[Updated 1/21/11 to reflect that I have decided to stop giving zeros, after reading this.]
[Updated again 7/7/11 to change how many past assessments I keep. See below.]
I’m hoping that sometime before June I’ll be able to write about the first third of the year. I am happy to announce I’ve changed my tagline – the old one was “Success is the ability to go from failure to failure without loss of enthusiasm. -Churchill.” I still have a million things to learn but things feel like they’re beginning to inch in the right direction.
A couple people have asked me about standards-based grading. If you already know SBG there’s nothing new here. But here’s an introduction.
Start by browsing these three introductions to what-and-why:
Here’s the details of what I’m doing so far. My gradebook for each course lists the skills we’ve studied so far. For each student I keep two scores per skill – the student’s best score ever, and her most recent score on that skill. So when an advanced algebra student logs on to snapgrades and looks at my class, here are the first few lines of what she sees:
Scores represent the student’s stage of development on that skill:
0. – blank response (counts as a 40% in grade computation. What? 40% for nothing?!? Read why I decided to stop giving zeros.)
1. – addresses the problem but no evidence of understanding what the question is asking (counts as a 50%)
2. – understands the question and looking for strategies to solve it (counts as a 60%)
3. – applying valid strategies (counts as an 80%)
4 – mastery: can explain and/or verify a correct solution (counts as a 100%)
Their grade is the average across all recorded scores on the usual scale: >90% A, >80% B etc.
By keeping the “best” and “most recent” score, students get some lasting credit for past demonstrations, but they are still expected to retain skills through the year. Keeping the best score also gives protection against a bad day. They can also request to reassess on a skill in order to improve their scores on that skill.
For this to work, homework and assessment must be cumulative. The curriculum I’m heavily drawing on (college preparatory mathematics) recommends at least 60-70% of each quiz or test be on previous material.
I allow students to demonstrate their skill in any way they’d like. Usually this is quizzes & tests. Or, they come in after class and I put a problem on the board; they solve and we discuss then I assess. I also will post scores for an explanation to the class from the board, or for a particularly lucid poster, or if I catch a student explaining something really well to someone else.
Update 7/7/11 – This year I don’t think I will keep the “best” score; only the “most recent”. My sense last year was that grades were too high for what the students had actually retained by the end of the year. (Why they didn’t retain is partly on me and partly on them, but anyway is a different story.) Deep enough understanding to cause retention is the goal and I think many students got complacent because as they stopped practicing and started to get confused between past concepts, their grades didn’t drop that fast, so they weren’t motivated to jump back on it.