The Science of Learning from Failure

Schliessen Icon

Designing the lecture contents according to the latest trends in the learning sciences, we systematically strived to preface any instruction with a problem-solving phase in which the students were involved in historically foundational or contemporarily relevant research topics circling how (academically) failing cognitively and affectively impacts learning. Thus, combining the nature of research with scientific literature through relevant scholarly articles built the foundation of the course's theoretical aspects. Yet, instead of directly presenting the results, the student's active participation shaped the core of the classes. The students had to discuss findings, generate hypotheses, participate in in-class experiments, and plan their studies. Overall, this resulted in a large portion of time allocated to constructive debates and students interacting with their colleagues with only passive support from the lecturer, positively impacting their learning process and engagement.

Designing for Problem-Solving prior to Instruction in University Classrooms to Enhance Student Engagement

Engaging students through active participation in research                                                                                      Designing the lecture contents according to the latest trends in the learning sciences, we systematically strived to preface any instruction with a problem-solving phase in which the students were involved in historically foundational or contemporarily relevant research topics circling how (academically) failing cognitively and affectively impacts learning.

Thus, combining the nature of research with scientific literature through relevant scholarly articles built the foundation of the course’s theoretical aspects. Yet, instead of directly presenting the results, the student’s active participation shaped the core of the classes. The students had to discuss findings, generate hypotheses, participate in in-class experiments, and plan their studies. Overall, this resulted in a large portion of time allocated to constructive debates and students interacting with their colleagues with only passive support from the lecturer. To achieve this, we developed three main modes for implementing diverse problem-solving opportunities.

Problem-solving prior to instruction (PS-I) and Productive Failure

The concept of problem-solving prior to instruction (PS-I) describes the idea of advancing instruction with a phase in which students are asked to solve problems or engage in sensemaking activities (Loibl et al., 2017). How such PS-I activities are integrated into the classroom can be diverse, from contrasting cases to scaffolding problems through historical narratives. All these activities, however, have in common that they were shown to activate prior knowledge, make students aware of knowledge gaps, and help introduce cognitive conflicts. Together, this was shown to support not only the learning process but also student engagement.

Laying the foundations of PS-I, the concept of Productive Failure (Kapur, 2008) describes a particular problem-solving case, namely one in which the exercise aims not to solve it correctly but rather to explore different potential (and suboptimal) solutions. Thus, students fail to come up with the most accurate explanation; however, they establish a much broader knowledge foundation for subsequent instruction, resulting in increased understanding and transfer, in contrast to the traditional tell-and-practice structure (Sinha & Kapur, 2021).

Modes of Implementation

Structuring lecture contents and discussions
Replicating or conducting novel experiments in the classroom might not be feasible regularly for many natural science courses. Yet, integrating the way scientific findings have been made or focusing on theory building and in-class hypothesizing before presenting results and discussing potential implications might be possible for a plethora of topics.

Likewise, the seminars in the course were structured. Providing the indispensable theoretical background and the emerging research questions, students were motivated to participate in think-pair-share exercises, in which, firstly, the individual thought process is supported, followed by small group discussions before gathering the conclusions in the plenum. Alternative implementation techniques could also start with the results and ask students to consider potential implications or underlying mechanisms. Instruction can then be provided on the topics that emerged from discussions, which might help students consolidate their knowledge and engage them in following the lecture input to deepen their understanding of the previously debated topic.

Especially when time is limited, however, think-pair-share exercises might be too time-consuming. Alternative tools, such as Mentimeter, were also used to obtain an overview of the current understanding in the class and engage them in critically evaluating presented information before proceeding.

Replicating studies in classroom experiments
Another method to increase students’ engagement and promote grappling with the lecture contents is replicating real studies or parts thereof in the classroom. Whereas searching and transforming appropriate scientific studies to a format in which they can be applied in the classroom is time-consuming, the results obtained in such settings are not only attractive to discuss but might be incredibly motivating for students as they can directly participate in replicating or refuting established scientific findings. Seeing the results of a study not only summarized in lecture slides but being able to verify them brings an entirely new perspective to the topic.

«I really like interactive classes! It was very nice with Mentimeter and doing a version of the experiments ourselves.»
– anonymous feedback from a student.

From the study design to the discussion of the results
The most significant innovation in this year’s course was the design and implementation of a behavioral study as part of the course. Having presented a literature background, students had to integrate prior knowledge from what they had learned in the course to generate hypotheses about potential outcomes. The selected topic was not only connected to what we discussed in the course but also to the students› everyday lives since it covered a subject that was frequently discussed in the news but lacked substantial scientific evidence. Hence, the students were involved in research without a scientific conclusion.

Generating hypotheses concerning the outcome represented the scientific process of motivating conjectures without knowing the outcome. As such, the discussions in the class for or against the investigated method were very engaged.

In the second phase of this project, data was collected with the students of the course as well as other participants. The obtained results were then presented in the classroom and discussed, also by taking the previously formulated hypotheses into account. Thus, in this project, the students actively participated in generating knowledge, from hypothesizing to discussing the findings.

Effects on student learning and engagement
Next to the direct assessment of students’ understanding through interactive discussions, which made up more than half of the seminar’s time, graded elements were also implemented. To quantitatively assess students’ learning in the course, we chose different formats that aimed to provide the students with some self-determination on what they want to focus on. The two primary student-based outcomes were a video presentation of a self-directed project applying the concept of Productive Failure in a relatable context, such as their studies, and an essay.

The goal of the video was to motivate students to consolidate their knowledge of the course contents and demonstrate their understanding by planning and eventually conducting a behavioral experiment themselves. By giving them the freedom to choose their topic of interest (in the context of Productive Failure) and the way of presenting the knowledge they gained, the finalized videos were particularly creative and thoughtful, showcasing the students’ motivation and engagement.

A comparable concept was followed for the essays. Providing the students only minimal guidance, they were asked to write a short critical reflection on one element of the course they chose. Combining the class contents with students’ everyday life experiences was not only a primary goal of the course but also allowed students to better relate to the contents and understand their relevance.

Practical implications and conclusion
The primary practical implication of our learnings from designing the course “The Science of Learning from Failure” according to PS-I principles is to put the students in the focus of the classroom. Providing the necessary theoretical background, they should be empowered to work themselves on problems, be it a discussion of potential study outcomes, the replication of an experiment, or even the design of their own, resulting in high learning gains and student engagement.

Course Description

Name:
The Science of Learning from Failure
Description:
Failing is an inseparable part of learning. Yet, understanding how we can benefit from failing remains often unexplored. This course combines research from the learning sciences, as well as educational and cognitive psychology, to address the question of how and why we can learn from failing within the context of human learning.
Objective:
Students will engage in the following activities:
- Critically read and analyze research articles addressing learning and failure in academic settings.
- Participate in in-class problem-solving and discussion activities around the research on failure.

Upon completion of the course, students will be able to:
- Demonstrate a critical understanding of the role that failure plays in learning.
- Discuss how, why, and when failure can benefit or hinder learning.

Department:
D-GESS
Level:
Bachelor / Master / Doctorate
Format:
Seminar
Size:
60
Teaching Power:
2
Assessment:
Video Project and Critical Reflection

ETH Competence Framework

loading