I’m implementing Dan’s grading program, with some minor variations. I’m finding it to be almost all positive:
- Even after 5 years, I’d still have grading guilt. That has gone away, because I know that my kids are going to have a chance to prove themselves again.
- Stretching the assessment out over weeks, rather than on one day, allows me to be more responsive with my teaching.
- I’m not stressed about makeup tests.
- The kids aren’t seeing the grading as punitive. They also don’t give up hope (for the most part) after a bad grade.
- My grading is now focused on perfection: no more 90% scores for someone who got the right idea, but had a minor math mistake. I want you to know it, and know it reliably. I think this, as much as anything, will pay off come the state tests.
I have a couple of modifications:
- I keep the two highest test scores. A B becomes the average of 4 & 3. A D is the average of 3 & 2.
- I miss the authentic assessment, so in my version the tests are a qualifying event for a “What I Learned About …” assignment (e.g. What I Learned About Integer Operations). This is worth an additional 2 points, for a total of 10 for each concept. Failing to do the assignment, or blowing it completely, will drop one of the test scores by one point.
- I’m not counting class/homework as part of the letter grade. I still collect it all in a separate portfolio, and it becomes part of their work habits grade. This may have to change.
- I have no final.
I think it keeps the major features, though: it makes students responsible for their own learning, it differentiates, it focuses on understanding the concepts, rather than just doing the work.
There are also some minor drawbacks:
- It’s good for individual concepts – it doesn’t do as well for combining them. Our books certainly seem to think that’s not an issue, but in my experience you can take a kid who’s shown proficiency in two different areas, and have them be completely befuddled when they get a problem that involves those two areas together. I’m addressing this by rolling in old concepts into the newer ones for the advanced version of the questions.
- Kids catching up don’t get a fair shot at the easier question. I suppose I could have two levels of questions, but that’s way more work for me. If they come in for tutoring, of course, I can tailor the question.
- The percent scale comes out just a bit weird.
- I’m still not entirely comfortable assessing with effectively two questions. The WILA should address some of that, but it still feels akward.