Still, I have seen a number of times written questions that were ambiguous enough to deserve two conflicting or even opposite answers, depending on the level at which you wanted to answer. When you are supposed to give a written answer, you can explain that there are different levels of answers (and hope that the corrector will understand what you mean, I have at least once been in a formal university exam situation where the corrector had no understanding of what I explained to be ambiguous, and I was able to check with the top professor that I was right making a two-sided answer and that the corrector simply did not understand the deepest thoughts that I was expressing, but that's rather uncommon).
And then, there is the multiple-choice type of tests, where you usually have no chance to explain your choice and are at a great difficulty to figure out whether the authority (college or university professor, job interviewer, whatever) is asking for the simple immediate answer, or for a more clever response. I really hate this type of situation. Ambiguity should really be banned from such tests.
| [reply] |
Some times the people writing test questions make mistakes and the reviewers also miss the mistakes. In this case, both the writer and reviewer may have been people whose primary programming language is not Perl, or at least the reviewer. These people were likely under pressure, so "if it looks right, it's good enough".
Unfortunately, I get far too many specifications that merely "look right".
| [reply] |