This allows changing individual tests in the future without having to adjust the expected output of all other tests.