Methods of in-class collaborative work: 421

This weekend I attended a workshop given by Kelly O'Shea and Danny Doucette.  They showed us their outstanding approach to lab practicals, which they assign as group tests.  

The discussion in the room at several points turned to balancing the group / individual dynamic in the classroom.  On one hand, physics is a collaborative endeavor. Cooperation and communication are skills which we must teach and assess.  On the other hand, we are teaching result-obsessed teenagers, who default to letting the (perceived) smart kids do all the work, probably while making fun of them behind their back.  

If we're going to encourage, let alone require, cooperative work in physics class, we must incentivize appropriate collaboration.  Remember, incentives can and should take forms other than mere grades.  Although others have found success in assigning a direct grade for the quality of participation in group work, I have not; I find students spend more time gaming the grade than actually collaborating.

My personal approach to encouraging effective collaboration is enforcement of the five foot rule.  As always, my way is not the only way.  Another workshop attendee -- I dearly wish I remember who -- mentioned an extraordinarily clever approach to evaluated group work, one that I'd like to try.

He called it the 421 method.  The laboratory exercise or problem to be solved is presented to the class, and then the class is divided randomly into groups of four.  Then, work proceeds in three stages, with clear time limits assigned to each.  (Yes, stages are numbered strangely.  You'll see.)

Stage 4: Discussion.  Each assigned group of four may discuss the problem together; but they may not write anything down.  No pen, no whiteboard, nothing.

Stage 2: Representation.  The groups are subdivided into pairs.  Each pair may communicate orally and using a whiteboard.  However, they may only write representations - no numbers or words.  This means they can use equations, free-body diagrams, energy bar charts, etc.  

Stage 1: Solution.  Now students separate to use pen and paper.  They are assigned to write a thorough response, including representations, numbers and words.  This is turned in for evaluation.

People in the workshop asked, do you evaluate the group work?  Thing is, by evaluating the individual solution in this case, you are evaluating the group work!  If the students were effectively working together, communicating clearly with one another, pooling their talents well, then necessarily the product should be that each individual student can communicate by him or her self.  The student who held back from the group, who didn't actively participate, won't have the benefit of the four folks working together.  

This method does require that you assign lab exercises or problems that are beyond the simplistic.  AP-level questions are good here, or a simpler version of Kelly's group test-style lab practicals could work in this style.  If the whole approach to the problem is immediately obvious to more than one or two students in your class, there's little incentive for high level students to converse in stages 1 or 2.

I'll need to experiment to figure out the precise level of difficulty for this approach.  Nevertheless, I love the idea.  Let me know if/how it works for you.