Interesting. Forced-choice is ok if the question is completely clear. In many cases you have to assume information that makes the difference in which choice you pick, I find. Unless that in itself is what is being tested (do you implicitly supply optimistic or pessimistic information? for example). The problem with these questionnaires is that they get a lot of information from people who don't take it seriously so they don't learn anything either. In other words, it's my preference for clarity that is making my responses uninformative for the survey-giver.