First, teachers were asked if they valued creativity and enjoyed working with creative students, and they overwhelmingly answered "yes".
Interestingly, there was a significant negative correlation between the degree of creativity of the student and his favorable rating by the teachers. This means that the most creative students were the least favorite of the teachers, across the entire sample surveyed. Additionally, the students that were rated as favorites of the teachers possessed traits that would seem counter-productive to creative behavior, such as conformity and unquestioning acceptance of authority. On the other hand, these are behaviors that fit well in a classroom setting. Even back in 1975, Feldhusen and Treffinger reported that 96% of teachers felt that creativity should be promoted in the classroom. However, when asked which students they actually liked to teach, they chose the students that were more compliant.
Teachers say they want creativity, but that is not the behavior that is rewarded. In this study (as well as in many others), they found that there is a discrepancy between what teachers, and schools in general, say they value and desire, and what behaviors they actually reward and encourage. Teachers don't want the student who is always raising their hand and questioning the assignment; they want the student who unquestioningly follows the outline given to them and turns the assignment in on time.
Professors love creative students, as demonstrated by those who extend their work in predictable yet difficult dimensions. The 'creativity' is evidenced merely by embracing their own paradigm shifting work. A truly creative person, like a true risk taker, is seen as a crank in real-time, and indeed on average they are. But sometimes the cranks are right, at which time they are then seen as geniuses.
Like risk-taking, what these creativity lovers mean is, they like positive surprises. Risk-taking of a regular sort that leads to failure is considered just dumb, or worse, full of hubris. Creativity is great if it generates an obviously valuable thing, otherwise, it's annoying, and just inappropriate.
I once read an article that accompanied an exposition on the Capital Asset Pricing Model, and its implication that expected return was positively related to risk (via the 'beta', now an anachronism). They asked a bunch of well-known portfolio managers about their greatest risks, and each one was an unqualified success. This gives one the feeling that risk is related to return, because in these biased samples they are. Kids, and adults, should know that risk-taking usually fails, and so the only key is to know when to stop and when to keep trying, learning along the way.