Pesticide Paradox in Software Testing
In this content, I will explain what the Pesticide Paradox is, which is one of the 7 principles of software testing.
What is the Pesticide Paradox aka Antibiotic Resistance
First of all, I will explain this situation biologically.
Basically, the decrease in the effect of drugs at a certain dose after repeated use at the same dose or the necessity of using them in higher doses to create the same effect is defined as the development of resistance to the effect of the drug. The same applies to drugs (antibiotics, antineoplastics) whose mechanism of action is to kill or suppress pathogens that cause disease in the body.
The concept of antibiotic resistance is often used as an example in software testing to indicate that when you run your automated regression tests multiple times, they stop being effective at catching bugs.
If the same tests are repeated over and over again, eventually these tests no longer find any new defects. To detect new defects, existing tests and test data may need changing, and new tests may need to be written. (Tests are no longer effective at finding defects, just as pesticides are no longer effective at killing insects after a while.) In some cases, such as automated regression testing, the pesticide paradox has a beneficial outcome, which is the relatively low number of regression defects.
In such a case, it will be better to write new test cases or update test cases that will cover newly added features to the product and parts of the product that was not covered in the regression tests before.
