tests if this is not for bug detection, but for some other reasons like demo, training purposes, achieving a certain system state, etc.
Mind the given arguments and think of repeating tests when it’s timely.