Given the comments here I think the problem demonstrated is that multi-guess exams are inherently poor in determining a candidates true understanding of a subject.
This is due in part to poor questions and/or answers, to the nature of examinations that force a person to use another persons words, and an examination system that allows for 'statistical' answers such as BPF suggests.
Unfortunately it appears it's not a UK-only phenomenon, as I live some 12,000 miles away. I had reason to expect a 100% pass in CPL Tech, given my 'other' live(s). Answering a question with its fundamentally correct (or least incorrect!) answer wasn't the right thing to do when explaining how a moving-coil meter works apparently. They were clearly looking for something, er, 'higher level' shall we say. CAA (or at least their designated examiners in this case)) had no reason to know I could probably have written a small book on the subject - without needing to research - and so marked me wrong when in fact I was definately correct. Actually to be more accurate; in answering as I did I was using a somewhat deeper level knowledge of the subject than was expected in most candiates. As such there were two 'correct' answers, one explaining the fundamental mechanism, one the mechanism itself more or less. To me, in explaining the operation of something, you get the fundamentals correct first before moving to the next level, hence why I answered that way and probably why I was 'wrong', the examiner may not even have realised the flaw - confirming P.Pilcher's thesis.
Incidentaly this wasn't the only case of a 'bad' Q & A(s) in this exam but it's the only one I can specifically remember at this point. There were also some others that I really did wonder why a pilot should need to know - an engineer perhaps but not someone driving an aircraft. Once again an opportunity lost to examine something that mattered in that particular context...
P.