Can a questionnaire to account holders on both ends (for example, to an entire office) reveal a spoofing or MIM attack?

...and if so what would be on it

Since there would be some differences between what account holders actually sent and what is received, it seems possible to detect a MIM or Spoofing attack with a questionnaire to both/all parties. What kinds of questions would you put on that questionnaire?

http://en.wikipedia.org/wiki/Questionnaire