From a psychological point of view, the objective of testing should be executing a software system with the intent of finding defects. Understanding the motivation of that claim can make the difference in the success of our tests.

Human beings tend to be goal oriented. If we carry out tests to demonstrate that a program has no errors, we will tend to implement tests selecting test data with a low probability of causing program failures. On the other hand, if the objective is to demonstrate that a program has errors, we will increase the probability of finding them, adding more value to the program than the former approach. For that reason, testing is often considered as a destructive process, since testers are supposed to prove that the SUT has errors.

Moreover, trying to demonstrate that errors are present in the software is a goal feasible, while trying to demonstrate their absence, as explained before, it is impossible. Again, psychology studies tell us that people perform poorly when they know that a task is infeasible.