Is there a pedagogical value to being confusing?

When a teacher makes a lot of mistakes, and you notice them and shoot up your hand to correct them, are you learning more or less than if the teacher was right all the time?

What if it isn't you shooting up your hand, but someone in the same class as you?

Since my degree is practical and science-oriented, most of what I learn is self-consistent rule systems (math, thermodynamics, physics). So, it is pretty easy to notice when a mistake occurs during teaching, and they are quickly rectified.

My opinion is that small mistakes make the transmission of information more clear, as it encourages you to correct information. So, should a teacher do it on purpose?

I am a T.A. for first-year calculus, and mostly I teach how to work through problems. When I am designing how to teach a problem, I build in common mistakes that people make and, of course, a few of my own by accident. Then, the process of doing a problem is much more interactive, and I think it makes for better students.

Comments?