function changes the behavior of the test so that a failure indicates
success and a pass indicates a failure.
Changing the expectation of the test to a failure is useful for test
cases that exercise a known bug: while the bug is live, the test case
will pass with a message describing the error encountered; however,
when the bug is fixed, the test will start failing indicating that it
needs to be revised.
can only be called once during the execution of a test case and its
effects cannot be reversed.