Filters help determine whether the reaction of the system under test to a specific action is suspicious.

Filters are used to avoid that a specific action is executed.

Filters can be used to avoid that a specific action that would unwantedly stop the test session (e.g., pressing the ‘close’ button) is executed. To be able to check whether the software under test actually would close the application (as expected) when executing a specific action, related filters should only be activated after that specific action has been tested once.

Filters can be used to avoid that a specific action, of which it is already known that it leads to a failure, is executed again. Thus, adding filters helps avoid that redundant tests leading to the same failure are executed repeatedly.