I’ve thought a great deal about assessment this year. The last school I worked at had a culture of grading a great deal of student work, in large part as an incentive to get students to work hard. My new school had a history of using standards-based grading, and offered me a great deal more freedom in how I wanted to grade and assess. I wrote a bit about my thoughts over the summer here.
I’ve also been influenced by Michael Pershan’s take on grading — that we often conflate three things. There’s evaluation — telling a kid what they know and don’t know, and putting a number on that. There’s feedback, which provides an opportunity to move learning forward. Then there are incentives, that encourage students to do things that we think will help them.
Here’s what I do
75% skills assessments. I try to isolate concepts so that they’re big enough that they aren’t just “do this calculation with these types of numbers”, but small enough that they provide a useful diagnostic of where to focus attention when students come up short. That whole thing is really really hard, and I’m still figuring out where the balance is. Students have the chance to retake assessments, and the new grade replaces the old grade. I also give review assessments on the most important topics, and the new grade replaces the old grade.
20% synthesis tasks. These are week-long tasks. Students work on them during dead minutes in class, and get about 10 minutes, about twice a week to focus on them. Each task takes content that we are learning, and asks students to apply it in a new way — a complex task writing a guide for a product using systems of equations, or writing a business plan for a Money Duck, or writing up rules for which pairs of function transformations are commutative and why. I care a lot about clarity and concision of writing for these, and students have the chance to turn them in to me for feedback before they’re due. No reassessments, though.
5% token homework completion grade. Don’t care if it’s right, just give it a shot.
I grade everything on the same 4-point scale that I inherited from my predecessors:
3.5: One small, non-fatal error.
3: Good understanding, but at least one significant error
2: Some understanding, but errors show significant misconceptions
1: Little understanding, work unclear, or significant confusion
0: Blank or off-topic
There are no 2.5s.
I have a system for turning these into letter grades. More or less, 3.5 is the bar for an A-, 3 is the bar for a B-, etc.
How this is going
I really like giving students their grades broken down by skill. It focuses their energy, particularly around reassessing, on what they need to focus on. I also like giving review assessments, because the message is that just because you knew how to do this once doesn’t mean you’ll remember it forever. I love the rubric. It’s simple, doesn’t pretend to differentiate between an 86 and an 88, and focuses on what students understand. Finally, it does a decent job messaging what I care about in class — you need to understand the different topics that we are learning in this course, and you need to be able to do some independent thinking and writing on more difficult questions that synthesize multiple standards.
I think this system gives me a great opportunity to give authentic feedback. First of all, it gets grading out of the way of much of class. I give students a great deal of tasks and sets of problems and activities during class, and none of these are graded. The point is to learn math, and I can give feedback, both formal and informal, that focuses them on learning math. I dont’ give a grade on the first round of synthesis tasks — just feedback on where they could be more clear or concise, point to an error or issue with their approach, or suggest an opportunity to go deeper. Then, they take that and ideally revise to improve their work.
This assessment system doesn’t do much in the way of incentives. I’m not grading very much stuff, and it’s not representative across the bulk of what we do in class. I’m happy about that — students aren’t trying to game the system — but it also means I need to engage the students each day and connect what we do in class to their outcomes on assessments. That last part isn’t easy, and often falls by the wayside as I get busy.
Stuff to make better
This is far from perfect. My synthesis tasks often miss the mark, and I want to better balance problem solving with connection to the curriculum. I’ve thought a great deal about further breaking down standards assessments, in particular differentiating between skills assessments and concepts assessments. I don’t give a consistent, clear roadmap for students at the start of a unit. I want to do more to review old standards as we move through the year, but there’s never time. The reassessment system can be time-consuming, and I’m still struggling to get the students who most need to reassess to reassess consistently. I don’t consistently provide resources for those students to make use of to prepare for their reassessments.
But this system is more consistent with my values, and how I want my students to learn math, than anything I’ve used before, and I’m enjoying immensely the challenges of making it work for as many students as possible.
Finally, a big thanks to the huge number of folks who have written about standards based grading over the last few years. I’ve read more blog posts than I can count or attribute here, but the sum of all of that is where this system comes from, and I’m immensely grateful for those ideas.