Students as partners in learning assessment

Co-Creating grading criteria in an alternatively graded STEM course

Photo by Hans-Peter Gauster on Unsplash

Sharon Stranford is a Professor of Biology at Pomona College and has been teaching in the biological sciences for over 25 years (previously at Mount Holyoke and Amherst Colleges). She teaches courses in cell biology, public health, and immunology. A longtime practitioner of Just-In-Time Teaching (the genesis of the flipped classroom), she’s a more recent convert to alternative grading. Sharon also served as the inaugural Co-Director of Pomona’s Institute for Inclusive Excellence, where she worked on community-building and professional development strategies to nurture inclusive teacher-scholars. She lives in Claremont CA with her wife, children, and two cats.

---

Several years ago, a conversation between two students during office hours changed the way I thought about exams and grading. One asked the other if they were ready for an upcoming exam in another class. The second student felt prepared and expected to do well, but added “I really like this topic. Someday, I’d like to actually learn this stuff.” This hit me. I remembered feeling the same way as an undergrad, telling myself that I just needed to get through this test and I could figure things out later.

That moment forced me to confront a contradiction: if my students were engaged enough to want to learn, why didn’t they feel like they could do it now and why didn’t an exam provide motivation? These questions kicked off a journey—part personal, part professional—into rethinking what grades are for, what motivates learning, and how I could do better.

Letting Go of Graded Work

The experimental part of this journey began with a leap: I stopped assigning grades in one course. No points, percentages, or letter grades on assignments or exams. Just feedback, revision, and reflection. The course was an upper level Immunology elective that typically enrolls about 24 undergraduates, mostly juniors and seniors. Pomona is a small liberal arts college with a low student-to-faculty ratio and teaching load. This, plus tenure, gave me the time and space to try something different.

Each week, I gave students detailed feedback and opportunities to revise responses to a set of online questions. They could also resubmit each of three exams, which we called summative assessments. I met one-on-one with students after each assessment, to discuss the immunology questions and reflect on their learning. At the end of the semester, they proposed a final grade.

I learned a lot. Students were energized and less stressed by the no-grades format. But they also had concerns. This was articulated in a collective debrief near the end of the semester. They asked “how do we assign grades that reflect growth? Are there ways to reward contributions that happen outside of class?” And, my favorite: “Isn’t this a bait and switch - ignore grades all semester and then ask us to assign them at the end?”

They were right. Grading can be opaque, and not talking about it didn’t make it clearer. Most traditional grading systems, especially in STEM, don’t account for growth, risk-taking, or time spent teaching peers, all things I wanted to encourage. While students appreciated the autonomy, they also wanted more guidance. Their advice? Keep the class structure, without grades, but engage students in the process of defining how learning will be assessed.

Students as Partners in Creating a Learning Rubric

The following year, I invited students to help build self-assessment criteria - essentially, co-creating a grading rubric. Informally, we called this “collaborative grading,” although that term is often used to describe what we were already doing, where student and instructor collaborate to settle on a student’s final grade¹. I wanted to go a step further and ask students to collectively engage with me to define grading expectations up front. We started by asking: What does learning look like in this course? What does it feel like? Students talked about mastering content, applying concepts, asking good questions, wrestling with hard problems, making mistakes, helping teammates, sticking with it, and so on.

From that discussion, a framework with three categories emerged – effort, mastery, and participation. We worked together to add specifics within each category, plus a fourth category - personal goals (this is something students propose individually, at the start of the semester). Students debated everything from what sufficient attendance meant (they landed on 90%, which was stricter than what I would’ve set!) to how to handle life disruptions fairly. We discussed accountability, flexibility, and how each student could stay engaged even if they missed class. A key agreement was personal responsibility for catching up and continuing to contribute meaningfully to their peer learning teams. For context, I assign students to stable teams of four in week two of the semester. These groups work together inside and outside of class, especially when they’re discussing weekly assignments or tackling exam questions. So, these teams are important learning tools and something I wanted students to invest in.

We continued shaping the rubric collaboratively on a shared Google Doc. Within two weeks, we had a version that everyone could support². It reflected collective wisdom and compromise, setting clear expectations while honoring individual agency. Every year since, we’ve started with the previous year’s criteria and updated it together to fit the current group. The process is iterative and dialogical - students propose ideas, I offer feedback, and we refine it together, sometimes coming back later in the semester to make adjustments. This has evolved into a formative conversation with students, or series of conversations, highly instructive for their success in the class.

The Self-Assessment Rubric in Practice

The rubric became a reference point, central to collective and individual behavior, as well as conversations with students. For several years, I’ve held two-stage exams. In stage one, students submit answers to a set of take-home applied questions. For stage two, peer teams discuss the exam material, after which each student can submit revised responses. After creating our assessment rubric, we added a set of metacognitive self-assessment questions to stage two, explicitly asking students to use the rubric to evaluate their work - both on that assessment/exam and the course overall. I also added individual meetings with each student, to discuss their final exam responses and answers to the metacognitive questions.

The rubric thus served as both a mirror and guide. It encouraged students to recognize less visible contributions to collective learning—like organizing study groups or helping a teammate understand a concept. It also created structure for individualized reflection, particularly around personal goals and growth. One-on-one meetings with me added depth and nuance. These conversations often surfaced insights that can get missed in traditional assessment schemes.

For instance, during individual meetings I could probe student responses to the immunology questions. Not infrequently, a student would say to me “Wait, I said that in my written response? What I really meant was…,” then proceeded to explain with surprising depth, sometimes even nailing off-the-cuff follow up questions. Other times, the reverse happened - their writing was spot on but they couldn’t explain why or apply it to new scenarios. In other words, we both knew pretty quickly how deep their learning went and whether they had achieved mastery, and we had agreed on this expectation in our rubric. By the end of that conversation, we typically agreed on what grade they would assign themselves.

Sometimes in these conversations, I had to convince them they deserved an A! It was hard for them to let go of “I didn’t get it right the first time” and the need to penalize. Other times, they could see that their knowledge was still skating the surface. In that case, we talked about opportunities for improvement. Often, these were places where they dropped the ball in our rubric.

What Students Said

After a couple of years of using this format of collaborative grading, I wanted to study the impact on students. So, I gathered feedback in post-exam reflections and an end of semester anonymous survey. The following themes emerged, with representative quotes³:

• More motivation and ownership- “I found myself studying to understand, not just to get the right answer.”

• Deeper self-awareness - “The self-assessment criteria helped me see how I was growing—not just as a student, but as a learner.”

• Reduced stress and more risk-taking - “Not having to stress about grades made me more willing to ask questions and take risks.”

• Greater trust in the system – “The biggest advantage [in our grading system] is that it is a much more fair and representative assessment of performance in a class.”

Students also valued the one-on-one meetings, though some admitted they were intimidating at first. Several asked for more guidance on how to prepare for these conversations - something we now address early in the semester.

Lessons Learned

These whole class discussions about grading and my one-on-one conversations with students have become some of the most meaningful teaching/learning moments in my career. This chance to “peak under the hood” of grading with my students always surfaces surprises and insights. In individual meetings, we get to connect on a human level and I discover their learning tricks. And, for the first time, I feel like I had a more authentic sense of what they’ve learned and whether they can apply it. Anecdotally, many students tell me their learning feels more durable and adaptable.

Of course, no approach is perfect. A few students - especially those accustomed to A’s - were uneasy without frequent graded benchmarks. Some struggled with perfectionism, unsure when work is “good enough.” To help address this, we amended the rubric, setting an 85% accuracy benchmark for short answer/multiple choice questions, and 90% for essays. They knew when they hit mastery because I gave them a “thumbs up” to that question, without any comments. Finally, they found the self-evaluation process uncomfortable, though most agreed it had value. In anonymous surveys, over 70% felt that the advantages of our system outweighed any disadvantages.

Final Thoughts

I’ll be honest. This collaborative grading model took more time and required more trust - from me and my students. I spent more time on feedback and I gave up complete control over grading. Students spent more time revising responses and reflecting on their learning. They also began to notice that the learning cycle didn’t work if they wouldn’t allow themselves to make mistakes. Those individual conversations with me encouraged them to (begin to) embrace that discomfort, and trust that I wasn’t trying to catch them out. Rather, I was trying to help them learn. Assessment therefore became a little more human, a little less transactional. And I noticed that talking with my students about assessment helped them develop more insight and motivation as learners. Plus, it clearly helped them learn a lot of immunology!

1. See this collaborative grading post for better definitions.

2. See this self-assessment criteria from fall 2024.

3. To further complicate the alternative grading name game, in early 2025 I published a paper describing this process of co-creating assessment criteria with my students. The results came from a class that I described to my students as both “ungraded” and “collaboratively graded,” a term suggested by the first group that engaged in co-creating an assessment criteria. Therefore, both terms show up in student responses to the survey. This was before I was aware of the use of collaborative grading to better describe the more “traditional ungrading” we had been engaged in previously. I know, a confusing alphabet soup of terms! If you’re interested, that paper also includes more on the structure of the class, a rubric from my 2022 Immunology class, and an analysis of results from student surveys.