Monday, August 1, 2011

Summer Learning, PCMI Edition: Formative Assessment

This is a series of posts that are reflections from the Park City Mathematics Institute Secondary School Teachers Program.

Our middle session every day was called Reflecting on Practice, and it was basically a mini ed-school class. The focus this year was on formative assessment or assessment for learning. These types of classes are not usually my favorite (make a fake assignment! watch a video of someone teaching! talk about your feelings!), but in this case they were exceptionally well planned and executed so I didn't have much time to feel sorry for myself.

Biggest takeaway - there needs to be deliberate feedback, not attached to a numerical grade, built in to classes. Because when there's a number there, kids don't pay attention to anything else. (On the flip side, in the absence of a grade you run into kids not taking the work seriously, so giving feedback on their marginal efforts feels like a waste of time.) At least some of the time, the attitude toward assessment should be less "judgment day" than "a conversation about learning and understanding." I was influenced especially by two articles we read: Classroom Assessment: Minute by Minute, Day by Day and Working Inside the Black Box: Assessment for Learning in the Classroom (which does not appear to be available online.)

These simple ideas lead me to rethink the whole process for "level 1" quizzes - the kids' first stab at a concept on an SBG quiz. I spent way too much time re-designing what a quiz paper should look like:


That's super-helpful for you, right? Sorry. After I try this in class I'll have more to say about it with a nicely-typed up version. But the idea is, there's half a page for the student to do his work, and predict his score. The bottom half of the page is set up for structured teacher, self, and peer feedback. I want the message for level 1 to be "I want to help you figure out what you still don't understand" instead of "Fear my red pen!"

I asked the cherubs to weigh in on Facebook and got mixed responses.


So anyway, the idea for a process will look like this for Level 1 questions:
  • Students take the quiz, and predict a score.
  • I collect the quizzes and write feedback in sentence form, like "right idea but computational errors" or "a more careful and accurate diagram would be helpful"
  • Next day, students give feedback to each other. One idea, so that they are working to understand and not to just get the right thing on the paper without understanding, is to not let them use pens or pencils, but communicate with mini-whiteboards. While they are doing this, I can be assessing/rewarding/publicizing helpful dialog that I hear.
  • Then, students have an opportunity to re-work the problem, or possibly a new but similar problem.
  • Then...what, grade the quality of their feedback to each other? Or never grade this part of the process? I am thinking that if I want the focus to be on the learning and not a numerical grade, I can't give a grade to this part ever.
Obviously there are questions here that will take time to sort out. But this seems like a step in the right direction.