The divide between teaching and learning and traditional assessment is real. Many of us think about the relationships we could develop and the authentic learning students could engage in if only we removed could grades. I’m here to tell you how the gradeless classroom can be more than an “if only…” fantasy. Instead of being a desert that remains just out of our reach, my journey shows how you can discard the grades and skip right to the good stuff.
First, let me tell you about an intermediate level programming course I teach at Northwestern University. Here is how the course works:
- You select an exercise and send me a working solution.
- I critique your solution and return it to you.
- You revise and resubmit your solution.
- We repeat this cycle until I have no more critiques.
- You repeat this process until the course ends.
I do not grade your solutions. Either a solution is fine, or it needs more work.
At the end of the term, your grade is based on three factors:
- Progress: How far did you get? How many different kinds of problems did you do? How difficult were these problems? How many different topics and skills did you explore?
- Quality: How good is your code by the end of the course? What level of critiques are you getting?
- Effort: How often and how steadily did you submit, including revisions? How hard did you push yourself from your personal starting point?
I call this critique-based assessment. It’s been my solution for how to go gradeless since 1997. It’s the model for all the online software courses I have designed and delivered for a company I work with. Twenty years without handing out and arguing about grades. How has this turned out for me and my students?
For my students…
It’s no big deal. This continues to surprise me. This is unlike any other course they take at Northwestern, and yet no student has said the class should switch back to the traditional graded model. No one has complained about lack of transparency in grading. What they (correctly) tell other students about is the importance of continuous resubmissions. For my students, “one and done” is never enough.
Gradeless doesn’t mean being in the dark about your progress. Northwestern requires final grades. My students care about those grades. They complain about teachers who are slow to return graded quizzes and homework assignments. How do I get around this problem? Technology! I developed a web application to store submissions and critiques. At any time, students can view a summary of what they’ve done so far: what they’ve submitted, what’s been completed, and how long since their last submission. They can also see how their statistics compare with the rest of the class.
Students no longer send me “how am I doing” emails because they know where they stand. Some ask “what should I focus on to improve?” That’s a question I’m happy to answer.
My interactions with students are centered around content and revision. I never worry about whether a submission is a 7 or an 8. I just discuss what needs to be fixed. This means no more arguments with students about points. Instead, students and I have worthwhile discussion about their feedback. I never feel like I am throwing feedback into the wind. Students don’t need to agree with every piece of my feedback, but they do need to give it a shot and show me the result. I never worry about being too abstract because my I link the feedback to specific examples in the student’s submission.
I have a more complete grasp of what my students do and do not understand. When I first began this process, I wrote down a dozen critiques I expected I would need. These critiques focused on general standards for good code, such as clarity of names, modularity, and documentation. But once I started critiquing actual submissions, I found that I needed to add a dozen repeating critiques. By the fall 2016 semester I had re-used over 500 distinct critiques for submissions from 86 different students.
I don’t often grade, but when I do it’s fast and fair. Until the end of the term, I don’t think about grades at all. When it comes time for grades, thanks to technology, I have a complete history of all submissions and feedback. I know who has submitted what, the amount of topics each student covered, and the summaries each critique received.
With this mix of details and summaries, assigning grades takes approximately an hour, even with 80 students and several thousand submissions. The process seems more equitable than traditional grades, and students seem to agree. I receive, at most, two to three post-course grade queries from students. During the last term I received no questions about final grades at all.
About that cake…
Teaching without grades is my cake. Every day I engage with students about how to improve their programming skills, and I never have to derail those conversations with tangents about who has what grade and why. Fortunately, I have found a way for my students and the university to have their grades while I have my cake.
Christopher Riesbeck is an Associate Professor in Electrical Engineering and Computer Science at Northwestern University, with a courtesy appointment in the Learning Sciences. He has a PhD in Computer Science from Stanford University.