In my 9th grade conceptual classes, I've done less and less talking as time's gone on. And I've actually seen improvement in test, quiz, and homework performance.
When I started teaching 9th grade last year, I'd spend maybe 20 minutes per class leading a discussion while students highlighted their text. Certainly this was more effective than when I talked to the class for 40 minutes per class back in 1996, the last time I taught 9th grade.
Later in the year, I handed out "fact sheets" and went over the facts with the class, again leading a brief discussion. These discussions would go for 10 minutes or so a few times per week. I'd also lead an interactive discussion about the previous night's homework sometimes.
This year, I pledged to do NO DISCUSSION, NO LECTURE.
Each class leads off with a three-minute quiz; students grade the quiz, so I dictate the answers to the quiz. But everyone focuses carefully on what I say, because they care very deeply about whether they get points on their quizzes. I make sure to explain the answer to each question first before explicitly stating the answer, so that I maintain attention as long as possible. This isn't lecture; this is going over a quiz.
Once a week or so, I have students grade each others' problem sets to a rubric. I stand in front of the class and dictate the rubric. Students pay very careful attention, because they are responsible for grading a peer's paper correctly. I find that 15 year olds are unabashed about writing total bull honkey for my reading pleasure, but they get quite embarrassed when their friends have to take off points for silly answers. Again, this isn't lecture at all, even though I'm talking. The class hangs on every word because they are immediately accountable for what I am saying.
Beyond grading items and (briefly) answering specific questions brought forward by students, I don't talk to the class. Important information -- definitions, equations, etc. -- is communicated exclusively in writing, via fact sheets. I turn on music and let them get on with an activity of some sort. Usually this activity involves both problem solving and experimentation; sometimes this is test corrections, or collaborative review of previous topics. In any case, the vast majority of the class is spent with students working on their own,* and showing me their work at regular intervals.
*Collaboration is allowed using the five-foot rule -- they can talk to each other as much as they want, but when they are writing anything to be turned in, they must sitting a minimum of five feet from any other student. I'm right there to enforce the rule if necessary.
Whatever we work on, students must show me every single part of every single problem before they move on. If their work is wrong, I explain the issue, and make them redo the work right then and there. If I don't see someone at my desk for more than about five minutes, I have a word with him -- either he's distracted and needs a (figurative) kick in the arse, or he needs to plow through some misunderstanding, which cannot be done by staring at a blank page. I make such a student write something to show me right away. We're modeling how to approach complicated homework problems, so that such a student knows what to do at night when he's flummoxed.
I see a lot of the same mistakes from multiple people. But it's amazing how well information flows through the class. I don't even need to prompt people much of the time -- they explain to each other how to approach problems, or what mistakes not to make.
The point is, the no-talking-by-me approach has produced better performance than ever, and by a good margin. If there's something I'm just dying to explain to the class, I don't say it; instead, I design a quiz question to make my point in the context of grading the quiz. If everyone is screwing something up, I make sure that several of the homework or in-class exercises address the issue.
So is there any point in me talking to the class at all?