There’s a major difference between wanting to accomplish something pedagogically and actually implementing it. I forget this important little fact every time I attempt something new. And every time, learning the tacit knowledge needed to implement my vision feels like recreating the wheel.
Ungrading was both the same and a different kind of experience than most pedagogical implementations (interventions?) I’ve initiated. Both more and less tacit knowledge was needed to implement ungrading. I’m not sure exactly why that was the case, but here’s my explanation of the process.
Making the first exam was fun. I ended up embedded this gif next to each question to analyze the confidence levels.
Figure 1. Confidence levels as emojis. Students needed to fill out their relative confidence in answering questions on the exam while the exam was in process.
While the “meh” (i.e. middle emoji) was a bit more unhappy than I would have liked, this set of emojis worked for the overall outcome. Letting my students know at the beginning of the exam and in the syllabus what the emojis would be helped them choose the right emoji for them on each question even if the emojis themselves weren’t ideal. And I was thrilled that I used emojis for the confidence levels (it was my wife’s idea) as my students didn’t need to spend too much time thinking about their confidence levels – it almost came second nature to them.
I also gave these instructions at the top of every page:
Confidence Levels (mostly in left margin): For each question (or part of a question), please mark your confidence level for your answer. CHOOSE ONLY 1 emoji per question/part.
The only trick with the confidence level emojis was that it started to get a little interesting in terms of space allotment when I asked multi-part problems. Where did the confidence level emojis go? Anywhere they could fit in.
Figure 2. A multi-part nomenclature question from my first exam. The confidence levels were a little confounding to place well in terms of space but my students seemed to “get” it nonetheless.
Amazingly, my students understood what was going on almost immediately with the confidence levels and answered all of them (i.e. I had 100% completion on the confidence levels throughout the entire semester).
It was almost…intuitive…
For the grading of the exams, the most difficult aspect in the entire process was giving thoughtful feedback that was critical but helpful. I also included positive feedback as that’s what I have a great deal more practice doing. Having not practiced ungrading for more than one semester, I think this is the part of ungrading that I would need to work on most. And in the midst of giving thoughtful feedback, I realized a major underlying assumption I’ve made throughout my grading life. I assumed that my students would thoughtfully consider the points I’d taken off of each question as a stand in for feedback. I made this assumption mainly because I thoughtfully took off those points, weighing exactly how much of the question had been missed and why it was necessary to take off that amount of points. My major ungrading realization was that students did not get this message. At all. My students did not even recognize this kind of grading was something we regularly agonized over until they had to do it themselves. It was only when they graded their own papers for the first time that they realized that taking off points had a reasoning behind it, and that the process of grading overall was muddy and difficult.
I returned the exams the first time with only feedback. The students had to submit an “ungrading” sheet with their returned exam in the next class session where they had detailed of how many points they thought they earned on each question and why. Many students also submitted corrections of their work because we had voted as a class whether I should provide the answer key while they were assigning their own point totals and the vote had returned a negative result. In other words, my students wanted to figure out the key on their own while they were ungrading their own exams.
God, I love my students. Even when I didn’t really care about the outcome, they made a really great choice that maximized their learning process EVERY TIME.
Before I initially returned the exams, I made a grading template for myself where I filled in the points I thought they had earned on each question. A beginning template is shown in Figure 3.
Figure 3. A beginning template for ungrading student exams. For a more complete template (or just to be able to actually see this one), please see the google sheet here.
The top row had the questions on the exam, and then the rows were divided in threes. The first row (in blue) of each triplet had the student name and my point totals for each question on the exam. The overall score column simply summed the points for each question across the row. The second row showed the student’s evaluation of their own work; their point totals were added to the sheet and, as you can see from Student 1, if they calculated their points differently, I accounted for that discrepancy. The last row in the triplet averaged my point totals and the student’s point totals for each exam question.
Between my exam score and the averaged score, I took the higher of the two scores as the student’s counted exam score (the 2nd to last column on the right). I calculated basic statistics (mean, median, standard deviation, max, and min) on each problem as well as the class set of exam scores for my exam scores, the average exam scores, and the counted exam scores. If the student’s overall exam score came within one standard deviation (calculated based on my exam scores) of my overall exam score for them, I gave them five bonus extra points on top of their final counted score for being metacognitive about their own performance on the exam vs. their peers (an ideal I *hope* my exam score reflects).
On their exams (returned for the final time), I wrote four grades: 1. YS (your score) – the exam score the students gave themselves (reiterating their work on the sheet they had turned in); 2. MS (my score) – the exam score I gave them before I handed the exam back with feedback only; 3. AS (average score) – the averaged score of their score and my score and 4. CS (counted score) – the exam score I counted with the bonus points possibly added for their score metacognition.
Overall, most students graded themselves more harshly than I graded them on the first exam. At times, some students decided that I was overly generous and actually argued on their ungrading sheets that I should take off more points because they really didn’t know what they were doing.
Some students, of course, argued for more points. But those instances were more rare than I expected when I initially took ungrading on as a methodology I would implement in my classes.
The majority of my students received the five bonus points. They were VERY savvy about their performance on the exam vs. their peers. Many of them, in fact, detailed why their performance was different than their peers on their ungrading sheet and how their relative confidence on specific questions factored into their performance.
Have I mentioned how much I love my students?
This first attempt went WAY better than I expected and, as the semester progressed, my students only learned better what helped them most in terms of ungrading. Their ungrading sheets often became multipage essays, complete with citations and arguments, about their grading and why they gave themselves the points they did. But more on that in part 4…