Pedagogy of Risk: Turning Failure into Learning Opportunities

This article is a translation of a text published on Eductive’s French website.

William’s story

Fear of failure

The idea of “failing forward” first caught my attention when I was a Teaching Fellow at Newcastle University. My pedagogical training led me to reflect with honesty upon my own experiences as a student. I realized how terrified I had always been of failure. During my studies, my fear of failure would sometimes manifest as feigned disinterest in an assessment (you can’t fail if you don’t try…), and at other times, conversely, through an obsessive dedication to my work. Whatever the case, my fear never disclosed itself in a healthy manner.

In the context of these reflections, I discussed with my students their own perceptions of failure. Without exception, they all admitted fearing failure at various levels: fear of failing a specific course, exam, or assignment. This fear of failure can sometimes result in a fear of the discipline being evaluated, which can directly impact students’ future success in their program.

Bouncing back after failure

In an attempt to eliminate (or at least reduce) the fear of failure, I wanted to encourage my students to embrace the idea of “failing forward”, of learning from their mistakes. To achieve this, I had to rethink the way I designed assessments.

The idea is simple enough in principle: give students multiple opportunities to take an evaluation and provide formative feedback each time they attempt it. In this way, failure becomes an integral part of the learning process. Students begin to see failure as an opportunity to improve their understanding. Moreover, since students know they will have the possibility of another attempt the following week, their fear of assessment is diminished. Of course, in practice it’s more complicated than it sounds, especially when it comes to maintaining the rigour of the assessment process.

Application in an introductory programming course

I tested the idea in an introductory programming course in the computing degree program I led at Coventry University. In this course, the assessment consisted of 2 parts: an exam and a project, both focused on the essential coding skills and basic programming knowledge necessary for students to progress in their studies. However, I noticed that several students were passing the course without truly mastering the fundamental concepts. They would pick up marks here and there by demonstrating an understanding of secondary (peripheral) elements assessed via the project, and this could sometimes be enough to get a passing grade (which is set at 40% in the United Kingdom).

My colleagues and I compiled a bank of 100 exam questions covering the fundamental concepts of the course. The questions were uploaded into Moodle in various formats (multiple-choice, numerical input, text input, drag-and-drop, etc.). The weightings of the questions varied based on their complexity. For each question, a detailed explanation of the answer and a walk-through of the required thinking were provided.

Starting in the 5^th week of the 12-week course, students could take a 20-question exam, with questions randomly selected from the question bank, for a total of 100 points. The passing grade was set at 70%, and a double threshold was applied to the course: to pass the course, students had to score at least 70% on the exam and at least 40% on the project. If a student scored 70% on their 1^st attempt, they passed! They did not need to “retake” other versions of the exam in the following weeks, but could do so if they wished to improve their grade. If a student scored below 70%, they could “retake” the exam the following week (or wait another week, as retake opportunities ended in the 10^th week of the course) to try to pass. Thanks to the formative feedback provided, each attempt became an opportunity to improve. When a student did not take the exam in a given week, they could use the time to work on the project.

Final project evaluation

For the evaluation of the final project, I provided my students with a detailed grading rubric so they could estimate their grade before submitting their work. Each criterion in the rubric included a list of elements that needed to be present in the computer program developed by the students. By checking which elements were present in their program and how they functioned together, students could determine the grade they would receive.

I asked students to indicate on the rubric which elements had been included and to submit the rubric along with their work. I then validated the content.

Students were also given 3 opportunities during the semester to receive formative feedback on their project. Either another member of the teaching team or I would review the work with the student, asking for their thoughts on each element of the project, and providing guidance on which aspects to focus on next to produce the best possible result.

Outcomes

The results met our expectations:

• All students took the exam at the 1^st opportunity in week 5. No student chose to avoid taking the exam, not even the weaker students, who might have been tempted to postpone a difficult assessment until the last minute.
• Students who performed poorly on the 1^st attempt performed better on each subsequent attempt.
• The failure rate for the course dropped from 25% to less than 10%.
• By the end of the course, students demonstrated better core programming understanding than some of their 3^rd-year peers who had passed the course on their 1st attempt 2 years earlier.

I believe this pedagogical approach, particularly in programming, reflects real life. Coding is an iterative process: you write code, compile it, identify the errors, fix them, compile again, and repeat the process until there are no more errors. Then you write new code and start the process again.

If adopting a similar approach can both reduce students’ fear of failure and, more importantly, encourage them to develop an interest in the more challenging topics within their studies rather than avoid them, then I believe it’s worth the effort!

Marie-Gervaise’s story

The right to make mistakes

When one learns a second language, mistakes are inevitable. A person learning a second language may hesitate over certain words, but if they want to communicate, they have to take risks and dare to form a sentence. The person they are talking to will do their best to understand and, if the message is not clear, will try to confirm their understanding by rephrasing the statement, and so on. In one sense, learning a second language involves constantly “failing forward”: one learns from one’s mistakes.

In second-language learning, therefore, risk-taking is essential. However, to be willing to take risks, students need to feel competent. Additionally, the risk must feel worth taking; it should be a controlled risk.

I believe that grades are an important factor in inhibiting students from taking risks. I teach many students who aim for an excellent R score and, as a result, do not allow themselves the right to make mistakes. In a way, this can make the classroom atmosphere a bit tense. Over the past 4 or 5 years, I have been gradually working on integrating practices that encourage risk-taking and restore the right to make mistakes.

Ungrading, metacognition, and scaffolding

To encourage students to take risks, I believe the ungrading approach is a promising avenue to explore. With ungrading, the focus is not on grades, but rather on the intrinsic value of the tasks to be completed and the importance of the feedback they provide.

Along the same lines, I try to foster metacognition in my students. I give them tasks that focus on personal growth and self-awareness rather than the accumulation of marks:

• I let students choose the topic for their work so they can select something they feel is meaningful.
• I encourage self-assessment by asking my students to annotate their work before I grade it. I then respond in writing to their annotations, either confirming or refuting their self-assessment. (In a way, this is a written version of the meetings that many teachers who practise ungrading have with their students, applied to a specific assignment rather than the final grade.)

I also rely on task scaffolding: I organize assignments into building blocks. Feedback on one task can help the student perform better on the next. For example, a speaking exam might cover the same topic as the following written exam (or vice versa). This way, my feedback on what the student expresses orally helps them structure their written work. I provide guidance and tips for the next step. Some reading quizzes also serve as personal notes and authorized preparation for exams. Both the midterm and final exams require long-term preparation: several prior assessments help students prepare for them step by step. (I have already discussed my approach in an Eductive featured report I wrote about preventing plagiarism and cheating in the age of AI.) In my advanced-level courses (103), the midterm itself is reused in the final exam: students bring their midterm to the final, rework it and reuse it.

Multiple assessments and opportunities for retakes

Paradoxically, to implement ungrading, I have chosen to include a very large number of graded assessments in my course; my students receive many grades! The idea is that by multiplying assessments, each worth only a few marks, the grade for each individual task becomes less important than the act of completing the work for its own sake. For students who are gamers, it can even seem like fun, “farming” points for the end of the semester. In reality, it’s their learning that they are cultivating. In addition, the homework assignments I give students are sometimes integrated into a system of “entry tickets”: they are graded for just a few marks (sometimes as little as 0.5), but completing them is essential for access to the learning community, especially when the assignment is part of a series of scaffolded activities.

Furthermore, students have the opportunity to correct themselves when they make mistakes. For example, for listening or reading comprehension skills, students can retake an assessment if they wish. There are 3 listed in the course plan, but I only count the 2 highest grades. Students can choose to take all 3 or only 2. The same approach applies to some at-home reading quizzes in my advanced course: I count the students’ 7 highest grades out of 11.

Obviously, the grading workload is quite heavy. However, I manage to keep it reasonable by designing self-graded assignments using Microsoft Forms. With this approach, the total time spent on grading is no greater than before.

Outcomes

Since I’ve been teaching this way, the classroom atmosphere has become more positive. Students realize they have the right to make mistakes. They quickly understand that a lower grade on a given assessment is not the end of the world: each assessment has limited weighting, and there are opportunities for retakes. On top of this, there are numerous formative assessments in class, both written (often on the whiteboard, since I teach in an active learning classroom) and oral (during in-class conversation practice).

As I had hoped, I’ve noticed that students are more engaged and willing to take risks. As a result, by the end of the semester, overall averages are higher than before, even if some grades were lower along the way. Most importantly, progress in learning, especially among intermediate students, is more noticeable.

Students can experience failure on an assessment and “live with it” for a few weeks without the situation becoming unhealthy, knowing they will have the chance to make it up. I believe this is both positive and healthy. As a learning support counsellor at my college once told me, we cannot and should not completely prevent our students from feeling anxious, but we can help them learn to live with a certain level of discomfort and accept it.

Moving forward

With regard to ungrading, I read Jesse Stommel’s book “Undoing the Grade: Why We Grade, and How to Stop” with great interest.

I have also learned about the experiences of Caroline Cormier, Bruno Voisard, François Arsenault-Hubert, and Véronique Turcotte [in French], who teach chemistry at Cégep André-Laurendeau, and I found them very inspiring. I would love to adapt their assessment approach to my own discipline … to be continued! In particular, the achievement levels they use (“In Progress,” “Almost There!,” “Mastered,” and “Advanced”) seems highly relevant to my discipline. I was especially intrigued by how they convert learning objectives and their achievement into a numerical grade, since we still need to provide one at the end. In short, I’ll keep reflecting on these alternative grading practices, and I hope to incorporate more of them into my assessments over the coming semesters!