If the value for FieldA come from some sort of input at runtime, then yes, you should write a failure test and verify that the result you get is the one you wanted (error code and/or exception).

If the value should be filled by a calling method at design time, then I don't think that you should write a test for that. The test for the calling method (the one that forgot to initialize FieldA) should trap this development error.

HTH

>WHile creating test, should we write test code for pass and also fail scenario?
>
>For example, If my class must not accept ZERO value for fieldA. I write test to check is the rule passed. Should I write test for scenario the test should fail?