As you can tell from our last few blog posts, it’s midterm time here at Grand Valley State University. Midterms are a good time to reflect on how classes are going and make mid-course adjustments, and that’s especially true when you’re trying a new assessment system.

This semester is the first time I’ve dipped my toes into *ungrading*. Last summer, I read Susan D. Blum’s Ungrading and was inspired to take the plunge and go fully gradeless in my Euclidean Geometry class. In this class, I give *only* feedback on student work — no marks at all. Students will propose their own final grades, with a portfolio of work that supports their proposal.

So, this week’s post is my honest and (hopefully) helpful reflection on how something brand new is working for me.

### What is this class, anyway?

My Euclidean Geometry class is really two sections with a total of 30 students. This is a junior-level class for pre-service math teachers, which is a unique (and wonderful) student population. By this point, the students have spent two years in our math education program. They are comfortable with each other as a cohort, and are *very *used to active involvement in class.

On the first day of class, I can — and do! — walk in, spend about a minute introducing myself, and then say “All right everybody, find three other people, go to a whiteboard, and start working on this problem.” And they *do* it, really well!

This gives me a lot of freedom. I use a guided inquiry framework that helps students learn about the process underlying Euclidean geometry. They attempt new constructions and proofs before class. Class time is mostly students presenting solutions to each other, asking questions, suggesting fixes, and working in teams to generate ideas as needed until everyone is satisfied with the result. I work hard to build a collaborative atmosphere where the goal is to build a common understanding.

Assessments are mostly focused on “homework”: Each week, students individually write up one problem that was previously presented. The idea is to demonstrate their *individual* understanding after our very collaborative class time.

Students can revise and resubmit one homework problem each week, including a reflective cover page. In my experience, class is filled with feedback loops that engage deep learning: Presenting and editing work on a problem is one loop, but writing and revising an individual write-up is especially effective.

I’ve previously used Specifications grading in this class. To earn an A/B/C/D or F, students had to earn a certain number of “Exemplary” and “Meets objectives” marks on homework, present a certain number of problems during class, and consistently complete some other ungraded work (such as completing pre-class work).

### What does ungrading look like for me?

The central goal of ungrading is to allow students to focus on growth and learning without the perverse incentives provided by grades. In my case, this means two things:

First, **there are no marks on any class work**. Most class work was already focused only on completion (pre-class work, presentations). For homework, I’ve always focused on giving detailed feedback, written in terms of specific objectives for each assignment. This semester, I just leave out the unnecessary final mark and leave *only* feedback.

Second, **final grades are proposed by students and supported by a portfolio of their work**. Gone are all hard cut-offs and required numbers of marks from my Specifications approach. Instead, I created a list of grade criteria: Descriptions of the types of things students can learn to do in this class. Some are focused on mathematical content

… and others focus on collaboration and other ways to engage with class:

Then I wrote some descriptions of final grades, using these criteria:

To earn an A,consistentlymeet almost all of these criteria (missing at most 2 or 3), including all of the bold criteria. Show meaningful progress on most criteria that you don’t fully meet.

Students will submit a final portfolio. In it, **they will include an essay that argues for a certain grade**. They will provide artifacts, such as homework or other work, that shows how they’ve met these criteria. There’s intentionally a lot of flexibility in the grade descriptions, giving students multiple ways to find their own path while completing a few critical items (the bold criteria).

My plan is to have students complete two “check-in” meetings in which we discuss their progress (I’ve previously done a version of this, even in the Specs-graded version of class). The goal is to get students comfortable with — and thinking about — the grade criteria.

We completed the first round of check-ins two weeks ago. We ended up doing these “meetings” mostly in written form: I asked students to describe their progress on each of these criteria, indicate what grade that progress pointed towards, and then make a specific plan for what they need to do for the rest of the semester. I responded with written comments and a quick (recorded) verbal response to their overall plan.

Students could request a Zoom meeting to discuss the criteria live, and I also asked a few students to schedule meetings if I wanted to make sure we discussed certain topics. In each case, I gave specific feedback on their progress descriptions, and if needed, suggested steps they needed to take. However, I found that I generally agreed with their self-assessments and occasionally suggested that a student might be doing more than they were giving themselves credit for.

The next round of check-ins will happen in a few weeks. This time, we’ll start practicing directly for the final portfolio: I’ll ask students to find a few initial portfolio artifacts and explain how they help meet some of the criteria.

### What do students think of ungrading?

Whew, that was a lot of description. Thanks for sticking with me.

In addition to the check-in meetings, I also asked students to complete an anonymous survey this past week. I asked questions like “What is something that is working well for you in this class?” and “What is something you’d like me (Dr. Clark) to change in how I run this class, its policies, etc.?”

From the anonymous survey, it seems like **the lack of grades is barely on anyone’s radar**. Exactly one survey response out of 30 even mentioned grades, and that only asked about a technical detail. Normally I get tons of comments — positive and negative — about grades and requirements. This time, silence.

In check-ins and meetings, students said a bit more. Overall, they like the list of criteria. They appreciate the clarity it provides, and found it helpful as a way to direct their efforts. They like the lack of pressure to “perform” and earn certain grades. Many had clarifying questions and made suggestions for edits to the grade criteria, which led me to create an updated and improved list that I shared with students.

But overall, student comments focused mostly on the *geometry* and what they’re *doing *and *learning* in class. They talked about how they were collaborating, how their teams

**More focus on learning, less on grades.**

### What do I think of ungrading?

From my point of view, some curious things have been happening. Even though students aren’t *saying* a lot about grades, I’ve observed some interesting changes in their behaviors:

**Students are focusing more on learning and growing as students of geometry.**My students seem to find much more motivation when they don’t have the fear of punishment (by a poor grade) or concern for achievement (of a good grade) hanging over their heads. I’ve realized that even a few-level grading system (like Specifications grading) can provide awkward incentives. Arbitrary cut-offs, like “complete 10 homeworks that meet all objectives” are meant to encourage a certain level of achievement and understanding. But they become goals for their own sake, with the focus entirely on getting to the magic number.One way that I’ve noticed this motivation is also one of the most startling changes:**students are revising homework****much****more often**, more thoroughly, and more thoughtfully. I use the same homework objectives as in previous semesters, and all feedback is written in terms of those objectives — but many students are pushing themselves far beyond what I used to require for “exemplary” work, much less “meets objectives”. I asked a student about this, pointing out that their work was already quite excellent (and that I’d already said as much to them). The response was telling: “Yes, but I know I can do better.”**Overall, I see more intrinsic motivation to learn, and less extrinsic motivation to meet grade requirements.****Narrative grade criteria give students clear directions for improvement.**Of course, this is a major reason for writing narrative grade criteria, but I’ve been amazed at how well students are acting on these criteria. One midterm check-in meeting with a student went like this: I read each of the grade criteria out loud, and she would either tell me how she’d been meeting it, or else explain what she could do to meet it in the next half of the semester. Repeat with each of the criteria, until she had spelled out her own clear plan to succeed in class. My contribution was to clarify what some of the criteria meant, and occasionally convince her that she was already meeting a criterion more than she believed. Other students made it clear that they’re using the grade criteria to direct their efforts, too.**The lack of grade-based incentives just… doesn’t seem to matter.**Many of us — myself included — sometimes think of grades as necessary for motivation. For example, in order to convince students to volunteer to present their work, I used to require each student to complete a certain number of presentations. This semester, this is rolled into two criteria, with no numbers attached: “Share your ideas by giving presentations […] that clearly communicate your ideas.” and “Volunteer to present your work.” Students are volunteering, as a whole,*much*more often*and*fewer students are avoiding volunteering. In addition, presentations are as good as ever. There is, after all, a lot of social pressure to do a good job in front of peers.

The examples I’ve given above of student revising homework more often also seem to be in line with this. The hard cut-offs for earning each grade did provide some extrinsic motivation, but it always felt like coercion. After I removed those cut-offs, students have found their own (better) intrinsic motivation.**Goodhart’s law is real.**Most of the changes I’ve noticed seem to be connected to Goodhart’s law: “When a measure becomes a target, it ceases to be a good measure.” I could see this in action in my old Specifications grading setup: Once students had completed the number of presentations required for an A, they often stopped volunteering. Once they had revised enough homeworks up to “Exemplary” level to qualify for an A, they stopped revising. In other words, their goal wasn’t necessarily to*learn*(as demonstrated through these measures), but rather to*meet the target*.

There is a question I’ve thought of, and perhaps you have too: Since I’ve removed the hard numerical cut-offs, **are students doing more work because they’re not sure what is “good enough”?** That is, are some students revising more often, volunteering and presenting more consistently, and asking for more revision opportunities because the target for “A-level” is too ambiguous, and they are over-estimating the work required?

So far, I *think* the answer is no — but this is something I’m monitoring carefully. I’ve noticed that students consistently push beyond my expectations *even when I explicitly tell them that they are doing great* and don’t need to (for example) revise that one tiny typo in an already exemplary proof. In addition, the question of “exactly how many do we need to do?” has *not* come up, in the anonymous surveys nor in the midterm check-ins. This is another way in which it looks to me** **like** students are developing more intrinsic, rather than extrinsic, motivation.**

Another reason is that the grade criteria are still clear and actionable, and students have done a good job of asking for clarifications when they need it. It’s just that the criteria don’t include numbers or marks any more. Instead, they describe skills students can develop, and actions they can take.

All of that said, I’m paying close attention to how students are working, and I expect that more questions might arise as the end of the semester approaches.

### Once more: Trust

As you can tell, I’m pretty happy with my first attempt at ungrading so far. I’ll also be the first to admit that this is probably the easiest possible class to use ungrading in. In addition to a student population that is already collaborative, active, and energetic, my students are used to all sorts of interesting assessment systems that my colleagues and I use. They’re willing to give me the benefit of the doubt and try things out, even if it seems a bit weird.

But in the end, that’s what all of this is about: *trust*. As Robert said in last week’s post, *When we talk about “getting buy-in”, we are really talking about trust. *In this class, students are willing to trust me with this ungrading thing partly because they’ve come up through a program filled with caring, hard-working faculty who use a wide variety of approaches to instruction and assessment. They’ve helped undo some of the trauma and fear of the “gotcha” that so many students suffer in their math classes. That makes things easier for me, but it’s not enough on its own.

I’ve also worked hard to deserve that trust. I try to be transparent and clear in my policies, to talk with students as adults, to trust *them* and believe *them*, and to ask them for help if things aren’t working out.

All of that together is helping to make this ungraded class work out pretty well. I’ll check in again and let you know how things go at the end of the semester.

*Click here to receive Grading for Growth in your inbox, every Monday.*

After our first round of check-in meetings, students proposed changes to the criteria, and I added comments to indicate what I had changed — hence the yellow highlights.

For everyone who’s asked, here are the complete descriptions. As you can see, the list of criteria are the same for each grade — what varies is which ones, and to what extent. While I tried to avoid hard numerical cut-offs, as you can see, they’re still lurking around the edges.

After the first few weeks, I assigned permanent teams for in- and out-of-class work. I createed these with student input, including making sure that each team has a time they can all meet. One question on this week’s anonymous survey asked if students wanted to keep the same teams or change. Results: 88% “keep the same teams”, 8% “change teams”, and 4% “no preference”.

I am a high school math teacher. My brother, a history professor at UNC-Chapel Hill, sent this to me and asked me if I liked this approach. My response was: "Yes, very much so. I'm going to read it a few more times. The one thing that concerns me is the reliance on homework. Homework is different in college than in high school. My students have 7 classes and spend every weekday at school, after which many of them have extracurricular activities and/or sports. College students have a lot more discretionary time. I do my best to keep the homework load (which I call practice problems and which I don't collect or grade) to a minimum. Still, I like this approach and could probably figure out a way to implement the essence of it." Have you any thoughts how this might work in a high school environment?

Glen Janken

This is awesome, and I appreciate you doing this. Could you post all of the grade criteria (not just for an A)?

Keep us posted on how this goes!