Research can't tell teachers what to do—classrooms are too complex for that ever to be likely. But research can identify "best bets" for improving practice, and the Great Teaching Toolkit Evidence Review is pretty much the state of the art in this regard:
greatteaching.com