This idea was presented at the last NCCTM conference. I didn’t go to the session, but one of my colleagues did. He has agreed to collude with me on my idea for interactive notebooks next semester if I will collude with him on this project. I actually really like the idea, and I’m thinking about implementing it in all three of my preps next semester (not just the geometry class we agreed to cooperate on.) The idea is just a rough sketch at the moment, I would appreciate any feedback you had to offer!

So it works like this. Each test is divided into three parts, roughly equivalent to developing, proficient, and advanced. Students know it is divided into three parts and the three parts are clearly labeled (although maybe using less confrontational terms than what I used above, I don’t know.)

The idea is that a student can select to answer only the “easier” questions and know they got an 80, but might feel bolder to try some of the more difficult questions if they knew they had that 80 in their pocket. A more advanced student could skip the easier questions and do only the more advanced sections to demonstrate their knowledge.

See sample below.
http://viewer.docstoc.com/
DIff Test Example

This is very much just an early attempt at this. I took most of the problems pretty directly out of Prentice Hall’s Geometry book (the one with the creepy bee on it) and I”m not totally convinced I’ve broken them down correctly into the sections. I need to spend more time with my state standards and unpacking them into what is absolutely essential (that would all go in section I) what is important for students who I expect to be successful in a course like precalculus (that would all go in section II) and some tasks that are really going to be a stretch (section III.) I would expect the sections to typically get shorter as they go along.

One thing I want to avoid is varying the difficulty level by just ratcheting up the algebra. That is harder, and it has it’s place, but I would like the differentiation to happen, primarily, on a more conceptual level. Of course, those tasks are much more difficult to write!

Problems I think of immediately:

1. Difficult to design.
2. What if there’s a disconnect between the sections, i.e. a student solves advanced problems well but can’t solve the easier problems (I suspect I can avoid this if the test is designed properly, but still.)
3. Does solving a more advanced task like an application really demonstrate the ability to answer straightforward knowledge questions?

There are also lots of other problems I haven’t thought of, and I’m hoping you will, but I’m intrigued enough by the idea to try it.

What do you think? Please feel free to leave any feedback or ideas below, or tweet me @seetur.