I thought this was an interesting essay from this week's Business Week, "Better Teachers: A Lesson Plan".
Two interesting data points here for K-12 teachers. Having a master's degree gets you a raise, but doesn't improve your teaching performance seems to make sense to me. Giving raises based on seniority only doesn't really work, because people stop getting better results just a few years into their careers. Once you get up the seniority ladder the pay is so good that you can't change jobs because of the opportunity loss.
There is an idea that the best teachers have the best test scores (I think they are talking about SAT type tests). But that makes sense when you look at general results without any other criteria.
An interesting read, check it out.